Island invasives:
scaling up to meet the challenge
Proceedings of the
international conference on island invasives 2017
Edited by C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West
Occasional Paper of the IUCN Species Survival Commission Nº 62
Island invasives:
scaling up to meet the challenge
Proceedings of the
international conference on island invasives
i
ii
Island invasives:
scaling up to meet the challenge
Proceedings of the
international conference on island invasives
Edited by: C.R. Veitch, M.N. Clout, A.R. Martin,
J.C. Russell and C.J. West
Occasional Paper of the IUCN Species Survival Commission No. 62
iii
The papers and abstracts published in this book are the outcome of the Island Invasives 2017 Conference co-hosted by
the University of Dundee and the South Georgia Heritage Trust, held at the University of Dundee, Scotland, from 10 to
14 July 2017.
The guidelines for this conference were: “any topic relating to invasive alien species on islands, where the term ‘island’
is broadly interpreted and (rather ironically from a classical perspective) may include a submarine island – e.g. a coral
reef. The invasive species involved may be flora or fauna. Particularly encouraged were papers that relate to the theme
of the conference – scaling up to meet the challenge – or to either biosecurity/quarantine or post eradication impacts on
native biota.”
The editors thank Carola Warner for her assistance with editing of many papers.
All papers have been peer reviewed and we thank all reviewers. The content of the papers is the choice of the authors. The
style of presentation has been modified in consultation with the editors. Nomenclature follows international published
standards.
The designation of geographical entities in this book, and the presentation of the material, do not imply the expression
of any opinion whatsoever on the part of IUCN concerning the legal status of any country, territory, or area, or of its
authorities, or concerning the delimitation of its frontiers or boundaries.
The views expressed in this publication do not necessarily reflect those of IUCN.
This publication has been made possible in part by funding from the South Georgia Heritage Trust.
Published by:
IUCN, Gland, Switzerland
Copyright:
© 2019 IUCN, International Union for Conservation of Nature and Natural Resources
Reproduction of this publication for educational or other non-commercial purposes is authorised without
prior written permission from the copyright holder provided the source is fully acknowledged.
Reproduction of this publication for resale or other commercial purposes is prohibited without prior
written permission of the copyright holder.
Citation:
Veitch, C.R., Clout, M.N., Martin, A.R., Russell, J.C. and West, C.J. (eds.) (2019). Island invasives:
scaling up to meet the challenge. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN. xiv +
734pp.
ISBN:
978-2-8317-1961-0 (pdf)
978-2-8317-1962-7 (print)
DOI:
https://doi.org/10.2305/IUCN.CH.2019.SSC-OP.62.en
Editor in chief:
C.R. Veitch (all correspondence with IUCN, authors, referees and editors; typography)
Editors:
M.N. Clout, A.R. Martin, J.C. Russell, C.J. West (quality control of content)
Copy editor:
C.J. West (application of IUCN style, checking typography, spelling and grammar)
Cover design:
C.R. Veitch
Cover photos:
Front cover, centre: Antipodes Islands, J. Doube. Clockwise from top right: Pleurophyllum hookeri,
Macquarie Island, M. Houghton; South Georgia pipit, A.R. Martin; Norway rat, A.R. Martin; Miconia
calvescens, Oʽahu Invasive Species Committee; Henderson crake, S. Oppel; Macquarie Island,
B. Horne; Eilean an Tighe, Shiant Isles, John Tayton; Macquarie Island, B. Horne; Ramsey Island,
G. Morgan; forest ringlet butterfly, C. Beard; Anolis desechensis, J. Herrera; mouse eating wandering
albatross chick, S. Schoombie; mink with gull chick, T Kolaas. Back page: South Georgia Island,
A.R. Martin.
Layout by:
C.R. Veitch
Produced by:
South Georgia Heritage Trust
Printed by:
Winter & Simpson, 16 Dunsinane Ave, Dunsinane Industrial Estate, Dundee DD2 3QT
Available from:
South Georgia Heritage Trust
Verdant Works
West Henderson’s Wynd
Dundee
Scotland DD1 5BT
United Kingdom
Tel: +44 1382 229792
Email: info@sght.org
www.iucn.org/resources/publications
iv
Contents
PREFACE
Addressing the challenge
A.R. Martin, M.N. Clout, J.C. Russell, C.R. Veitch and C.J. West. ....................................................................... xiii
OPENING ADDRESS
Her Royal Highness, the Princess Royal
Princess Anne ............................................................................................................................................................1
KEYNOTE ADDRESS
Protecting the biodiversity of the UK Overseas Territories
Lord Gardiner of Kimble ...........................................................................................................................................3
PRESENTED PAPERS
Chapter 1: Rodents
A potential new tool for the toolbox: assessing gene drives for eradicating invasive rodent populations
K.J. Campbell, J.R. Saah, P.R. Brown, J. Godwin, F. Gould, G.R. Howald, A. Piaggio, P. Thomas,
D.M. Tompkins, D. Threadgill, J. Delborne, D.M. Kanavy, T. Kuikin, H. Packard, M. Serr and A. Shiels ........6
Black rat eradication on Italian islands: planning forward by looking backward
D. Capizzi, P. Sposimo, G. Sozio, F. Petrassi, C. Gotti, E. Raganella Pelliccioni and N. Baccetti ........................15
Control of house mice preying on adult albatrosses at Midway Atoll National Wildlife Refuge
M. Duhr, E.N. Flint, S.A. Hunter, R.V. Taylor, B. Flanders, G. Howald and D. Norwood .......................................21
Eradicating black rats from the Chagos – working towards the whole archipelago
G.A. Harper, P. Carr and H. Pitman .......................................................................................................................26
Bridging the research-management gap: using knowledge exchange and stakeholder engagement to aid
decision-making in invasive rat management
G. Maggs, M.A.C. Nicoll, N. Zuël, D.J. Murrell, J.G. Ewen, C. Ferrière, V. Tatayah, C.G. Jones and K. Norris .31
Timing aerial baiting for rodent eradications on cool temperate islands: mice on Marion Island
J.P. Parkes ...............................................................................................................................................................36
South Africa works towards eradicating introduced house mice from sub-Antarctic Marion Island: the
largest island yet attempted for mice
G.R. Preston, B.J. Dilley, J. Cooper, J. Beaumont, L.F. Chauke, S. L. Chown, N. Devanunthan,
M. Dopolo, L. Fikizolo, J. Heine, S. Henderson, C.A. Jacobs, F. Johnson, J. Kelly,
A.B. Makhado, C. Marais, J. Maroga, M. Mayekiso, G. McClelland, J. Mphepya, D. Muir,
N. Ngcaba, N. Ngcobo, J.P. Parkes, F. Paulsen, S. Schoombie, K. Springer, C. Stringer,
H. Valentine, R.M. Wanless and P.G. Ryan ........................................................................................................40
Improving the efficiency of aerial rodent eradications by means of the numerical estimation of rodenticide density
E. Rojas-Mayoral, F.A. Méndez-Sánchez, B. Rojas-Mayoral and A. Aguirre-Muñoz ............................................47
Preventing extinctions: planning and undertaking invasive rodent eradication from Pinzon Island, Galapagos
D. Rueda, V. Carrion, P.A. Castaño, F. Cunninghame, P. Fisher, E. Hagen, J.B. Ponder,
C.A. Riekena, C. Sevilla, H. Shield, D. Will and K.J. Campbell .......................................................................51
First results from a pilot programme for the eradication of beavers for environmental restoration in
Tierra Del Fuego
A. Schiavini, J. Escobar, E. Curto and P. Jusim ......................................................................................................57
Towards a genetic approach to invasive rodent eradications: assessing reproductive competitiveness
between wild and laboratory mice
M. Serr, N. Heard and J. Godwin ............................................................................................................................64
Considerations and consequences when conducting aerial broadcast applications during rodent eradications
D. Will, G. Howald, N. Holmes, R. Griffiths and C. Gill ........................................................................................71
It’s not all up in the air: the development and use of ground-based rat eradication techniques in the UK
E.A. Bell...................................................................................................................................................................79
The Isles of Scilly seabird restoration project: the eradication of brown rats (Rattus norvegicus) from
the inhabited islands of St Agnes and Gugh, Isles of Scilly
E. Bell, K. Floyd, D. Boyle, J. Pearson, P. St Pierre, L. Lock, P. Buckley, S. Mason, R. McCarthy,
W. Garratt, K. Sugar and J. Pearce ...................................................................................................................88
Rat eradication in the Pitcairn Islands, South Pacific: a 25-year perspective
M.de L. Brooke ........................................................................................................................................................95
House mice on islands: management and lessons from New Zealand
K. Broome, D. Brown, K. Brown, E. Murphy, C. Birmingham, C. Golding, P. Corson, A. Cox and R. Griffiths .100
v
Simultaneous rat, mouse and rabbit eradication on Bense and Little Bense Islands, Falkland Islands
P.W. Carey .............................................................................................................................................................108
The history of the aerial application of rodenticide in New Zealand
P. Garden, P. McClelland and K. Broome .............................................................................................................114
Successes and failures of rat eradications on tropical islands: a comparative review of eight recent projects
R. Griffiths, D. Brown, B. Tershy, W.C. Pitt, R.J. Cuthbert, A. Wegmann, B. Keitt, S. Cranwell and G. Howald 120
Eradication of mice from Antipodes Island, New Zealand
S. Horn, T. Greene and G. Elliott ..........................................................................................................................131
Scaling down (cliffs) to meet the challenge: the Shiants’ black rat eradication
C.E. Main, E. Bell, K. Floyd, J. Tayton, J. Ibbotson, W. Whittington, P.R. Taylor, R. Reid,
K. Varnham, T. Churchyard, L. Bambini, A. Douse, T. Nicolson and G. Campbell.........................................138
The effect of Norway rats on coastal waterbirds of the Falkland Islands: a preliminary analysis
S. Poncet, K. Passfield, A. Kuepfer and M.A. Tabak .............................................................................................147
Applying lessons learnt from tropical rodent eradications: a second attempt to remove invasive rats
from Desecheo National Wildlife Refuge, Puerto Rico
D.J. Will, K. Swinnerton, S. Silander, B. Keitt, R. Griffiths, G.R. Howald, C.E. FiguerolaHernandez and J.L. Herrera-Giraldo ..............................................................................................................154
The eradication of black rats (Rattus rattus) from Dog Island, Anguilla, using ground-based techniques
E. Bell, J. Daltry, F. Mukhida, R. Connor and K. Varnham ..................................................................................162
Recovery of introduced Pacific rats following a failed eradication attempt on subtropical Henderson
Island, South Pacific Ocean
A.L. Bond, R.J. Cuthbert, G.T.W. McClelland, T. Churchyard, N. Duffield, S. Havery, J. Kelly,
J.L. Lavers, T. Proud, N. Torr, J.A. Vickery and S. Oppel ...............................................................................167
Bait colour and moisture do not affect bait acceptance by introduced Pacific rats (Rattus exulans) at
Henderson Island, Pitcairn Islands
A.L. Bond, S. O’Keefe, P. Warren and G.T.W. McClelland ...................................................................................175
Containment of invasive grey squirrels in Scotland: meeting the challenge
J. Bryce and M. Tonkin ..........................................................................................................................................180
Testing auto-dispensing lure pumps for incursion control of rats with reduced effort on a small,
re-invadable island in New Zealand
A. Carter, R. van Dam, S. Barr and D. Peters.......................................................................................................187
Survival analysis of two endemic lizard species before, during and after a rat eradication attempt on
Desecheo Island, Puerto Rico
J.L. Herrera-Giraldo, C.E. Figuerola-Hernández, N.D. Holmes, K. Swinnerton, E.N. BermúdezCarambot, J.F. González-Maya and D.A. Gómez-Hoyos ................................................................................191
Long term rodent control in Rdum tal-Madonna yelkouan shearwater colony
P. Lago, J.S. Santiago Cabello and K. Varnham ...................................................................................................196
Seasonal variation in movements and survival of invasive Pacific rats on sub-tropical Henderson
Island: implications for eradication
S. Oppel, G.T.W. McClelland, J.L. Lavers, T. Churchyard, A. Donaldson, N. Duffield, S. Havery,
J. Kelly, T. Proud, J.C. Russell and A.L. Bond ................................................................................................200
Assessing the critical role that land crabs play in tropical island rodent eradications and ecological restoration
A. Samaniego-Herrera, S. Boudjelas, G.A. Harper, and J.C. Russell ...................................................................209
Trail cameras are a key monitoring tool for determining target and non-target bait-take during rodent
removal operations: evidence from Desecheo Island rat eradication
A.B. Shiels, D. Will, C. Figuerola-Hernández, K.J. Swinnerton, S. Silander, C. Samra and G.W. Witmer ..........223
Rat and lagomorph eradication on two large islands of central Mediterranean: differences in island
morphology and consequences on methods, problems and targets
P. Sposimo, D. Capizzi, T. Cencetti, F. De Pietro, F. Giannini, C. Gotti, F. Puppo, G. Quilghini,
E. Raganella Pelliccioni, G. Sammuri, V. Trocchi, S. Vagniluca, F. Zanichelli and N. Baccetti .....................231
Chapter 2: Other taxa
Big island feral cat eradication campaigns: an overview and status update of two significant examples
D. Algar, M. Johnston and C. Pink........................................................................................................................238
Safeguarding Orkney’s native wildlife from non-native invasive stoats
M. Auld, B. Ayling, L. Bambini, G. Harper, G. Neville, S. Sankey, D.B.A. Thompson and P. Walton...................244
Rhesus macaque eradication to restore the ecological integrity of Desecheo National Wildlife Refuge,
Puerto Rico
C.C. Hanson, T.J. Hall, A.J. DeNicola, S. Silander, B.S. Keitt and K.J. Campbell ..............................................249
Eradication of red deer from Secretary Island, New Zealand: changing tactics to achieve success
N. Macdonald, G. Nugent, K-A. Edge and J.P. Parkes .........................................................................................256
Large scale eradication of non-native invasive American mink (Neovison vison) from the Outer
Hebrides of Scotland
I.A. Macleod, D. Maclennan, R. Raynor, D.B.A. Thompson and S. Whitaker ......................................................261
vi
Ecological restoration of Socorro Island, Revillagigedo Archipelago, Mexico: the eradication of feral
sheep and cats
A. Ortiz-Alcaraz, A. Aguirre-Muñoz, F. Méndez-Sánchez, E. Rojas-Mayoral, F. Solís-Carlos,
B. Rojas-Mayoral, E. Benavides-Ríos, S. Hall, H. Nevins and A. Ortega-Rubio ............................................267
Removing introduced hedgehogs from the Uists
Thompson, R.C. and J.M. Ferguson ......................................................................................................................274
Five eradications, three species, three islands: overview, insights and recommendations from invasive
bird eradications in the Seychelles
N. Bunbury, P. Haverson, N. Page, J. Agricole, G. Angell, P. Banville, A. Constance, J. Friedlander,
L. Leite, T. Mahoune, E. Melton-Durup, J. Moumou, K. Raines, J. van de Crommenacker
and F. Fleischer-Dogley ..................................................................................................................................282
House sparrow eradication attempt on Robinson Crusoe Island, Juan Fernández Archipelago, Chile
E. Hagen, J. Bonham and K. Campbell ................................................................................................................289
Predation pressures on sooty terns by cats, rats and common mynas on Ascension Island in the South Atlantic
B.J. Hughes, R.C. Dickey and S.J. Reynolds .........................................................................................................295
Eradication and control programmes for invasive mynas (Acridotheres spp.) and bulbuls (Pycnonotus
spp.): defining best practice in managing invasive bird populations on oceanic islands
S. Saavedra Cruz and S.J. Reynolds......................................................................................................................302
Invasion by the red-vented bulbul: an overview of recent studies in New Caledonia
M. Thibault, E. Vidal, M.A. Potter, F. Masse, A. Pujapujane, B. Fogliani, G. Lannuzel, H. Jourdan,
N. Robert, L. Demaret, N. Barré and F. Brescia..............................................................................................309
Population assessment of a novel island invasive: tegu (Salvator merianae) of Fernando de Noronha
C.R. Abrahão, J.C. Russell, J.C.R. Silva, F. Ferreira and R.A. Dias ....................................................................317
Life-history comparisons between the native range and an invasive island population of a colubrid snake
S.R. Fisher, R.N. Fisher, S. Alcaraz, R. Gallo-Barneto, C. Patino-Martinez, L.F. López Jurado and
C.J. Rochester ..................................................................................................................................................326
Control of the ladder snake (Rhinechis scalaris) on Formentera using experimental live-traps
G. Picó, M.J. Fernández, J.E. Moreno and V. Colomar........................................................................................332
Spatial dynamics of invasion and distribution of alien frogs in a biodiversity hotspot archipelago
A.N. Pili, C.E. Supsup, E.Y. Sy, M.L.L. Diesmos and A.C. Diesmos......................................................................337
In situ evaluation of an automated aerial bait delivery system for landscape-scale control of invasive
brown treesnakes on Guam
S.R. Siers, W.C. Pitt, J.D. Eisemann, L. Clark, A.B. Shiels, C.S. Clark, R.J. Gosnell and M.C. Messaros ..........348
The potential detrimental impact of the New Zealand flatworm to Scottish islands
B. Boag and R. Neilson ..........................................................................................................................................356
Management of an invasive avian parasitic fly in the Galapagos Islands: is biological control a viable option?
R.A. Boulton, M. Bulgarella, I.E. Ramirez, C.E. Causton and G.E. Heimpel .......................................................360
Feasibility of eradicating the large white butterfly (Pieris brassicae) from New Zealand: data gathering
to inform decisions about the feasibility of eradication
K. Brown, C.B. Phillips, K. Broome, C. Green, R. Toft and G. Walker .................................................................364
Effort required to confirm eradication of an Argentine ant invasion: Tiritiri Matangi Island, New Zealand
C. Green ................................................................................................................................................................370
Analysis of the secondary nest of the yellow-legged hornet found in the Balearic Islands reveals its
high adaptability to Mediterranean isolated ecosystems
C. Herrera, A. Marqués, V. Colomar and M.M. Leza ...........................................................................................375
Methods for monitoring invertebrate response to vertebrate eradication
M. Houghton, A. Terauds and J. Shaw ..................................................................................................................381
Introduction of biological control agents against the European earwig (Forficula auricularia) on the
Falkland Islands
N. Maczey, D. Moore, P. González-Moreno and N. Rendell .................................................................................389
Assessment of snail exposure to the anticoagulant rodenticide brodifacoum in the Galapagos Islands
C.E. Parent, P. Fisher, W. Jolley, A. Alifano and K.J. Campbell ...........................................................................394
Criteria to help evaluate and guide attempts to eradicate terrestrial arthropod pests
C.B. Phillips, K. Brown, K. Broome, C. Green and G. Walker .............................................................................400
Eradication of invasive alien crayfish: past experiences and further possibilities
R. Sandodden.........................................................................................................................................................405
Modelling invasive plant alien species richness in Tenerife (Canary Islands) using Bayesian
Generalised Linear Spatial Models
D. Da Re, E. Tordoni, Z. Negrín-Pérez , J. M. Fernàndez-Palacios, J. R. Arévalo, R. Otto,
D. Rocchini and G. Bacaro ..............................................................................................................................410
Using expert knowledge and field surveys to guide management of an invasive alien palm in a Pacific
Island lowland rainforest
M.J.B. Dyer, G. Keppel, D. Watling, M. Tuiwawa, S. Vido and H.J. Boehmer .....................................................417
Persistence, accuracy and timeliness: finding, mapping and managing non-native plant species on the
island of South Georgia (South Atlantic)
K. Floyd, K. Passfield, S. Poncet, B. Myer and J. Lee ..........................................................................................424
vii
Eradication programmes complicated by long-lived seed banks: lessons learnt from 15 years of
miconia control on O'ahu Island, Hawai'i
R. Neville, J.Y. Fujikawa and M. Halabisky ..........................................................................................................430
Weed eradication on Raoul Island, Kermadec Islands, New Zealand: progress and prognosis
C.J. West and D. Havell ........................................................................................................................................435
Successful eradication of signal crayfish (Pacifastacus leniusculus) using a non-specific biocide in a
small isolated water body in Scotland
L. Ballantyne, D. Baum, C.W. Bean, J. Long and S. Whitaker..............................................................................443
Small- and large-scale eradication of invasive fish and fish parasites in freshwater systems in Norway
H. Bardal ...............................................................................................................................................................447
First report of marine alien species in mainland Ecuador: threats of invasion in rocky shores
M. Cárdenas-Calle, J. Pérez-Correa, P. Martinez,, I. Keith, F. Rivera, M. Cornejo, G. Torres, F.
Villamar, R. Zambrano, A. Cárdenas, M. Triviño, L. Troccoli, G. Bigatti, J. Coronel and E. Mora ...............452
Lessons on effectiveness and long-term prevention from broad-scale control of invasive alien species
in Scotland’s rivers and lochs
J.C. Horrill, M.K. Oliver and J. Stubbs Partridge ................................................................................................458
Chapter 3: Strategy
Biosecurity on St Helena Island – a socially inclusive model for protecting small island nations from
invasive species
J.R. Balchin, D.G. Duncan, G.E. Key and N. Stevens...........................................................................................468
Proactive planning and compliance for a high-priority invasive species rapid response programme
C.L. Boser, P. Power, A. Little, J. Matos, G.R. Howald, J.M. Randall and S.A. Morrison ...................................473
How do we prevent the obstacles to good island biosecurity from limiting our eradication ambitions?
E.S. Kennedy and K.G. Broome ............................................................................................................................478
Mexico’s island biosecurity programme: collaborative formulation and implementation
M. Latofski-Robles, F. Méndez-Sánchez, A. Aguirre-Muñoz, C. Jáuregui-García, P. Koleff-Osorio,
A.I. González-Martínez, G. Born-Schmidt, J. Bernal-Stoopen and E. Rendón-Hernández ............................484
Achieving post-eradication biosecurity on South Georgia
M.G. Richardson and J.P. Croxall.........................................................................................................................489
Beyond biodiversity: the cultural context of invasive species initiatives in Gwaii Haanas
K.X.T. Bellis, R.T. Peet, R.L. Irvine, G. Howald and G.J. Alsop ...........................................................................494
Cooperative natural resource and invasive species management in Hawai'i
D.C. Duffy and C. Martin ......................................................................................................................................497
Invasive species management in Mauritius: from the reactive to the proactive – the National Invasive
Species Management Strategy and its implementation
J.R. Mauremootoo, S. Pandoo, V. Bachraz, I. Buldawoo and N.C. Cole ..............................................................503
Islander perceptions of invasive alien species: the role of socio-economy and culture in small isolated
islands of French Polynesia (South Pacific)
J.-Y. Meyer and M. Fourdrigniez ..........................................................................................................................510
Married bliss and shotgun weddings: effective partnerships for island restoration
C. Stringer, S. Boudjelas, K. Broome, S. Cranwell, E. Hagen, G. Howald, J. Kelly, J. Millett, K.
Springer and K. Varnham ................................................................................................................................517
Winning the hearts and minds – proceeding to implementation of the Lord Howe Island rodent
eradication project: a case study
A. Walsh, A. Wilson, H. Bower, P. McClelland and J. Pearson .............................................................................522
Recovery and current status of seabirds on the Baja California Pacific Islands, Mexico, following
restoration actions
Y. Bedolla-Guzmán, F. Méndez-Sánchez, A. Aguirre-Muñoz, M. Félix-Lizárraga, A. Fabila-Blanco,
E. Bravo-Hernández, A. Hernández-Ríos, M. Corrales-Sauceda, A. Aguilar-Vargas, A. AztorgaOrnelas, F. Solís-Carlos, F. Torres-García, L. Luna-Mendoza, A. Ortiz-Alcaraz, J. HernándezMontoya, M. Latofski-Robles, E. Rojas-Mayoral and A. Cárdenas-Tapia ......................................................531
The recovery of seabird populations on Ramsey Island, Pembrokeshire, Wales, following the
1999/2000 rat eradication
E.A. Bell, M.D. Bell, G. Morgan and L. Morgan ..................................................................................................539
Practical considerations for monitoring invasive mammal eradication outcomes
J.P. Bird, K. Varnham, J.D. Shaw and N.D. Holmes .............................................................................................545
Community-based conservation and recovery of native species on Monuriki Island, Fiji
R.N. Fisher, J. Niukula, P. Harlow, S. Rasalato, R. Chand, B. Thaman, E. Seniloli, J. Vadada, S.
Cranwell, J. Brown, K. Lovich and N. Thomas-Moko .....................................................................................552
Costs and benefits for biodiversity following rat and cat eradication on Te Hauturu-o-Toi/Little Barrier Island
R. Griffiths, E. Bell, J. Campbell, P. Cassey, J.G. Ewen, C. Green, L. Joyce, M. Rayner, R. Toy,
D. Towns, L. Wade, R. Walle and C.R. Veitch ..................................................................................................558
Restoring plant-pollinator communities: using a network approach to monitor pollination function
C.N. Kaiser-Bunbury .............................................................................................................................................568
viii
Ten years after feral goat eradication: the active restoration of plant communities on Guadalupe Island, Mexico
L. Luna-Mendoza, A. Aguirre-Muñoz, J.C. Hernández-Montoya, M. Torres-Aguilar, J.S. GarcíaCarreón, O. Puebla-Hernández, S. Luvianos-Colín, A. Cárdenas-Tapia and F. Méndez-Sánchez .................571
Canna seabird recovery project: 10 years on
R. Luxmoore, R. Swann and E. Bell ......................................................................................................................576
Conservation gains and missed opportunities 15 years after rodent eradications in the Seychelles
J.E. Millett, W. Accouche, J. van de Crommenacker, M.A.J.A. van Dinther, A. de Groene,
C.P. Havemann, T.A. Retief, J. Appoo and R.M. Bristol ..................................................................................580
Eradication of invasive animals and other island restoration practices in Seychelles: achievements,
challenges and scaling up perspectives
G. Rocamora .........................................................................................................................................................588
No detection of brodifacoum residues in the marine and terrestrial food web three years after rat
eradication at Palmyra Atoll, Central Pacific
A. Wegmann, G. Howald, S. Kropidlowski, N. Holmes and A.B. Shiels................................................................600
‘Island’ eradication within large landscapes: the remove and protect model
P. Bell, H. Nathan and N. Mulgan .........................................................................................................................604
Multi island, multi invasive species eradication in French Polynesia demonstrates economies of scale
R. Griffiths, S. Cranwell, D. Derand, T. Ghestemme, D. Will, J. Zito, T. Hall, M. Pott and G. Coulston .............611
Trialling gene drives to control invasive species: what, where and how?
T. Harvey-Samuel, K.J. Campbell, M. Edgington and L. Alphey ..........................................................................618
Tracking invasive species eradications on islands at a global scale
N.D. Holmes, B.S. Keitt, D.R. Spatz, D.J. Will, S. Hein, J.C. Russell, P. Genovesi, P.E. Cowan
and B.R. Tershy ................................................................................................................................................628
Going to scale: reviewing where we’ve been and where we need to go in invasive vertebrate eradications
B. Keitt, N. Holmes, E. Hagen, G. Howald and K. Poiani ....................................................................................633
Tackling invasive non-native species in the UK Overseas Territories
G.E. Key and N.P. Moore ......................................................................................................................................637
A little goes a long way when controlling invasive plants for biodiversity conservation
D.A. Knapp, J.J. Knapp, K.A. Stahlheber and T. Dudley ......................................................................................643
Achieving large scale, long-term invasive American mink control in northern Scotland despite short term funding
X. Lambin, J.C. Horrill and R. Raynor .................................................................................................................651
Battling invasive species in the Pacific
D. Moverley ...........................................................................................................................................................658
Maximising conservation impact by prioritising islands for biosecurity
S. Oppel, S.J. Havery, L. John, L. Bambini, K. Varnham, J. Dawson and E. Radford ..........................................663
Working with the local community to eradicate rats on an inhabited island: securing the seabird
heritage of the Isles of Scilly
J. Pearson, P. St Pierre, L. Lock, P. Buckley, E. Bell, S. Mason, R. McCarthy, W. Garratt, K. Sugar
and J. Pearce ...................................................................................................................................................670
A plan for the eradication of invasive alien species from Arctic islands
J.K. Reaser, G.R. Howald and S.D. Veatch ...........................................................................................................679
Invasive species removals and scale – contrasting island and mainland experience
P.A. Robertson, S. Roy, A.C. Mill, M. Shirley, T. Adriaens, A.I. Ward, V. Tatayah and O. Booy ...........................687
Strategic environmental assessment for invasive species management on inhabited islands
J.C. Russell and C.N. Taylor .................................................................................................................................692
Towards a guidance document for invasive species planning and management on islands
A. Tye ....................................................................................................................................................................698
Chapter 4: Abstracts
Mexico’s progress and commitment to comprehensive island restoration
A. Aguirre-Muñoz, F. Méndez-Sánchez, L. Luna-Mendoza, A. Ortiz-Alcaraz,
J. Hernández-Montoya, Y. Bedolla-Guzmán, M. Latofski-Robles, E. Rojas-Mayoral,
N. Silva-Estudillo, F. Torres-García, M. Félix-Lizárraga, A. Fabila-Blanco, A. Hernández-Ríos,
E. Bravo-Hernández, F. Solís-Carlos, C. Jáuregui-García and D. Munguía-Cajigas ....................................704
The Pacific invasives partnership – a model for regional collaboration on invasive alien species
P.C. Andreozzi, R. Griffiths, D. Moverley, J. Wainiqolo, R. Nias, S. Boudjelas, D. Stewart,
S. Cranwell, M. Smith and P. Cowan...............................................................................................................704
A review of monitoring of biodiversity responses to island invasive species eradications
J. Bird, J. Shaw, R. Alderman and R. Fuller .........................................................................................................705
A review of seabird recovery on Lundy Island, England, over a decade following the eradication of
brown and black rats
H. Booker, D. Appleton, D. Bullock, R. MacDonald, E. Bell, D. Price, P. Slader, T. Frayling,
A. Taylor and S. Havery ...................................................................................................................................705
Eradicating invasive ants in conservation areas
C.L. Boser..............................................................................................................................................................705
ix
Big island, small invader: eradicating invasive fish on a national scale
M. Brazier..............................................................................................................................................................706
Population growth of seabirds after the eradication of introduced mammals
R. Buxton and M. Brooke ......................................................................................................................................706
Assessment of the possible effects of biological control agents of Lantana camara and Chromolaena
odorata in Davao City, Mindanao, Philippines
C. Canlas, C. Gever, P. Rosialda, Ma. N. Quibod, P. Buenavente, N. Barbecho, C. Layusa and M. Day ............706
Black rat eradication from Linosa Island: work in progress
M. Cecchetti, G. Dell’Omo and B. Massa .............................................................................................................707
Effects of cat, rat, and human predation on Scopoli’s shearwater (Calonectris diomedea) breeding
success and nest-site occupancy on Linosa Island
M. Cecchetti, L. Nelli, B. Massa and G. Dell’Omo...............................................................................................707
Invasive plants: what can be done about this continuing threat to biodiversity?
C. Clubbe...............................................................................................................................................................707
Partnerships in the restoration of tropical Pacific islands
S. Cranwell ............................................................................................................................................................708
Vespapp: citizen science to detect the invasive species Vespa velutina
M. del Mar Leza, A. Marqués, C. Herrera, M. Ángel Miranda, M. Ruiz, A. Pou and C. Guerrero ......................708
Wild ginger, a beautiful menace to island ecosystems – can a natural solution be found?
D. Djeddour, N. Maczey and C. Pratt ...................................................................................................................708
Is poisoning rodents a health hazard?
J. Doube ................................................................................................................................................................709
When our enemy is our friend: new approaches to managing alien vegetation in Seychelles catchment forest
K. Fleischmann, S. Massy, M. Schmutz, B. Seraphine and J. Millett ....................................................................709
Eleonora’s falcon (Falco eleonorae) benefiting from cat eradication – the case of Andros, Greece
J. Fric, T. Dimalexis, V. Goritsas, A. Evangelidis and I. Nikolaou .......................................................................709
A review of 12 years of rat eradication operations for the conservation of priority island nesting birds in Greece
J. Fric and A. Evangelidis .....................................................................................................................................710
Improving nesting habitats for the Eleonora’s falcon and seabirds
J. Fric, A. Evangelidis, T. Dimalexis, N. Tsiopelas, S. Xirouchakis, C. Kassara and S. Giokas...........................710
Broadening the context of invasive species eradications
P. Genovesi ............................................................................................................................................................710
Recovery of Santa Luzia Nature Reserve and translocation of the globally endangered Raso lark
P. Geraldes, T. Melo, P. Oliveira and V. Paiva......................................................................................................711
Setting-up a predator-free area on a Macaronesian island using a pest-proof fence
P. Geraldes, T. Pipa, N. Oliveira, C. Silva and S. Hervías ...................................................................................711
Green iguana (Iguana iguana) monitoring and control efforts on Grand Cayman
J. Haakonsson, F. Rivera-Milan and E. Radford ..................................................................................................711
Predicting the potential habitat of the invasive coral vine (Antigonon leptopus) using remote sensing
and species distribution modelling
E. Haber, M. Eppinga, M. Ferreira dos Santos, M. Rietkerk and M. Wassen .......................................................712
The diet of ‘Viking mice’ on Nólsoy, Faroe Islands
S. Hammer and J. Russell......................................................................................................................................712
Seabird restoration and advances towards the eradication of feral cats on Guadalupe Island, Mexico
J.C. Hernández-Montoya, L. Luna-Mendoza, A. Aguirre-Muñoz, F. Méndez-Sánchez,
A. Duarte-Canizales, E. Rojas-Mayoral, S. Hall, Z. Peña-Moreno, S. Figueroa-Flores,
D. Cosio-Muriel and M. Latofski-Robles ........................................................................................................712
From island studies to mainland management
S. Hudin .................................................................................................................................................................713
Differential effects of human impact and habitat type on exotic and native species diversity on oceanic islands
W. Jesse, J. Ellers, J. Behm and M. Helmus ..........................................................................................................713
Genetic pest management technologies to control invasive rodents
D. Kanavy and D. Threadgill ................................................................................................................................713
A new look at Galapagos fouling communities
I. Keith, J. Carlton and G. Ruiz.............................................................................................................................714
Planning processes for eradication of mice on Gough Island
J. Kelly, K. Springer, C. Stringer, A. Schofield and T. Glass .................................................................................714
Perils of saving the smallest for the last: lessons learnt about sequencing eradications on Santa Cruz Island, CA
J. Knapp, C. Boser, J. Randall, E. O’Byrne and S.A. Morrison............................................................................714
Citizens’ attitude towards the removal of grey squirrels in Italy: what support do we need?
V. La Morgia, D. Paoloni, P. Aragno and P. Genovesi..........................................................................................715
Computer modelling of complex interstitial spaces to protect endemic island lizards from invasive mice
Z. Lennon, H. Wittmer and N. Nelson ...................................................................................................................715
x
An integrated physical control method on Spartina alterniflora
J. Li, C. Zhao and X. Zhao ....................................................................................................................................715
Predicting the risk of plant invasion on islands: the case of Miconia calvescens in the Marquesas,
French Polynesia (South Pacific)
M. Libeau, R. Pouteau, R. Taputuarai and J.-Y. Meyer ........................................................................................716
The secret life in Switzerland of an island pest, the house mouse
A. Lindholm and B. König .....................................................................................................................................716
Catalysing conservation of islands through collaboration: a North American perspective
A. Little, A. Aguirre-Muñoz, G. Seutin, L. Wein, P. Nantel, H. Berlanga, F. Méndez-Sánchez,
J. Putnam, E. Iñigo-Elías and G. Howald .......................................................................................................716
The value of monitoring and the price of uncertainty in the management of an invasive population
E. McHenry, X. Lambin, T. Cornulier and D. Elston ............................................................................................717
Invasive arthropods of ecological, agricultural and health importance recently introduced in the
Balearic Islands (Spain)
M. A. Miranda, C. Barceló, D. Borràs, A. González, M. Leza and C. Paredes-Esquivel .....................................717
Using key-informant surveys to reliably and rapidly estimate the distributions of multiple insular invasive species
N.P. Mohanty, G.J. Measey, A. Sachin, G. Selvaraj and K. Vasudevan ................................................................717
Time germination response to temperature and light conditions in Ulex
Z. Negrín Pérez, D. Da Re, M. Bernardos and B. Garrido ...................................................................................718
Rat eradication from Berlengas Island, Portugal
N. Oliveira, P. Geraldes, I. Fagundes, P. Oliveira and J. Andrade ......................................................................718
Response of an open feral cat population to an intensive control programme for improving the
critically endangered Fatu Hiva monarch conservation strategy
P. Palmas, R. Gouyet, T. Ghestemme, A. Matohi, E. Terorohauepa, I. Tauapaohu, C. Blanvillain,
J. Zito, D. Beaune and E. Vidal .......................................................................................................................718
Feral cats threaten the outstanding endemic fauna of the New Caledonia biodiversity hotspot:
implications for feral cat management strategy
P. Palmas, H. Jourdan, E. Bonnaud, F. Rigault, L. Debar, H. De Méringo, E. Bourguet, R.
Adjouhgniope and E. Vidal. .............................................................................................................................719
Scaling up invasive plant management for ecosystem restoration in Mauritius: successes and challenges
S. Pandoo, P. Ragen, B. Vishnuduth, Z. Jhumka and J. Mauremootoo .................................................................719
Implementing an early detection programme on Catalina Island: prioritising landscaped grasses
J. Parish ................................................................................................................................................................719
Challenges and opportunities for lethal and non-lethal management of non-native ungulates on islands:
feral pigs, goats and cows
D. Parrott, G. Massei, R. Ridley, J. Sandon, M. Lambert, D. Cowan and M. Sutton-Croft..................................720
Diet of introduced black rats (Rattus rattus) on Christmas Island: setting the scene with stomach and
stable isotope analysis
C. Pink, D. Algar and P. Green .............................................................................................................................720
The prospects for biological control of Rubus niveus in the Galapagos Islands
K. Pollard, D. Kurose, A. Buddie and C. Ellison ..................................................................................................720
A tool for biodiversity conservation within Chile: renewed interest in island eradications sparked by
successful European rabbit (Oryctolagus cuniculus) eradication
M. Pott, E. Hagen, P. Martínez and M. Díaz.........................................................................................................721
Finders keepers? Discovering and securing the rare species rediscovered in weeded restoration plots
P. Ragen, S. Pandoo, V. Bachraz, Z. Jhumka and J. Mauremooto ........................................................................721
Impacts and control of invasive species: trading off actions
M. Roberts, W. Cresswell and N. Hanley ..............................................................................................................721
Incorporating interaction networks into conservation: Tasmania as a case study
A. Rogers, J. Shaw and S. Kark.............................................................................................................................722
Invasive plants of the Caribbean: application of herbarium collections to protect a regional biodiversity hotspot
J. Rojas-Sandoval, P. Acevedo-Rodríguez, M. Datiles, S. Dube, H. Diaz-Soltero, L. Charles,
G. Richards, M. Angel Duenas, D. Simpson, E. Ventosa-Febles, J. Ackerman,
F. Areces-Berazain, M. Caraballo-Ortiz, A. Carvajal-Vélez, J. Chabert-Llompart, S. Kaufman,
J. Thompson and J. Vélez-Gavilán ..................................................................................................................722
Can large database mining inform invasive non-native species management on islands?
L. Ruffino and T. Cornulier ...................................................................................................................................722
Management of numerous introduced plants on Matiu (Somes Island), Wellington, New Zealand
P. Russell and S. Weaver........................................................................................................................................723
Managing Vespula wasp invasion in New Zealand
J.M. Schmack, M.C. Barron, D.F. Ward and J.R. Beggs .......................................................................................723
Invasive rat colonisation history and movement dynamics in Haida Gwaii
B. Sjodin, R. Irvine, G. Howald and M. Russello ..................................................................................................723
Garden cans and river rafts – equipped to approach invasive freshwater fish
B. Skei ....................................................................................................................................................................724
xi
What happens after the helicopters have gone – assessing post-eradication changes on Macquarie Island
K. Springer ............................................................................................................................................................724
Predicting the distribution of island invader bird species under climate change
M. Thibault, F. Brescia and M. Barbet-Massin .....................................................................................................724
Biosecurity Plan for invasive ants in the Pacific
C. Vanderwoude, S. Boudjelas, P. Andreozzi, P. Cowan and J. Wainiqolo ...........................................................725
Rat control to protect the Turks and Caicos rock iguana: monitoring and responding to rat activity on a
Caribbean island Nature Reserve
K. Varnham, E. Radford, S. Busuttil, C. Forbes, E. Gibbs-Williams and G. Gerber ............................................725
An innovative programme to protect the UK’s seabird islands
K. Varnham, S. Thomas, L. Bambini, S. Havery and L. Lock................................................................................725
Prioritising islands for the eradication of invasive vertebrates in the Arctic
S.D. Veatch ............................................................................................................................................................726
Changes in forest passerine numbers on Hauturu following rat eradication
C.R. Veitch .............................................................................................................................................................726
Habitat features that influence predation of endangered Hawaiian common gallinule nests by invasive
vertebrates in Hanalei and Huleia National Wildlife Refuges
B. Webber, K. Uyehara, T. Luxner and D. Dewey .................................................................................................726
Using sUAS to direct trap placement in support of feral cat eradication on islands
D. Will, T. Hall, M. Khalsa and J. Bruch ..............................................................................................................727
Removal of invasive, black rats increases activity levels and population density of Christmas Island’s
last remaining endemic reptile
M. Wynn and D. Driscoll .......................................................................................................................................727
Effect of Spartina alterniflora invasion on benthic macro-invertebrate communities in Guangxi
Zhuang Autonomous Region
C. Zhao, J. Li and X. Liu .......................................................................................................................................727
xii
A.R. Martin, M.N. Clout, J.C. Russell, C.R. Veitch and C.J. West.
Martin, A.R.; M.N. Clout, J.C. Russell, C.R. Veitch and C.J. West. Addressing the challenge
PREFACE
Addressing the challenge
A.R. Martin1, M.N. Clout2, J.C. Russell3, C.R. Veitch4 and C.J. West5
Centre for Remote Environments, University of Dundee, Dundee, DD1 5BT, UK. 2Centre for
Biodiversity and Biosecurity, School of Biological Sciences, Private Bag 92019, Auckland, New
Zealand. 3School of Biological Sciences and Department of Statistics, University of Auckland,
Private Bag 92019, Auckland, New Zealand. 448 Manse Road, Papakura, New Zealand.
<dveitch@kiwilink.net.nz>. 5Department of Conservation, P.O. Box 10420, Wellington 6143,
New Zealand.
1
The papers in this volume were, with a few exceptions, presented at the third Island Invasives
conference, held in Dundee, Scotland in July 2017. The conference was attended by 254 people
from 41 countries or territories, reflecting growing global recognition of the problems caused by
invasive alien species (IAS) on islands and recent progress in solving those problems.
The prefaces of the Proceedings of the two earlier conferences in this series (Veitch & Clout,
2002, Veitch, et al., 2011) discussed many of the threats posed by IAS on islands, conditions for
eradication feasibility, possible complications and successes to date. They remain as relevant today
as when they were first published. So, what has changed in this field of conservation in the seven
years since the last conference proceedings? In a nutshell, scale, diversity and experience.
The first two conferences in the series were influential in bringing together people from many
parts of the globe to discuss and exchange ideas, to learn from the experience of others, to inform
and inspire. The principal motive for the South Georgia Heritage Trust and University of Dundee
to host the third conference in Dundee was a desire to give something back to the island invasives
community in recognition of the enormous support and assistance provided by so many people to
the South Georgia Habitat Restoration Project. The 2010 Auckland conference, and the contacts
made there, were undoubtedly pivotal in guiding the South Georgia operation to success (Veitch,
et al., 2011).
The sub-title of the Dundee conference was 'Scaling up to meet the challenge'. The papers in
this volume and in the recent literature demonstrate up-scaling in several aspects of eradication
operations – not least in ambition, land area, operational size, global reach and of course financial
cost. In the space of a few decades, the size of islands treated for invasive species has increased by
five orders of magnitude – from a few hectares to over 100,000 ha or 1,000 km2. Meanwhile, the
diversity of species being tackled has increased, as has the range of countries now actively carrying
out island restoration work. Inspired by pioneers from New Zealand and Australia, principally,
today the movement has spread to islands in all oceans and off all continents. This expansion has
been informed by, and has in turn produced, growing experience in all aspects of this field, from
non-target impacts to ecological responses to factors affecting eradication success. We now know
much more about why some eradication attempts fail, and consequently how to prevent subsequent
failures. We know how much operations will cost, and what level of budget contingency to allow
– hugely important considerations for potential organisers, fund-raisers and sponsors. Crucially,
and due in large measure to the internationally recognised work of the Island Eradication Advisory
Group (IEAG – staff of the New Zealand Government's Department of Conservation), operation
planners now have access to Best Practice guidelines, and these have underpinned much of the
work reported upon in this volume. This field of conservation is remarkable in the degree of mutual
support and encouragement between individuals, organisations, countries, and between Government
and non-Government institutions.
The Dundee conference was opened by Her Royal Highness The Princess Royal, Patron of
the South Georgia Heritage Trust. Lord Gardiner, Parliamentary Under Secretary of State for
Rural Affairs and Biosecurity, spoke to the conference about the British Government's support of
the South Georgia Habitat Restoration Project and its commitment to confront problems caused
by Invasive Alien Species in the United Kingdom and its overseas territories more widely. Both
addresses are published in these Proceedings, with kind consent.
Indicative of the level of ambition now influencing the field, several papers in this volume
address topics related to the unveiling of Predator Free 2050, a campaign to rid New Zealand of
its most damaging invasive mammalian predators by the year 2050. If this bold objective is to be
achieved, novel tools will be needed to complement the existing arsenal of traps, bait, shooting
etc. Among the concepts being considered is that of gene drives – a means of reducing an invasive
population to zero by genetic engineering. The potential power of such developments generates
both excitement and concern, was the subject of much discussion at the conference and illustrates
how the field is adapting to the ever-increasing challenges posed by invasive species worldwide.
Learning from mistakes is a vital driver of progress, yet authors and journal editors alike are
often reluctant to publish papers that discuss failure. The editors of this volume have consequently
xiii
encouraged practitioners to write about what went wrong or could have been improved in their
own projects. Such openness is a sign of confidence and of a desire to advance the field, with
each operation being informed by the experiences of those that have gone before.
The editors of this volume have summarised what they consider to be key conclusions and
lessons to be drawn from the many, diverse papers published within it. They are:
●
The size of islands successfully cleared of invasive alien species that have been the
target of eradications has increased by an order of magnitude since the previous Island
Invasives conference.
●
Successful and large-scale eradications of invasive mammals other than rodents from
islands continue to occur, although some (e.g. mustelids) present significant challenges.
●
There are still relatively few examples of successful eradications of invasive birds,
but some have been achieved, despite management challenges and the threat of new
incursions.
●
The herpetofauna papers really highlight the need for effective border biosecurity to
exclude pests as well as information to guide the importation of exotic organisms.
●
For invertebrate eradications, principles are the same as for mammal eradications however
revisions to criteria to guide terrestrial arthropod eradications are proposed. Adaptive
management during eradication attempts is a consistent theme and methodologies to
evaluate the response of invertebrate communities to mammal eradications and nontarget impacts of vertebrate toxins on endemic molluscs are proposed.
●
Plant eradications require persistence over the long term because many species have a
seed bank (or similar cryptic life-stage) of high and often unknown longevity: regular
surveillance is essential to detect and remove plants as they germinate from the seed
bank and before they reproduce. In many situations, eradication is the optimal solution
rather than ongoing control.
●
Successful eradications of invasive aquatic species continue to be reported and can be
achieved using tools and knowledge currently available.
●
The presence of human populations can raise the cost and complexity of invasive
eradication operations, but investment in community engagement and participation
may remove barriers and should be factored in to all future operations on inhabited
islands.
●
Reviews of single or across multiple operations show the breadth of scope of invasive
rat eradications and are important for knowledge sharing and understanding failure.
●
Lessons from invasive rat eradications, particularly from those facing complex or novel
challenges, are important to inform attempts on other islands.
●
Effective biosecurity is essential to prevent new invasions and re-invasions and requires
community involvement, proactive planning, monitoring and rapid response.
●
Cooperation between indigenous (local) and national governments may allow projects
to expand beyond biodiversity conservation to become culturally significant as well,
i.e. restoring or aligning with existing traditional knowledge or resource use practices
or refining and improving these and bringing them into the realm of the total, diverse
human population locally, regionally or nationally.
●
Successful eradications of invasive species often yield significant benefits to native
species, natural ecosystems and local communities.
●
The effect of climate change on invasive species impacts is poorly known, so further
research is needed as well as application of the precautionary principle.
●
Genetic techniques such as gene drives offer the potential to facilitate eradications on
very large scales, but must be treated with caution until more research is conducted on
impacts and feasibility.
●
As ambition grows, so does the need for new techniques to facilitate eradications on
geographical scales never previously considered.
Our field is as much practical as it is academic, and a major aim of publishing these
Proceedings is to inform people who are, or will in the future be, planning new projects to
free islands of invasive species. Regardless of its location or the target species involved, each
successive operation builds on the experience of those who have gone before, and the papers in
this volume represent an invaluable wealth of such experience.
References
C.R. Veitch and M.N. Clout (eds.) (2002). Turning the tide: the eradication of invasive species. Occasional Paper SSC
no. 28 IUCN SSC Invasive Species Specialist Group. IUCN, Gland, Switzerland and Cambridge, UK.
C.R. Veitch, M.N. Clout and D.R. Towns (eds.) (2011). Island invasives: eradication and management. Occasional
Paper SSC no. 42 Gland, Switzerland: IUCN and Auckland, New Zealand: CBB.
xiv
OPENING ADDRESS
Her Royal Highness, the Princess Royal
Princess Anne
VICE-CHANCELLOR, LADIES AND GENTLEMEN,
it is a real pleasure to be able to join you for this, the
third of the Island Invasive conferences. I can work out by
the very fact of their subject matter being islands that these
conferences are few and far between, but judging by the
programme I have seen for the next five days you are really
making up for the gap in between. There are a lot of good
things to report and to talk about and to share.
I am delighted that one of those discussions will be about
the South Georgia Heritage Trust and the rat eradication
scheme that has been underway. It has been a real privilege
to visit South Georgia, not once but twice, in order to see
some of that work and understand the environment in
which the Trust is functioning. I am hugely impressed by
just how successful that work appears to have been, though
we are not counting our chickens, or indeed any of the
other birds that appear to be doing better as a result of the
rat eradication so far, and we look forward to being able to
prove the project’s success in the future.
The progress that has been made in many of the
eradication schemes must give enormous encouragement
to others, to know now what is really possible and to
believe that they can do the same. It perhaps underlines just
how important it is to have your methodology, logistics and
various other aspects in line before you can begin to have
a successful project.
You have here an astonishing gathering of those who
have been most involved at both the research and the
practical delivery end of these eradication schemes, which
covers a pretty wide range of talents. Looking at your
conference book there is an enormous diversity of success
stories to tell. You have seven of the world’s foremost
specialists in their respective fields who have agreed to
give keynote presentations at this conference. They will
talk on a wide range of topics, from island biosecurity to
invasive plants and international policy.
Contemplating invasive plants, I have to say, makes
eradicating rats look positively straightforward, but I’m
sure we will come up with a process for tackling those too.
Scotland of course has its own conservation challenges, as
we will hear in the keynote speech, which will talk about
the Shiant Islands, somewhere that I have also had the
pleasure of visiting, not officially but as part of a sailing
trip. If you time it right and the puffins are still there it is an
extraordinary place, well worth a visit, but it faces similar
challenges as South Georgia.
In other islands around Scotland there are other issues
to face. An invasive species I can think of in Coll came
all the way from New Zealand and is composed of some
large and rather successful worms, which everybody hoped
would eat themselves out of house and home but have
failed to do so yet!
The challenges facing the organisations attempting to
remove invasive species from islands are diverse. Progress
can be made from the experience of earlier operations and
oddly the operations that failed can teach us even more.
The South Georgia Heritage Trust has hugely benefitted
from the knowledge of the Island Invasives community
in tackling South Georgia’s eradication work. In deciding
to help host this next Island Invasives conference in
partnership with the University of Dundee, we really do
believe that that will inspire the next generation of island
conservationists. Dundee is a very good place to be able
to do that, and to the Vice-Chancellor and everybody
here from the University of Dundee, thank you for your
hospitality.
On our last trip to the island we were in Possession Bay
to celebrate Captain Cook’s first arrival at South Georgia
two hundred years before. When we looked at Possession
Bay at that particular moment one could sympathise with
Cook as he wondered what on earth he was doing there, but
it would be very nice to be able to say that South Georgia
had returned to the condition in which Captain Cook found
it and claimed possession. We still believe that might be
possible, so we look forward to the next couple of years
and being able to revisit South Georgia to really prove that
the eradication has been a success before we declare the
island rodent free. We all understand that 99% success is
not quite enough when it comes to removing rats and mice.
I am sure that this conference is something that you
have all been looking forward to, but I hope that you enjoy
it on a number of different levels, not least for the chance
to make friends because however successful you are at
communicating online, it is really nice and possibly more
encouraging to meet the people who have been involved
and can give you that very personal information about
what really works, what didn’t work and the little things
that caused the big problems. And those valuable lessons
from each other’s experiences will be something that you
can all take away, as well as happy memories of your time
in Dundee.
1
2
KEYNOTE ADDRESS
Protecting the biodiversity of the UK Overseas Territories
Lord Gardiner of Kimble
LADIES AND GENTLEMEN,
It is a great privilege to be here at this Conference and I
am delighted to learn more about how we can better protect
the biodiversity of our island ecosystems from the threat of
invasive alien species.
As you will hear, this week, much has been achieved
by passionate and committed conservationists around the
world since you last met in Auckland seven years ago. We
have a great opportunity this week to celebrate successes
and learn from these experiences.
The eradication work completed by the South Georgia
Heritage Trust, which you will hear about later this evening
from Professor Tony Martin is undoubtedly among the
most remarkable of recent island conservation efforts.
As we approach 2020, it is also a good time to reflect
on our progress towards the ambitious targets adopted by
the global community on invasive alien species as part of
the Aichi Targets and the Sustainable Development Goals.
UK successes
The past decade has seen a step-change in how the UK
responds to invasive non-native species. We now have
a co-ordinating secretariat, a risk analysis mechanism, a
GB Strategy and are prioritising species and pathways for
action.
My ministerial colleagues and I meet each month to
consider emerging threats across the biosecurity spectrum
– including animal diseases, plant pests and invasive nonnative species. The UK Department for Environment, Food
and Rural Affairs and the Scottish and Welsh Governments
will soon be putting in place contingency plans to stop over
30 high-risk invasive species getting a foothold in the UK.
This approach was tested last autumn when one of our
top threats – Asian hornet – was spotted in the south west of
England. My Department had a team on the ground within
48 hours and had successfully eliminated this specific
threat within 10 days.
The UK has also completed three further national
rapid response eradications, targeting two fish and one
amphibian species. Another six eradication campaigns
are underway. The biggest of these, the eradication of the
Ruddy duck, a world class effort covering the whole of the
UK, is now almost complete after more than a decade of
concerted effort.
Sadly, some non-native species are here to stay, yet we
seek to mitigate their impact. To this end, we have invested
over £1m in research on biocontrol agents for several
invasive plants.
The Department has also invested in a public-private
partnership to research novel methods of grey squirrel
control. Scotland is fortunate to have the Red Squirrel still
relatively widespread here.
We also recognise that effective awareness-raising is
key – for example, the GB Non-Native Species Secretariat
is leading an awareness-raising campaign called Check
Clean Dry, aimed at encouraging anglers and boaters
to reduce the risk of moving invasive species between
waterways. We adapted this from an excellent New Zealand
campaign of the same name.
I believe that it is vital to learn from the experiences
and good practice of others. We within the UK are keen to
share the lessons that we have learnt and the expertise that
we have developed – and to put these at the disposal of the
Overseas Territories and in collaboration more widely.
Focus on Overseas Territories
It is clear that much remains to be done to tackle the
issue of invasive non-native species. This is why the
International Union for Conservation of Nature launched
the global Honolulu Challenge last year. In December,
the UK Government pledged £2.75 million for priority
activities to tackle invasive species in the UK Overseas
Territories.
The UK is proud to be custodian of the precious and
unique biodiversity of 14 Overseas Territories, which
account for over 90% of UK biodiversity and contain two
World Heritage sites.
Our pledge shows our commitment to working in
partnership with the Territories to address what is probably
the single greatest threat to these unique places.
Many of the Overseas Territories are small island
environments that are highly vulnerable to environmental
change. They contain rare species found nowhere else on
the planet; species that have often evolved over thousands
of years in isolation from predators, competitors and
diseases, and are therefore highly susceptible to invasive
threats.
Sadly, we have already seen the loss of some unique
species, like the giant earwig of St Helena. Others, like the
endangered Henderson petrel, and the Cayman blue iguana,
remain under pressure from invasive species. Crucial work
is underway to save the Monserrat mountain chicken, a
unique frog which nearly disappeared from the island
following the incursion of an aggressive fungal disease.
Territory Governments are increasingly alive to these
issues and addressing them head on, putting in place
biosecurity regimes to prevent new introductions and
manage existing threats.
On the ground, the National Trusts of Monserrat and
Saint Helena, for instance, have worked in partnership with
the RSPB to protect critical habitats and manage the impact
of invasives on two unique bird species: the Monserrat
Oriole, the national bird of Monserrat, and the Saint Helena
Plover, the island’s last remaining unique bird. In no small
part thanks to these efforts, as of last December, these
unique birds are no longer listed as critically endangered.
I am delighted that the Governments of Ascension, the
British Virgin Islands, the Falkland Islands, and Tristan
da Cunha are here in Dundee this week to share their
experiences so we can learn from them.
3
Honolulu Challenge projects
As part of the UK’s contribution to the Honolulu
Challenge, the UK Government has committed £1m
to support Territories in improving their biosecurity.
The GB Non-Native Species Secretariat has already
identified the key gaps in practices and capacity. They
will now take targeted action to address them, sharing UK
expertise on pathway management, horizon scanning, pest
identification, and the development of effective legislation.
The UK Government is also contributing £1.75m to
support the work led by the Royal Society for the Protection
of Birds to restore Gough Island. Seabird populations on
the island – including the critically endangered Tristan
albatross and Gough bunting – are threatened by invasive
house mice that have evolved to become the largest in the
world. Every year, an estimated 900,000 seabird chicks are
killed. The aerial eradication operation planned for 2019
can turn things around for this precious World Heritage
Site. The RSPB, working in partnership with the Tristan
da Cunha Government, is making excellent progress in
preparing for the operation and is working hard to attract
further support for this vital project.
The teams delivering both of these projects are here in
Dundee this week. I know that they have planned useful
discussions and are eager to draw on your expertise to
advance their work. I wish you all possible success.
Learning from experience: South Georgia
Over the past 20 to 30 years, the pace of island
eradications has quickened and projects have become
increasingly ambitious. But each project tends to build on
what has gone before.
We in the UK have learnt a lot from the ground-breaking
work that New Zealand, Australia and South Africa have
carried out in this field. It is their systems that have been
adapted for recent work in South Georgia and that will be
applied on Gough Island in 2019.
We have also learnt from the failure of the eradication
project on Henderson Island. I know that the island
restoration community remains committed to solving such
difficult issues.
Tonight, you will hear about what has undoubtedly
been one of the most ambitious island eradications carried
out to date.
Less than a decade ago, seabirds in South Georgia,
including the unique South Georgia pipit, were in decline
and increasingly confined to a small number of rodent-free
areas.
Remarkably, by 2015, the South Georgia Heritage Trust
had completed the final steps of what has been the largest
island rodent baiting operation ever attempted.
Great credit is due to the Trust for completing this
ambitious operation – and for raising the majority of the
funds needed to support it.
When you see Tony’s pictures, you will appreciate the
harsh terrain and weather conditions faced by the Trust in
delivering this project.
4
Initial reports suggest that the endangered South
Georgia pipits are already returning to areas where
populations had previously been decimated by rats. This
is a great success for the Trust, South Georgia, and the
protection of UK biodiversity.
The UK Department for Environment, Food and
Rural Affairs is proud to have supported the Trust’s work
with £885k, including through our dedicated Overseas
Territories Environment and Climate Change grant scheme,
Darwin Plus.
The South Georgia Habitat Restoration Project sets
an outstanding example for island eradications in the UK
Overseas Territories and beyond. I hope future island
restorations will be able to emulate the project’s success.
The Trust’s work also well and truly establishes the
place of conservation charities – working in partnership
with local governments – in the field of island eradications.
I would like to pay tribute to the Trust and our other
partners in the Overseas Territories, both Governments and
charitable organisations, for their vital work.
I am delighted to now be able to hand over to Professor
Tony Martin for a full account of the Trust’s exceptional
work in South Georgia.
I regret that I will not be able to stay for Tony’s lecture
and the fascinating discussions planned for the rest of this
week, as I have to return to London to answer questions in
Parliament tomorrow.
It has been a privilege to join your discussions today.
I am sure that this will be an inspirational week, which
will be the basis for vital progress in the years to come: 44
countries collaborating together is an inspirational force.
Invasive alien species management is an area where
we can have a real and immediate positive impact and in
many cases reverse the errors of our ancestors and leave
the environment in a better state than we found it. It is our
generation’s responsibility to rise to this challenge and the
expertise of all of you at this Conference give confidence
and, importantly, hope.
PRESENTED PAPERS
Chapter 1: Rodents
With Sections: A. Planning
B. Review
C. Lessons
5
K.J. Campbell, J.R. Saah, P.R. Brown, J. Godwin, F. Gould, G.R. Howald, A. Piaggio, P. Thomas, D.M. Tompkins, D. Threadgill, J. Delborne, D.M. Kanavy, T. Kuikin, H. Packard, M. Serr and A. Shiels
Campbell, K.J.; J.R. Saah, P.R. Brown, J. Godwin, F. Gould, G.R. Howald, A. Piaggio, P. Thomas, D.M. Tompkins, D. Threadgill, J. Delborne, D.M. Kanavy, T. Kuikin, H. Packard, M. Serr and A. Shiels. A
potential new tool for the toolbox: assessing gene drives for eradicating invasive rodent populations
A potential new tool for the toolbox: assessing gene drives for
eradicating invasive rodent populations
K.J. Campbell1,2, J.R. Saah1, P.R. Brown3, J. Godwin4, F. Gould5,6, G.R. Howald1, A. Piaggio7, P. Thomas8,
D.M. Tompkins9, D. Threadgill10, J. Delborne11, D.M. Kanavy10, T. Kuiken6, H. Packard1, M. Serr4 and A. Shiels7
Island Conservation, 2100 Delaware Ave, Santa Cruz, CA, 95060, USA. <karl.campbell@islandconservation.
org>. 2School of Geography, Planning & Environmental Management, The University of Queensland, St Lucia 4072,
Australia. 3CSIRO Agriculture & Food, Black Mountain, Canberra ACT 2601, Australia. 4Department of Biological
Sciences, North Carolina State University, Raleigh, NC 27695, USA. 5Department of Entomology and Plant Pathology,
North Carolina State University, Raleigh, NC 27695, USA. 6Genetic Engineering and Society Center, North Carolina
State University, Raleigh, NC 27695, USA. 7US Department of Agriculture, National Wildlife Research Center, 4101
LaPorte Avenue, Fort Collins, CO 80521, USA. 8Department of Biological Sciences, University of Adelaide, Adelaide,
SA 5005, Australia. 9 Managing Invasives, Landcare Research Manaaki Whenua, 764 Cumberland Street, Dunedin
9016, New Zealand. 10Department of Veterinary Pathobiology, Texas A&M University, College Station, TX 77843 USA.
11
Department of Forestry and Environmental Resources, Genetic Engineering and Society Center, North Carolina State
University, Raleigh, NC 27695, USA.
1
Abstract Invasive rodents have significant negative impacts on island biodiversity. All but the smallest of rodent
eradications currently rely on island-wide rodenticide applications. Although significant advances have been made in
mitigating unintended impacts, rodent eradication on inhabited islands remains extremely challenging. Current tools
restrict eradication efforts to fewer than 15% of islands with critically endangered or endangered species threatened by
invasive rodents. The Genetic Biocontrol of Invasive Rodents partnership is an interdisciplinary collaboration to develop
and evaluate gene drive technology for eradicating invasive rodent populations on islands. Technological approaches
currently being investigated include the production of multiple strains of Mus musculus with a modified form of the native
t-complex, or a CRISPR gene drive, carrying genes or mechanisms that determine sex. These systems have the potential
to skew the sex ratio of offspring to approach 100% single-sex, which could result in population collapse. One goal
proposed is to test the ability of constructs to spread and increase in frequency in M. musculus populations in biosecure,
captive settings and undertake modelling to inform development and potential deployment of these systems. Structured
ecologically-based risk assessments are proposed, along with social and cultural engagement to assess the acceptability
of releasing a gene drive system. Work will be guided by an external ethics advisory board. Partners are from three
countries with significant regulatory capacity (USA, Australia, New Zealand). Thus, we will seek data sharing agreements
so that results from experiments may be used within all three countries and treat regulatory requirements as a minimum.
Species-specific, scalable, and socially acceptable new eradication tools could produce substantial biodiversity benefits
not possible with current technologies. Gene drive innovation may provide such a tool for invasive species management
and be potentially transformative and worthy of exploring in an inclusive, responsible, and ethical manner.
Keywords: conservation, CRISPR, genetic biocontrol, invasive species, mice, Mus musculus, pest management, public
engagement, risk assessment, transgenic
INTRODUCTION
Three Rattus species (R. rattus, R. norvegicus, R.
exulans) and house mice (Mus musculus) are, outside of
their native ranges, globally widespread invasive species
(Capizzi, et al., 2014). These invasive rodents negatively
impact stored foods, crops, and infrastructure and can
carry pathogens that impact the health of people and their
livestock (Stenseth, et al., 2003; Meerburg, et al., 2009;
Banks & Hughes, 2012). Invasive rodents cause population
declines and extinctions of island floras and faunas and
interrupt ecosystem processes with negative cascading
effects (Towns, et al., 2006; Jones, et al., 2008; Kurle, et
al., 2008; Doherty, et al., 2016). To recover endangered
populations and restore ecosystem processes, invasive
rodents on islands are increasingly targeted for eradication,
with at least 650 eradication attempts of introduced Rattus
spp. populations to-date (Russell & Holmes, 2015). These
and other island-based invasive mammal eradications have
resulted in positive responses by native species with few
exceptions (Jones, et al., 2016).
Anticoagulants are the most common control method for
invasive rodents (Capizzi, et al., 2014). Rodent eradication
on any island typically >5 ha has relied exclusively on the
use of anticoagulant toxicants incorporated into cereal or
wax baits (DIISE, 2016). Second generation anticoagulants
are most commonly used and have had the highest success
rate (Howald, et al., 2007; Parkes, et al., 2011). However,
their broad-spectrum toxicity to vertebrates, duration
of persistence, ability to biomagnify, mode of death and
negative public perception limit their responsible use
(Eason, et al., 2002; Fitzgerald, 2009; Broome, et al., 2015).
These features can lead to negative impacts, including for
conservation targets (e.g. Rueda, et al., 2016), although
significant advances in strategies to mitigate these impacts
have been made (e.g. Rueda, et al., 2019). Inhabited
islands with children, livestock and pets present significant
challenges because eradication is currently limited by a
lack of species-specific methods, animal welfare issues,
high fixed costs, and socio-political opposition (Campbell,
et al., 2015). Hence, even with optimistic assessments for
current methods (islands up to 30,000 ha and/or 1,000
people), eradications are possible on fewer than 15% of
islands with critically endangered or endangered species
threatened by invasive rodents (Campbell, et al., 2015).
New species-specific, scalable tools are needed if we are
to prevent extinctions.
Genetic biocontrol in the form of gene drives coupled
with sex-determining genes to produce single-sex offspring,
offers a potentially transformative new tool to add to the
rodent eradication toolbox, by offering species-specificity
not readily achievable in existing technology (Campbell, et
In:
6 C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 6–14. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Campbell, et al.: Assessing gene drives
al., 2015). Gene drives cause a gene to spread throughout
a population at a rate higher than would normally occur
(Champer, et al., 2016). Gene drives occur naturally and
are not recent phenomena (Lindholm, et al., 2016); for
example, mice with the native t-complex gene drive were
first described in 1927 (Schimenti, 2014). Attempts to
harness naturally-occurring gene drive systems, primarily
for invertebrate pests and disease vectors have had mixed
results (Sinkins & Gould, 2006; Champer, et al., 2016). In
2012, the Genetic Biocontrol of Invasive Rodents (GBIRd)
partnership was formed between North Carolina State
University (NCSU), Island Conservation (IC) and later
Texas A&M University (TAMU). GBIRd started exploring
opportunities for harnessing the native t-complex gene
drive in mice to eradicate invasive mouse populations
on islands (Kanavy & Serr, 2017; Piaggio, et al., 2017).
Other partners were identified through professional
networks and during searches for specific skillsets. GBIRd
currently includes seven partners in three countries:
TAMU, NCSU, University of Adelaide (UA), USA
Department of Agriculture’s National Wildlife Research
Center (NWRC), the Agriculture and Food Business Unit
of the Commonwealth Scientific and Industrial Research
Organisation (CSIRO), Landcare Research (LR), and IC.
Beginning in 2013, a harnessed bacterial immune
response system called CRISPR/Cas9 revolutionised the
field of genetic engineering. CRISPR/Cas9 can be used
to delete, modify or insert new genes more precisely,
effectively, time- and cost-efficiently than previous gene
editing tools (NASEM, 2016). Multiple genes can also
now be edited simultaneously. In 2014, a landmark paper
(building upon earlier concepts of Burt, 2003), described
how a cassette encoding the CRISPR/Cas9 machinery
could be precisely inserted into an organism’s DNA,
creating a self-replicating gene drive with potential to
modify wild populations by design (Esvelt, et al., 2014).
Since then, CRISPR/Cas9 gene drives have been developed
in yeast Saccharomyces cerevisiae (DiCarlo, et al., 2015),
fruit fly Drosophila melanogaster (Gantz & Bier, 2015)
and both Anopheles stephensi (Gantz, et al., 2015) and A.
gambiae (Hammond, et al., 2016) mosquitoes as proofof-concept demonstrations in biosecure laboratories.
This field has become a significant focus of research, and
USA and Australian Academies of Science have provided
recommendations aimed at guiding its development
(NASEM, 2016; AAS, 2017). GBIRd, with its partnership
already established, adopted CRISPR as a gene editing and
potential gene drive tool.
Gene drives are a technology platform. GBIRd
partnership considers Mus musculus the logical starting
point for developing, exploring, and providing proof-ofconcept for a genetics-based invasive vertebrate eradication
tool. They are the model vertebrate species for genetics,
possess a short generation-time, are small, husbandry is
straight-forward, and they are invasive around the world
including on many islands (Guénet & Bonhomme, 2003;
Phifer-Rixey & Nachman, 2015). Mice are also among
the best studied species in terms of mammalian sex
determination, reproductive biology, behaviour, genetic
manipulation and genetic control of phenotypic traits
(Guénet & Bonhomme, 2003; Eggers, et al., 2014; PhiferRixey & Nachman, 2015; Singh, et al., 2015). If proofof-concept, safety, and efficacy are demonstrated in Mus
musculus, it should be possible to apply this approach to
Rattus species.
The
GBIRd
programme
(<http://www.
geneticbiocontrol.org/>) aims to develop multiple gene
drive systems in mice for simultaneous evaluation of
safety and efficacy, while carefully assessing the social,
cultural and policy acceptability of such an approach. Our
staged inclusive approach reflects USA and Australian
Academies of Sciences’ recommendations (NASEM,
2016; AAS, 2017) that we treat as our minimum standards.
The GBIRd partnership aims to provide vital data for
conducting risk assessments, determining efficacy, and
engaging stakeholders and communities in order to inform
and enhance progress, or identify limitations, of future
research. A potential longer-term goal is submission of
an application to a regulatory agency for release of gene
drive constructed mice on a small, biosecure island to test
eradication of the wild, invasive mouse population.
This paper provides an overview of the GBIRd
programme as it has developed to-date, including the
risks and opportunities as they are currently envisioned
and understood. These will certainly evolve, and the
programme must strategically evolve with them.
Genetic Biocontrol of Invasive Rodents programme
The programme’s guiding principles provide context
for decision making:
● Proceed cautiously, with deliberate step-wise
methods and measurable outcomes;
● Engage early and often with the research community,
regulators, communities and other stakeholders;
● Maintain an uncompromising commitment to
biosafety, existing regulations, and protocols as
minimum standards (e.g. NASEM, 2016; AAS,
2017);
● Use, and participate in developing best practices;
● Only operate in countries with appropriate regulatory
capacity; and
● Be transparent with research, assessments, findings,
and conclusions.
1. Governance and Coordination
GBIRd involves seven organisations from Australia,
New Zealand and the USA; three universities (NCSU,
TAMU, UA), three governmental research (CSIRO, LR,
NWRC) and one non-governmental non-profit (IC). Each
has specific roles and responsibilities (Fig. 1) as detailed
in the memorandum of understanding that formalises the
partnership. A steering committee comprised of one or two
representatives from each organisation provides direction
and decision making, and a programme coordinator
facilitates activity. The consortium is inclusive and,
indeed, strengthened by a transparent internal dialogue in
both the scientific positioning (e.g. Gemmell & Tompkins,
2017) and societal/values realm (e.g. Webber, et al., 2015).
GBIRd has 14 component areas and three cross-cutting
themes (Fig. 1) being investigated, as follows.
2. Gene drives
Three gene drives are currently being investigated;
a modified t-complex, a CRISPR/Cas9 and a CRISPR/
Cpf1 gene drive. The t-complex on chromosome 17 in
mice is a natural male-transmitted meiotic drive (Lyon,
2003; Schimenti, 2014). The t-complex impairs sperm not
carrying the t-complex, leading to an increased frequency
of t-complex carrying sperm fertilising ova. The frequency
of the t-complex in natural populations of house mice is
typically lower than predicted given the often very strong
transmission ratio distortion displayed. This phenomenon
is not completely understood (see Lindholm, et al., 2016),
but may imply that a sex-biasing system based on the
t-complex would require ongoing releases to be effective
(Backus & Gross, 2016). The t-complex haplotype we are
using is free of recessive lethals and has a high rate (>95%)
of inheritance, also called transmission distortion (Kanavy
7
Island invasives: scaling up to meet the challenge. Ch 1A Rodents: Planning
Fig. 1 Programme map, showing 14 component areas being investigated by partners of the Genetic Biocontrol of Invasive
Rodents programme. The three components not linked to any organisation are cross-cutting themes.
& Serr, 2017; Piaggio, et al., 2017). The remaining
offspring (<5%) would not carry the gene drive or exhibit
the phenotypic traits of the genes being driven (Piaggio, et
al., 2017).
CRISPR/Cas9 gene drives are capable of >94%
inheritance (Gantz, et al., 2015; Hammond, et al., 2016).
Once inserted within one individual’s genome, a gene
drive can work in one of two ways. A zygotic gene drive
works when that individual’s ova or sperm are fertilised.
If the gene drive cassette is activated in the fertilised egg
(zygote), the guide RNA (gRNA) directs Cas9 to produce
a double-stranded break in the DNA at the target site in the
chromosome lacking the gene drive. This triggers the cell’s
repair mechanism to repair the break using the gene drivecontaining chromosome as a template resulting in selfreplication of the gene drive. Alternatively, in a germline
gene drive, germ cells can be targeted as the stage for selfreplication of the gene drive.
3. Targeted genes
Genes can be targeted for deletion, modification or
insertion of new genes in conjunction with a gene drive
to increase inheritance of specific traits. Investigations
currently focus on the appropriateness of two target genes
(Sry, Sox9) to be inserted and one chromosome to be
deleted (Y-’shredder’), each in coordination with a gene
drive. The Sry gene is found on the Y chromosome and
is considered the master sex-determining gene in most
mammals (Kashimada & Koopman, 2010; Eggers, et
al., 2014). Another key component of the testis pathway
is the autosomal gene Sox9, which acts immediately
downstream of Sry (Eggers, et al., 2014). Both genes
drive the development of male testes in mammals and sex
reversal has been demonstrated in transgenic female (XX)
mice (Koopman, et al., 1991; Vidal, et al., 2001; Eggers,
et al., 2014). A Y-shredder (Adikusuma, et al., 2017)
8
promotes solely offspring with one (XO) or two X (XX)
chromosomes, i.e. females. Initial developments focus on
t-complex with Sry inserted (t-Sry), and CRISPR/Cas9 and
CRISPR/Cpf1 gene drives with Sox9 and Y-shredder.
As of June 2018, partners attempting to incorporate
Sry into a t-complex drive have been challenged by the
large construct size of Sry. If that technological hurdle can
be overcome, these mice are expected to produce >95%
phenotypically male offspring (Kanavy & Serr, 2017;
Piaggio, et al., 2017). The mice currently under development
in Australia are expected to test the functionality of a split
CRISPR/Cas9 gene drive that uses phenotypic coat markers
as genetic ‘cargo’. A ‘split gene drive system’ has the gene
drive in two separate ‘cassettes’ (DiCarlo, et al., 2015).
This design is a safety feature for laboratory testing where
the separation of the cassettes results in drive components
being inherited separately even if a drive carrier were to
escape, thus preventing drive function (since both are
necessary for function). Development of CRISPR/Cpf1
gene drives and incorporating Sox9 and the Y-shredder are
underway.
4. Spatial control of gene drive
Spatially or temporally limiting drive function is one of
the major research challenges for CRISPR gene drives, e.g.
restricting a gene drive to affect only a single island’s rodent
population. Our programme is investigating genome-level
targeting of population-specific locally-fixed alleles as a
potential spatial control mechanism. It is likely that through
the process of invasion, founder effects and population
bottlenecks, certain alleles across the genome have
become fixed in any island population (Britton-Davidian,
et al., 2000; Hartl & Clark, 2006). This pattern of fixation
is likely a unique genomic signature in every genetically
isolated island population. Similar to the molecular
confinement strategy being implemented in the laboratory
Campbell, et al.: Assessing gene drives
Multiple biocontainment strategies accompany all
laboratory work and are part of our staged testing pathway
(following the recommended approach by NASEM,
2016). Recommended containment standards for gene
drives include at least two stringent confinement strategies
wherever possible, in addition to containment (Akbari, et
al., 2015; NASEM, 2016), and our programme exceeds
these standards. For example, the CRISPR gene drive
studies are using physical containment at the currently
required level (PC2) (AAS, 2017) and three containment/
confinement methods; a ‘split gene drive system’ as
explained above (DiCarlo, et al., 2015); coat colour (white
or black) to identify the zygotic homing in offspring – white
mice (Cas9-positive) are less likely to survive in the wild
(Vignieri, et al., 2010); and gRNA exclusively targeting
a synthetic sequence not present in wild mice, providing
molecular confinement to transgenic laboratory mouse
populations. For scaled laboratory trials, CSIRO and
NWRC state-of-the-art facilities provide the opportunity
to safely conduct trials with colonies of mice that could
originate from islands.
generations. Experiments in the 1980s introducing Isle of
Eday mice to the Isle of May (57 ha) demonstrate the power
of selecting appropriate stock for facilitating introduced
individuals ‘invading’ another population (Berry, et al.,
1991; Jones, et al., 1995). A Y-chromosome (i.e. male)
linked marker spread across the Isle of May site within six
months and in 18 months only hybrids could be detected
(Berry, et al., 1991; Jones, et al., 1995). The 42 Isle of Eday
males introduced were estimated at <5% of May’s resident
mouse population, demonstrating differential success of
introduced versus resident males (Berry, et al., 1991; Jones,
et al., 1995). We aim to rank the ‘invasability’ of males
from laboratory strains, selected islands and mainlands so
that appropriate stock may be selected for backcrossing in
gene drives and their cargo. Initial trials involve t-complex
carrying laboratory mice (C57BL/6/129 strain), Southeast
Farallon Island, and F1 hybrid Farallon-laboratory mice in
small cages with single males and females, to determine
if mating would occur (Serr & Godwin, 2019). (Note:
Southeast Farallon Island is not considered a potential
site for field trials at this time). Larger arenas were used
to determine mate choice and male competition where
males from different populations would have to compete
for females and resources (Serr & Godwin, 2019).
Behavioural experiments to-date indicate that t-complex
carrying lab mice can successfully mate with island mice in
captivity (Serr & Godwin, 2019). Other mate competition
results indicate that male F1 hybrid Farallon-laboratory
mice may be able to outcompete male Farallon island mice.
6. Safety and efficacy experiments
8. Island selection
Experiments demonstrating that constructs work
effectively and efficiently, are species-specific and
safe to the environment are needed. Data needs for risk
assessments and field trial applications have yet to be
determined in conjunction with regulatory agencies, and
this will dictate minimum requirements for experiments.
Experiments will inform risk assessments to reduce
uncertainty surrounding outcomes and probabilities.
Phased testing and experiments are viewed as part of
the development process, and occur at each tier (i.e.
molecular level, individual mice, mouse population,
ecological community). This phased development process
incorporates feedback loops to developers, and evaluates
efficiency, stability, specificity and safety to determine
whether a specific construct proceeds to the next stage (e.g.
molecular to insertion in a mouse or going from individual
mice to a colony). Constructs that pass will go on to more
rigorous testing, and those that don’t will either be dropped
or modified and then re-evaluated. No functional CRISPR
drives have yet been reported for vertebrates. Attempting
development of multiple combinations of gene drives and
gene targets within our programme increases the likelihood
of success, and, if successful, would provide opportunities
for comparative analyses and risk assessments. Highquality data for modelling and risk analyses will be
necessary.
As part of our staged, stepwise approach, if biosecure
laboratory studies support safety and efficacy in biasing
sex ratios and supressing test populations, the next stage
will involve studies in natural settings under conditions
where dispersal or persistence of the organisms outside
the evaluation area is restricted (NASEM, 2016). We have
identified a suite of ecological criteria for initial selection
of potentially appropriate islands for trials, including 1.
the island is biosecure (i.e. closed to public or infrequent/
controlled visitation; and remote enough (>1 km from
other land masses) to avoid unassisted immigration or
emigration), 2. no significant challenges exist to treatment
using traditional toxicant-based methods to eradicate mice
(e.g. no major non-target species, regulatory environment
allows the use of brodifacoum bait products, single land
manager), 3. M. musculus are the only rodent present
or could be introduced, and 4. the island is reasonably
economical and feasible to visit year-round (see HarveySamuel et al., 2019 for a more detailed account and
rationale). By selecting islands where the use of traditional
eradication methods could readily be used to eradicate
all rodents (Howald, et al., 2007) a contingency (i.e. exit
strategy) explicitly exists. However, these ecological
criteria are just a first filter and additional steps would be
required prior to any field trial, including engagement with
stakeholders (e.g. land managers, local communities) and
regulators to determine final approval (Harvey-Samuel et
al., 2019).
(see Biosafety), population-specific locally-fixed alleles
(and their sequence) could act as unique gRNA targets for a
CRISPR gene drive that will not function outside the island
population. Others are investigating alternative approaches
to temporally and/or spatially contain gene drives and their
relative effectiveness (e.g. Dhole, et al., 2018).
5. Biosafety
7. Mate choice
Behavioural barriers to mating success and resulting
gene flow must be considered, as to how (or if) a gene
drive will successfully spread through a population, and
if understood and used correctly may provide significant
advantage. Key characteristics influencing male
reproductive success in mice include aggressive dominance
for securing territories, and a preference among females
for unfamiliar males (Gray & Hurst, 1998; Cunningham,
et al., 2013). Promiscuity of male mice and their ability to
inseminate many females provides males the potential to
disproportionately influence the genetic makeup of future
9. Population genetic characterisation
Genetic characterisation of mouse populations from
islands selected for potential trials will occur using nextgeneration sequencing technologies (e.g. Illumina MiSeq). Analyses of these data will inform the feasibility of
using population-specific fixed allele sequences as gRNA
targets to provide spatial control of any gene drive trialled.
They will also provide baseline assessments of genetic
characteristics of target island populations, and potentially
inform future strategies.
9
Island invasives: scaling up to meet the challenge. Ch 1A Rodents: Planning
10. Modelling
12. Social engagement
Modelling can be used to inform broad strategies, such
as male or female biasing gene drives and, within those
strategies, to identify heritable traits or environmental
conditions that provide disproportionate advantages (Bax
& Thresher, 2009; Backus & Gross, 2016). Modelling is
contemplated at each development stage (i.e. molecular,
individual mouse, mouse population, ecological
community), incorporating data from experiments and
trials, and providing feedback to developers and trial
designs. It aims to predict outcomes, reduce the number
of animals required in experiments and trials and
provide insight on strategies. At the molecular level, for
example, the efficiency and stability of homing and nonhomologous end joining for Cas9 and Cpf1 zygotic and
germline homing approaches can be modelled based on
data from experiments informing on likelihood of failure
(Prowse, et al., 2017). Models also consider individual
mouse characteristics and the effects these may have at
the population level. A population model would estimate
the number of gene drive mice with certain characteristics
required for release to a specific island, the optimal
frequency, timing and location of releases, and time until
eradication. The impacts of changes to specific mouse
characteristics (or other variables) can then be estimated.
As data sets accumulate, the accuracy and sophistication
of models will increase. The opportunity exists to leverage
a 30+ year dataset and existing mouse population models,
which will facilitate sophisticated analyses and allow
the development of advanced deployment strategies that
optimise seasonal and climatic variation (Singleton, et
al., 2005; CSIRO, unpub. data). The use of these and
other models will be critical in the development of robust
ecologically-based risk assessments.
The emergence of gene drives and other genetic
technologies will force not only technologists, but
conservationists, other environmentalists and the public
to “negotiate with unfamiliar interest groups and perhaps
compromise on deeply held positions if they are going to
succeed in a complex world of contradictory perspectives”
(McShane, et al., 2011, p. 969). We hope to develop guiding
principles to establish dialogue between these disparate
groups to identify and eventually negotiate trade-offs,
things that should not be traded off, and also to “render
explicit the relevant justice dimensions and principles at
play in particular contexts” (Martin, et al., 2015, p. 176).
The programme aims to establish a transparent process
that both encourages public participation and offers a
trustworthy and responsible decision pathway for making
decisions about releases of gene drive organisms.
Specifically, members of our team have developed
a three-part plan for social engagement. First, we will
conduct a stakeholder landscape analysis to understand the
mix of interests, priorities, concerns, and hopes of diverse
stakeholders that surround the programme. Second, we
will convene a stakeholder workshop to create a forum for
discussion, provide feedback to the technical project team,
and strategise the design of community engagements.
Third, we propose to organise community focus groups
near potential island release sites to engage relevant publics
sufficiently early to influence technological innovation
and field trial research (see Chapter 7, NASEM, 2016).
Importantly, the international nature of our partnership will
foster the sharing of best practices – and challenges – of
social engagement across different cultural contexts.
To-date, engagements have occurred with publics,
scientists, conservationists, indigenous groups and other
stakeholders (including those opposing gene drive research,
Borel, 2017; Reese, 2017), but more work is required.
11. Risk assessment
There is the possibility that releases of gene drivemodified organisms will lead to unpredicted and undesirable
side effects. Ecologically-based risk assessments (EBRA)
aim to reduce some types of uncertainty surrounding
outcomes and probabilities (NASEM, 2016; AAS, 2017).
They are used to estimate the probability of immediate and
long-term environmental and public health harms. EBRAs
allow alternative strategies to be compared (e.g. traditional
use of toxicants), incorporate the concerns of relevant
publics, and can be used to identify sources of uncertainty,
making them well-suited to inform research directions
and support public policy decisions about emerging gene
drive technologies. EBRAs provide the ability to trace
cause-and-effect pathways and the ability to quantify the
probability of specific outcomes. We regularly consult with
risk assessment experts leading other gene drive EBRAs
and plan to apply specific tools to identify where, within
our development process, additional studies are required
to reduce uncertainties, complementing regulatory
requirements. The large existing body of work on rodent
eradications, including the potential ecological impacts
from toxicant use (Broome, et al., 2015) and probability of
success of traditional methods (DIISE, 2016), along with
meta-data analyses on the ecological impacts of removing
invasive rodents (Jones, et al., 2016) will facilitate rigorous
EBRAs. Our staged experimental approach prior to any
potential release would culminate in trials within biosecure
simulated natural environments with colonies of mice
imported from the target island(s) with the most efficacious
gene drive mice. This allows simulations of various
ecological scenarios and increases the power of predictive
analyses, resulting in increased levels of certainty around
potential outcomes and ecological impacts.
10
13. Communications and outreach
The investigation requires clear, concise, and
transparent communications to ensure public perceptions
by target audiences are based on facts, and not unduly
influenced by scientifically-unsubstantiated fears and
hyperbole. Communicating to stakeholders, researchers,
communities, and decision-makers interested in this
evaluation is the foundation of the programmatic principle
of transparency. Coordinated external communications
by the partnership’s representatives through media, in
peer-reviewed publications, presentations, and one-onone outreach have and will continue to be core to our
mission. Informing stakeholders and decision-makers in
fora such as the IUCN’s World Conservation Congress and
the United Nations’ Convention on Biological Diversity
encourages public discourse about this innovation,
engages thought leaders in making our investigations
more robust, ensures that fact-based concerns can be
addressed while unsubstantiated fears can be allayed, and
helps guide decision-makers in developing policies and
guidelines complementary to the precautionary, stepwise
research guiding principle, even as the technology is being
developed.
14. Ethics
There are considerable potential benefits of this
technology and we are committed to exploring it in
a responsible and inclusive manner. But the question
remains, if the technology works, should it be used? This
key ethical question is best answered once robust EBRAs
have been completed and in the context of rigorous social
and regulatory engagement. The USA and Australian
Campbell, et al.: Assessing gene drives
Academies of Science recommend that research continue
and decisions to release gene drives continue to be made
on a case-by-case basis following a comprehensive
environmental risk assessment that includes ecological
and evolutionary modelling (NASEM, 2016; AAS,
2017). We have volunteered our programme as a case
study for discussion at various fora, including ethical
deliberations amongst ethicists and peers (e.g. NCSU
Genetic Engineering and Society Center, 2016; Leitschuh,
et al., 2018), on national radio (Barclay, 2017) and for the
USA National Academies of Sciences Engineering, and
Medicine’s report on gene drives (case study 4, NASEM,
2016). Emulating the Target Malaria partnership (<http://
targetmalaria.org/>), an independent ethics advisory board
has been established to provide advice on ethical matters
and identify issues for the partnership’s consideration.
constructs to ensure appropriate characteristics is clear.
Technical issues may arise, and data needs for risk
assessments and field trial applications have yet to be
determined in conjunction with regulatory agencies. The
timeline for completion of experimental biocontained
trials is also uncertain as not all funding has been secured,
processes are of uncertain duration in some cases and
requirements for experiments have not yet been determined
in conjunction with regulators. Considering these caveats,
we estimate US$16–22M will be needed over the next 4–5
years to complete experimental biocontained trials.
All programme areas are unfunded or partially funded
at this time. We are actively pursuing opportunities for
complementary funding.
15. Regulatory
Unlike incremental advancements in current technology
or tools, the development of transformative applications
cannot be undertaken within existing rodent eradication
projects on islands or as part of rodent control on mainlands.
Transformative innovations require deliberate intent and
focussed programmes. GBIRd includes interdisciplinary
scientists, varied experience, backgrounds and viewpoints.
An analysis of the hazards associated with a hypothetical
split gene drive is underway. If proof of concept of the
gene drive can be established in laboratory populations,
and suitable target populations can be identified, funding
will be sought to perform a risk assessment building on
the results of the hazard analysis. GBIRd is also engaging
with independent external ethicists to develop best practice
ethical conduct for gene drives. Indeed, as a programme
we have attempted to maintain a balanced approach and
wish to inform future decisions with the best science at that
time. This does not preclude pursuing a pathway to broader
deployment of this type of technology if, indeed, it proves
to be safe, efficacious, and socially accepted.
Our regulatory engagement strategy is to ensure
transparent and early engagement with the regulatory
agencies responsible for the oversight and review of the
program. Varying regulatory maturity exists around the
world, with Australia and New Zealand having possibly
the most developed and mature biotechnology regulatory
review processes. The USA is revising regulatory guidelines
through the Coordinated Framework for the Regulation of
Biotechnology (Barbero, et al., 2017). Currently, in the
USA it is likely the Food and Drug Administration will
lead regulatory review of GBIRd.
Regulatory data-sharing agreements for registration of
pesticides exist between Australia, New Zealand, and USA,
and we anticipate that this will carry over to review of
biotechnology. The design, execution, and data collection
will be compliant with all three countries’ regulatory
agency requirements or under data sharing agreements.
The regulatory oversight and testing is intended to
demonstrate efficacy and safety of the construct, i.e. does it
work and what are the ecological consequences. Managing
risks associated with its potential release, including capacity
to “shut off” in vivo in case of unanticipated consequences
is one hallmark of our programme. Testing will take
place in a step-wise manner, laboratory development and
characterisation, laboratory testing, pen trials and field
trials. With the lack of clarity of regulatory pathways at this
time, we are engaging regulators early, and have done so
in Australia, New Zealand and USA to inform and ideally
strengthen regulatory standards, while ensuring open
dialogue and regulatory awareness of GBIRd exists.
16. Intellectual property
A patent for RNA-guided gene drives was filed in
2014 and two competing patents exist over CRISPR gene
editing technology (Egelie, et al., 2016; AAS, 2017).
However, there may be little scope for commercialisation
for CRISPR/Cas9 gene drives for conservation and
public health purposes (AAS, 2017). The intent of our
partnership is to safely and effectively develop and assess
this technology in a socially responsible manner that
democratises the science involved with the innovation. Our
partnership is composed of organisations that are dedicated
to the public good potential of this technology. We intend
for intellectual property to be secured in a manner that
prevents unintended use but allows maximum benefit for
communities and environments in need. The mechanisms
with which to do this have not yet been identified.
17. Financial
Budget estimates until completion of experimental
biocontained trials are uncertain until refinement of
DISCUSSION
In addition to impacting biodiversity on islands,
invasive rodents also negatively impact the health of
people and their livestock, and greatly reduce agricultural
productivity, stored food stocks and damage infrastructure.
In the future, these problems may also benefit from the
application of gene drive systems in invasive rodents.
However, the GBIRd programme is currently focussed on
the development and evaluation of gene drives in invasive
rodents on islands to prevent biodiversity loss. We are
committed to a deliberate and step-wise approach following
National Academies’ recommendations (NASEM, 2016;
AAS, 2017).
Eradication is a biological extreme involving all
individuals in a population (Parkes & Panetta, 2009).
Populations hold a diversity of genes that provide
plasticity in behaviours and susceptibilities (e.g. Buckle
& Prescott, 2012; Cunningham, et al., 2013). Eradication
of a population requires that eradication method(s)
overcome this variability (Parkes & Panetta, 2009). That
we are looking to develop an eradication (i.e. complete
and permanent removal of a population), and not a control
(i.e. frequent removal of a portion of a population for
perpetuity) tool, is intentional and strategic. Eradication
provides permanent solutions and for invasive species is
nearly always desirable when it can be achieved (Parkes
& Panetta, 2009). Eradication methods may be used for
control, but not necessarily vice-versa. Our methods must
be robust enough to eradicate populations independent
of their variability but specific enough, or controlled in
some way, that the global population (especially native
populations) are not at risk. The concept of eradication
units is a useful way to think of this (Robertson & Gemmell,
11
Island invasives: scaling up to meet the challenge. Ch 1A Rodents: Planning
2004). Are there alleles shared by all individuals (i.e. fixed)
within invasive populations that are not found in the native
population, or only a subset of individuals have? Gene
drive could be contained under either of these scenarios.
GBIRd is attempting to identify island-specific locallyfixed alleles that would provide molecular confinement
of the gene drive to the target island population. If this
is possible, potential exists for the approach to be scaled
(e.g. where locally-fixed alleles can be identified for
archipelagos, or for invasive but not native populations).
Further, our programme is also researching differential
mating success of males between populations to be able to
select the most effective stock for transmitting a gene drive
and associated genes to a target population.
CRISPR has transformed gene editing and CRISPR
gene drives are providing similar transformational
opportunities for genetic pest management (Webber, et
al., 2015; Harvey-Samuel, et al., 2017). Our partnership
was formed prior to these revolutionary tools, providing a
ready foundation upon which we expanded our partnership
and incorporated these tools, increasing the number of
technical approaches and likelihood of success. CRISPR,
as an editing tool, has also increased the efficacy of
inserting large genetic sequences (e.g. 10kb Sry) and due
to its precision, efficacy and high success rate has often
reduced the number of animals required compared to
previous approaches. We anticipate there will be other
opportunities, technological or otherwise, that emerge
throughout the life of our programme.
CRISPR has been shown to be able to edit DNA in a
range of taxa (NASEM, 2016; AAS, 2017) and a CRISPR
gene drive has advantages when developing a technology
platform, when compared to the t-complex drive which
may not be effective in species other than mice. However,
the t-complex provides options and, being naturally
occurring in mice, may increase social acceptability, or be
technically more appropriate for certain situations. Having
multiple gene drives and target genes or mechanisms
allows for many potential combinations and simultaneous
comparisons in efficacy, safety and acceptability. We
are currently investigating various combinations of
gene drive mechanisms (i.e. t-complex, CRISPR/Cas9,
CRISPR/Cpf1) and target genes or deletion mechanisms
(i.e. Sry, Sox9, Y-shredder), providing multiple potential
combinations.
Spatial control and remediation of CRISPR/Cas9 gene
editing and gene drives has been a major concern and is
the focus of significant research. We are keeping abreast
of advances in this field and will look to incorporate
mechanisms developed where appropriate. Recent research
identified CRISPR/Cas9 inhibitors that can block genome
editing, providing a means to spatially, temporally, and
conditionally control Cas9 activity (Pawluk, et al., 2016;
Rauch, et al., 2017). As a nascent field, it is understandable
that not all technological concerns have yet been addressed
(NASEM, 2016; AAS, 2017), but a significant amount of
research is underway to do so.
Few, if any, people are opposed to preventing
extinctions but there is mixed opinion about the methods
by which this is done. Rodent eradication on islands of
any significant size can currently only be implemented
with toxicants, the least publicly accepted of all control
methods (Fitzgerald, 2009). Gene drives hold promise as
putting an additional tool in the practitioner’s toolbox that
could increase the feasibility and scale of conservation
efforts. In contrast to toxicant-based invasive rodent
eradication campaigns characterised by a short duration of
implementation and high fixed costs (Howald, et al., 2007;
Holmes, et al., 2015), gene drive approaches could provide
12
an alternative and flexible financial model. Alternative
financial mechanisms such as endowments covering
annual costs instead of single campaigns costing tens
of millions of dollars may be feasible. If the anticipated
species specificity holds true, risks from methods to nontarget species (e.g. raptors, Rueda, et al., 2016) would be
eliminated and the ability for non-specialists to implement
projects would increase. Animal welfare concerns over
the mode of death of rodents and non-target species from
toxicants could be alleviated by gene drives that bias
the sex of invasive populations as no animals would be
killed (Dubois, et al., 2017). This approach could also
facilitate potential future developments with other invasive
mammals beyond rodents, including foxes (Vulpes vulpes)
and rabbits (Oryctolagus cuniculus) in Australia (Kinnear,
et al., 2016; AAS, 2017), brushtail possums (Trichosurus
vulpecula), and stoats (Mustela erminea; Owens, 2017) in
New Zealand. New Zealand has set a goal of eradicating
invasive mammal predators from their country (‘Predator
Free New Zealand 2050’ – New Zealand, 2016). One
interim 2025 goal in this strategy is to develop a scientific
breakthrough capable of removing at least one small
mammalian predator from New Zealand entirely (New
Zealand, 2016), and gene drive is one of a suite of potential
innovations currently being considered. Globally, invasive
rodents are linked to 30% of all extinctions (Doherty, et al.,
2016), and currently threaten 88% of all insular critically
endangered or endangered terrestrial vertebrates (TIB
Partners, 2014). New, scalable, species-specific tools are
needed to prevent further extinctions. The opportunity that
gene drives as a transformative technology may bring to
invasive species management is significant and worthy of
exploring in a responsible and inclusive manner.
ACKNOWLEDGEMENTS
This programme’s financial support has come from
CSIRO, IC, LR, NCSU, NWRC, TAMU, UA, US National
Science Foundation (to NCSU; NSF IGERT grant #
000166685), and The Seaver Institute (to IC). Thanks to N.
Holmes, G. Baxter and two anonymous peer reviewers for
suggestions that improved the manuscript.
REFERENCES
AAS. (2017). Synthetic Gene Drives in Australia: Implications of
Emerging Technologies. Australian Academy of Science. <www.
science.org.au/gene-drives>. Accessed: 17 November 2017.
Adikusuma, F., Williams, N., Grutzner, F., Hughes, J. and Thomas, P.
(2017). ‘Targeted deletion of an entire chromosome using CRISPR/
Cas9’. Molecular Therapy 25: 1736–1738.
Akbari, O.S., Bellen, H.J., Bier, E., Bullock, S.L., Burt, A., Church, G.M.,
Cook, K.R., Duchek, P., Edwards, O.R., Esvelt, K.M., Gantz, V.M.,
Golic, K.G., Gratz, S.J., Harrison, M.M., Hayes, K.R., James, A.A.,
Kaufman, T.C., Knoblich, J., Malik, H.S., Matthews, K.A., O’ConnorGiles, K.M., Parks, A.L., Perrimon, N., Port, F., Russell, S., Ueda, R.
and Wildonger, J. (2015). ‘Safeguarding gene drive experiments in the
laboratory’. Science 349: 927–929.
Backus, G.A. and Gross, K. (2016). ‘Genetic engineering to eradicate
invasive mice on islands: Modeling the efficiency and ecological
impacts’. Ecosphere 7: e01589.
Banks, P.B. and Hughes, N.K. (2012). ‘A review of the evidence for
potential impacts of black rats (Rattus rattus) on wildlife and humans in
Australia’. Wildlife Research 39: 78–88.
Barbero, R., Kim, J., Boling, T. and Doherty, J. (2017). Increasing the
Transparency, Coordination, and Predictability of the Biotechnology
Regulatory System. The White House. < https://obamawhitehouse.
archives.gov/blog/2017/01/04/increasing-transparency-coordinationand-predictability-biotechnology-regulatory>. Accessed 9 July 2017.
Campbell, et al.: Assessing gene drives
Barclay, P. (2017). ‘New Weapons in the Battle against Invasive Pests’. In:
P. Barclay (ed.) Big Ideas. Australian Broadcasting Corporation. <http://
www.abc.net.au/radionational/programs/bigideas/new-weapons-in-thebattle-against-invasive-pests/8580602>.
Bax, N.J. and Thresher, R.E. (2009). ‘Ecological, behavioral, and
genetic factors influencing the recombinant control of invasive pests’.
Ecological Applications 19: 873–888.
Berry, R., Triggs, G., King, P., Nash, H. and Noble, L. (1991).
‘Hybridization and gene flow in house mice introduced into an existing
population on an island’. Journal of Zoology 225: 615–632.
Borel, B. (2017). ‘How Genetically Modified Mice Could One Day Save
Island Birds’. Audubon magazine Summer 2017. <http://www.audubon.
org/magazine/summer-2017/how-genetically-modified-mice-couldone-day-save>.
Britton-Davidian, J., Catalan, J., da Graça Ramalhinho, M., Ganem, G.,
Auffray, J.C., Capela, R., Biscoito, M., Searle, J.B. and da Luz Mathias,
M. (2000). ‘Environmental genetics: Rapid chromosomal evolution in
island mice’. Nature 403: 158.
Broome, K.G., Fairweather, A.A.C. and Fisher, P. (2015). Brodifacoum
Pesticide Information Review. Version 2015/1. Hamilton, NZ:
Department of Conservation internal document DOCDM-25436.
Buckle, A. and Prescott, C. (2012). The Current Status of Anticoagulant
Resistance in Rats and Mice in the UK. Unpublished report from the
Rodenticide Resistance Action Group of the United Kingdom to the
Health and Safety Executive. Reading, UK: The University of Reading.
<http://www.pestmagazine.co.uk/_attachments/Resources/624_
S4.pdf>.
Burt, A. (2003). ‘Site-specific selfish genes as tools for the control and
genetic engineering of natural populations’. Proceedings of the Royal
Society of London B: Biological Sciences 270: 921–928.
Campbell, K.J., Beek, J., Eason, C.T., Glen, A.S., Godwin, J., Gould, F.,
Holmes, N.D., Howald, G.R., Madden, F.M., Ponder, J.B., Threadgill,
D.W., Wegmann, A. and Baxter, G.S. (2015). ‘The next generation
of rodent eradications: Innovative technologies and tools to improve
species specificity and increase their feasibility on islands’. Biological
Conservation 185: 47–58.
Capizzi, D., Bertolino, S. and Mortelliti, A. (2014). ‘Rating the rat: Global
patterns and research priorities in impacts and management of rodent
pests’. Mammal Review 44: 148–162.
Champer, J., Buchman, A. and Akbari, O.S. (2016). ‘Cheating evolution:
Engineering gene drives to manipulate the fate of wild populations’.
Nature Reviews Genetics 17: 146–159.
Cunningham, C., Ruff, J., Chase, K., Potts, W. and Carrier, D. (2013).
‘Competitive ability in male house mice (Mus musculus): Genetic
influences’. Behavior Genetics 43: 151–160.
Dhole, S., Vella, M.R., Lloyd, A.L. and Gould, F. (2018). ‘Invasion and
migration of spatially self‐limiting gene drives: A comparative analysis’.
Evolutionary Applications 11: 794–808.
DiCarlo, J.E., Chavez, A., Dietz, S.L., Esvelt, K.M. and Church, G.M.
(2015). ‘Safeguarding CRISPR-Cas9 gene drives in yeast’. Nature
Biotechnology 33: 1250–1255.
DIISE. (2016). ‘The Database of Island Invasive Species Eradications,
developed by Island Conservation, Coastal Conservation Action
Laboratory UCSC, IUCN SSC Invasive Species Specialist Group,
University of Auckland and Landcare Research New Zealand’. <http://
diise.islandconservation.org>. Accessed 16 February 2016.
Doherty, T.S., Glen, A.S., Nimmo, D.G., Ritchie, E.G. and Dickman, C.R.
(2016). ‘Invasive predators and global biodiversity loss’. Proceedings
of the National Academy of Sciences113: 11261–11265.
Dubois, S., Fenwick, N., Ryan, E.A., Baker, L., Baker, S.E., Beausoleil,
N.J., Carter, S., Cartwright, B., Costa, F., Draper, C., Griffin, J.,
Grogan, A., Howald, G., Jones, B., Littin, K.E., Lombard, A.T., Mellor,
D.J., Ramp, D., Schuppli, C.A. and Fraser, D. (2017). ‘International
consensus principles for ethical wildlife control’. Conservation Biology
31: 753–760.
Eason, C.T., Murphy, E.C., Wright, G.R.G. and Spurr, E.B. (2002).
‘Assessment of risks of brodifacoum to non-target birds and mammals
in New Zealand’. Ecotoxicology 11: 35–48.
Egelie, K.J., Graff, G.D., Strand, S.P. and Johansen, B. (2016). ‘The
emerging patent landscape of CRISPR-Cas gene editing technology’.
Nature Biotechnology 34: 1025–1031.
Eggers, S., Ohnesorg, T. and Sinclair, A. (2014). ‘Genetic regulation of
mammalian gonad development’. Nature Reviews Endocrinology 10:
673–683.
Esvelt, K.M., Smidler, A.L., Catteruccia, F. and Church, G.M. (2014).
‘Concerning RNA-guided gene drives for the alteration of wild
populations’. eLife: 10.7554/eLife.03401.
Fitzgerald, G. (2009). Public Attitudes to Current and Proposed Forms
of Pest Animal Control. A summary and review of the Australasian
and selected international research. Unpublished report. Canberra,
Australia: Invasive Animals Cooperative Research Centre. <https://
www.pestsmart.org.au/public-attitudes-to-current-and-proposed-formsof-pest-animal-control/>.
Gantz, V.M. and Bier, E. (2015). ‘The mutagenic chain reaction: A method
for converting heterozygous to homozygous mutations’. Science 348:
442–444.
Gantz, V.M., Jasinskiene, N., Tatarenkova, O., Fazekas, A., Macias,
V.M., Bier, E. and James, A.A. (2015). ‘Highly efficient Cas9-Mediated
gene drive for population modification of the malaria vector mosquito
Anopheles stephensi’. Proceedings of the National Academy of Sciences
112: E6736–E6743.
Gemmell, N.J. and Tompkins, D.M. (2017). ‘Gene drives and rodent
control: Response to Piaggio et al.’. Trends in Ecology and Evolution
32: 314–315.
Gray, S.J. and Hurst, J.L. (1998). ‘Competitive behaviour in an island
population of house mice, Mus domesticus’. Animal Behaviour 56:
1291–1299.
Guénet, J.-L. and Bonhomme, F. (2003). ‘Wild mice: An ever-increasing
contribution to a popular mammalian model’. Trends in Genetics 19:
24–31.
Hammond, A., Galizi, R., Kyrou, K., Simoni, A., Siniscalchi, C.,
Katsanos, D., Gribble, M., Baker, D., Marois, E., Russell, S., Burt,
A., Windbichler, N., Crisanti, A. and Nolan, T. (2016). ‘A CRISPRCas9 gene drive system targeting female reproduction in the malaria
mosquito vector Anopheles gambiae’. Nature Biotechnology 34: 78–83.
Hartl, D.L. and Clark, A.G. (2006). Principles of Population Genetics.
Oxford: Sinauer Associates.
Harvey-Samuel, T., Ant, T. and Alphey, L. (2017). ‘Towards the genetic
control of invasive species’. Biological Invasions 19: 1683–1703.
Harvey-Samuel, T.O., Campbell, K.J., Edgington, M., and Alphey, L.
(2019).
). ‘Trialling gene drives to control invasive species: what, where
and how?’ In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and
C.J. West (eds.) Island invasives: scaling up to meet the challenge, pp.
618–627. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Holmes, N.D., Campbell, K.J., Keitt, B.S., Griffiths, R., Beek, J., Donlan,
C.J. and Broome, K.G. (2015). ‘Reporting costs for invasive vertebrate
eradications’. Biological Invasions 17: 2913–2925.
Howald, G., Donlan, C.J., Galván, J.P., Russell, J., Parkes, J., Samaniego,
A., Wang, Y., Veitch, D., Genovesi, P., Pascal, M., Saunders, A. and
Tershy, B. (2007). ‘Invasive rodent eradication on islands’. Conservation
Biology 21: 1258–1268.
Jones, C.S., Noble, L.R., Jones, J., Tegelstrom, H., Triggs, G.S. and Berry,
R. (1995). ‘Differential male genetic success determines gene flow in
an experimentally manipulated mouse population’. Proceedings of the
Royal Society of London. Series B: Biological Sciences 260: 251–256.
Jones, H.P., Tershy, B.R., Zavaleta, E.S., Croll, D.A., Keitt, B.S.,
Finklestein, M.E. and Howald, G.R. (2008). ‘Severity of the effects of
invasive rats on seabirds: A global review’. Conservation Biology 22:
16–26.
Jones, H.P., Holmes, N.D., Butchart, S.H.M., Tershy, B.R., Kappes,
P.J., Corkery, I., Aguirre-Muñoz, A., Armstrong, D.P., Bonnaud, E.,
Burbidge, A.A., Campbell, K., Courchamp, F., Cowan, P., Cuthbert,
R.J., Ebbert, S., Genovesi, P., Howald, G.R., Keitt, B.S., Kress, S.W.,
Miskelly, C.M., Oppel, S., Poncet, S., Rauzon, M.J., Rocamora, G.,
Russell, J.C., Samaniego-Herrera, A., Seddon, P.J., Spatz, D.R., Towns,
D.R. and Croll, D.A. (2016). ‘Invasive mammal eradication on islands
results in substantial conservation gains’. Proceedings of the National
Academy of Sciences 113: 4033–4038.
Kanavy, D. and Serr, M. (2017). ‘Sry gene drive for rodent control: Reply
to Gemmell and Tompkins’. Trends in Ecology and Evolution 32: 315–
316.
Kashimada, K. and Koopman, P. (2010). ‘Sry: The master switch in
mammalian sex determination’. Development 137: 3921–3930.
Kinnear, J., Pentland, C., Moore, N. and Krebs, C. (2016). ‘Fox control
and 1080 baiting conundrums: Time to prepare for a CRISPR solution’.
Australian Mammalogy 39: 127–136.
Koopman, P., Gubbay, J., Vivian, N. and Goodfellow, P. (1991). ‘Male
development of chromosomally female mice transgenic for Sry’. Nature
351: 117–121.
13
Island invasives: scaling up to meet the challenge. Ch 1A Rodents: Planning
Kurle, C.M., Croll, D.A. and Tershy, B.R. (2008). ‘Introduced rats
indirectly change marine rocky intertidal communities from algae- to
invertebrate-dominated’. Proceedings of the National Academy of
Sciences 105: 3800–3804.
Reese, A. (2017). ‘New Zealand aims to eradicate invasive predators,
but winning public support may be big challenge’. Science 10 July
2017. <http://www.sciencemag.org/news/2017/07/new-zealand-aimseradicate-invasive-predators-winning-public-support-may-be-big>.
Leitschuh, C.M., Kanavy, D., Backus, G.A., Valdez, R.X., Serr, M., Pitts,
E.A., Threadgill, D. and Godwin, J. (2018). ‘Developing gene drive
technologies to eradicate invasive rodents from islands’. Journal of
Responsible Innovation. 5(supl): S121–S138.
Robertson, B.C. and Gemmell, N.J. (2004). ‘Defining eradication units to
control invasive pests’. Ecology 41: 1042–1048.
Lindholm, A.K., Dyer, K.A., Firman, R.C., Fishman, L., Forstmeier, W.,
Holman, L., Johannesson, H., Knief, U., Kokko, H. and Larracuente,
A.M. (2016). ‘The ecology and evolutionary dynamics of meiotic
drive’. Trends in Ecology and Evolution 31: 315–326.
Lyon, M.F. (2003). ‘Transmission ratio distortion in mice’. Annual
Review of Genetics 37: 393–408.
Martin, A., Akol, A. and Gross-Camp, N. (2015). ‘Towards an explicit
justice framing of the social impacts of conservation’. Conservation and
Society 13: 166–178.
McShane, T.O., Hirsch, P.D., Trung, T.C., Songorwa, A.N., Kinzig,
A., Monteferri, B., Mutekanga, D., Van Thang, H., Dammert, J.L.
and Pulgar-Vidal, M. (2011). ‘Hard choices: Making trade-offs
between biodiversity conservation and human well-being’. Biological
Conservation 144: 966–972.
Meerburg, B.G., Singleton, G.R. and Kijlstra, A. (2009). ‘Rodentborne diseases and their risks for public health’. Critical Reviews in
Microbiology 35: 221–270.
NASEM. (2016). Gene Drives on the Horizon: Advancing Science,
Navigating Uncertainty, and Aligning Research with Public Values.
Washington DC: Committee on Gene Drive Research in Non-Human
Organisms: Recommendations for Responsible Conduct; Board
on Life Sciences; Division on Earth and Life Studies; National
Academies of Sciences, Engineering, and Medicine. <https://www.
nap.edu/catalog/23405/gene-drives-on-the-horizon-advancing-sciencenavigating-uncertainty-and>.
NCSU Genetic Engineering and Society Center. (2016). A Roadmap to
Gene Drives: A Deliberative Workshop to Develop Frameworks for
Research and Governance. North Carolina State University Genetic
Engineering and Society Center. <https://research.ncsu.edu/ges/
research/projects/gene-drives-grant/>. Accessed 30 June 2017.
New Zealand. (2016). Predator Free NZ 2050 to be a Massive Team Effort.
Wellington, NZ: New Zealand Government, 25 July 2016. <https://
www.beehive.govt.nz/release/predator-free-nz-2050-be-massive-teameffort>. Accessed 15 July 2017.
Owens, B. (2017). ‘The big cull’. Nature 541: 148–150.
Parkes, J.P. and Panetta, F.D. (2009). ‘Eradication of Invasive Species:
Progress and Emerging Issues in the 21st Century’. In: M.N. Clout
and P.A. Williams (eds.) Invasive Species Management. A Handbook
of Principles and Techniques, pp. 47–60. New York, USA: Oxford
University Press.
Parkes, J., Fisher, P. and Forrester, G. (2011). ‘Diagnosing the cause of
failure to eradicate introduced rodents on islands: Brodifacoum versus
diphacinone and method of bait delivery’. Conservation Evidence 8:
100–106.
Pawluk, A., Amrani, N., Zhang, Y., Garcia, B., Hidalgo-Reyes, Y., Lee,
J., Edraki, A., Shah, M., Sontheimer, E.J. and Maxwell, K.L. (2016).
‘Naturally occurring off-switches for CRISPR-Cas9’. Cell 167: 1829–
1838.
Phifer-Rixey, M. and Nachman, M.W. (2015). ‘The natural history of
model organisms: Insights into mammalian biology from the wild house
mouse Mus musculus’. eLife 4: e05959.
Piaggio, A.J., Segelbacher, G., Seddon, P.J., Alphey, L., Bennett, E.L.,
Carlson, R.H., Friedman, R.M., Kanavy, D., Phelan, R., Redford, K.H.,
Rosales, M., Slobodian, L. and Wheeler, K. (2017). ‘Is it time for
synthetic biodiversity conservation?’. Trends in Ecology and Evolution
32: 97–107.
Prowse, T.A.A., Cassey, P., Ross, J.V., Pfitzner, C., Wittmann, T. and
Thomas, P. (2017). ‘Dodging silver bullets: Good CRISPR gene-drive
design is critical for eradicating exotic vertebrates’. Proceedings of the
Royal Society B: Biological Sciences 284(1860): 20170799.
Rauch, B.J., Silvis, M.R., Hultquist, J.F., Waters, C.S., McGregor, M.J.,
Krogan, N.J. and Bondy-Denomy, J. (2017). ‘Inhibition of CRISPRCas9 with bacteriophage proteins’. Cell 168: 150–158.
14
Rueda, D., Campbell, K.J., Fisher, P., Cunninghame, F. and Ponder, J.B.
(2016). ‘Biologically significant residual persistence of brodifacoum
in reptiles following invasive rodent eradication, Galapagos Islands,
Ecuador’. Conservation Evidence 13: 38.
Rueda, D., Carrion, V., Castaño, P.A., Cunninghame, F., Fisher, P.,
Hagen, E., Ponder, J.B., Riekena, C.A., Sevilla, C., Shield, H., Will,
D. and Campbell, K.J. (2019). ‘Preventing extinctions: planning and
undertaking invasive rodent eradication from Pinzon Island, Galapagos’
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West
(eds.) Island invasives: scaling up to meet the challenge, pp. 51–56.
Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Russell, J.C. and Holmes, N.D. (2015). ‘Tropical island conservation: Rat
eradication for species recovery’. Biological Conservation 185: 1–7.
Schimenti, J. (2014). ‘The Mouse T Complex’. In: E.C.R. Reeve and I.
Black (eds.) Encyclopedia of Genetics, pp. 287–292. Chicago, USA:
Fitzroy Dearborn.
Serr, M. and Godwin, J. (2019). ‘Towards a genetic approach to invasive
rodent eradications: assessing reproductive competitiveness between
wild and laboratory mice’. In: C.R. Veitch, M.N. Clout, A.R. Martin,
J.C. Russell and C.J. West (eds.) Island invasives: scaling up to
meet the challenge, pp. 64–70. Occasional Paper SSC no. 62. Gland,
Switzerland: IUCN.
Singh, P., Schimenti, J.C. and Bolcun-Filas, E. (2015). ‘A mouse
geneticist’s practical guide to CRISPR applications’. Genetics 199:
1–15.
Singleton, G.R., Brown, P.R., Pech, R.P., Jacob, J., Mutze, G.J. and
Krebs, C.J. (2005). ‘One hundred years of eruptions of house mice in
Australia–a natural biological curio’. Biological Journal of the Linnean
Society 84: 617–627.
Sinkins, S.P. and Gould, F. (2006). ‘Gene drive systems for insect disease
vectors’. Nature Reviews Genetics 7: 427–435.
Stenseth, N.C., Leirs, H., Skonhoft, A., Davis, S.A., Pech, R.P.,
Andreassen, H.P., Singleton, G.R., Lima, M., Machang’u, R.S. and
Makundi, R.H. (2003). ‘Mice, rats, and people: The bio-economics of
agricultural rodent pests’. Frontiers in Ecology and the Environment 1:
367–375.
TIB Partners. (2014). The Threatened Island Biodiversity Database
(TIB): Developed by Island Conservation, University of California
Santa Cruz Coastal Conservation Action Lab, Birdlife International
and IUCN Invasive Species Specialist Group. Version 2014.1. <www.
tib.islandconservation.org>. Accessed 23 December 2016.
Towns, D.R., Atkinson, I.A.E. and Daugherty, C.H. (2006). ‘Have
the harmful effects of introduced rats on islands been exaggerated?’.
Biological Invasions 8: 863–891.
Vidal, V.P., Chaboissier, M.-C., de Rooij, D.G. and Schedl, A. (2001).
‘Sox9 induces testis development in XX transgenic mice’. Nature
Genetics 28: 216–217.
Vignieri, S.N., Larson, J.G. and Hoekstra, H.E. (2010). ‘The selective
advantage of crypsis in mice’. Evolution 64: 2153–2158.
Webber, B.L., Raghu, S. and Edwards, O.R. (2015). ‘Opinion: Is CRISPRbased gene drive a biocontrol silver bullet or global conservation
threat?’. Proceedings of the National Academy of Sciences 112: 10565–
10567.
D. Capizzi, P. Sposimo, G. Sozio, F. Petrassi, C. Gotti, E. Raganella Pelliccioni and N. Baccetti
Capizzi, D.; P. Sposimo, G. Sozio, F. Petrassi, C. Gotti, E. Raganella Pelliccioni and N. Baccetti.
Black rat eradication on Italian islands: planning forward by looking backward
Black rat eradication on Italian islands: planning forward
by looking backward
D. Capizzi1, P. Sposimo2, G. Sozio1, F. Petrassi1, C. Gotti3, E. Raganella Pelliccioni3 and N. Baccetti3
Regione Lazio, Direzione Capitale Naturale, parchi e Aree Protette, via del Pescaccio 96 - 00166 Rome, Italy.
<dcapizzi@regione.lazio.it>. 2NEMO s.r.l., Piazza D’Azeglio 11 – 50121 Florence, Italy. 3Istituto Superiore per la
Protezione e la Ricerca Ambientale, via Ca’ Fornacetta 9 – 40064 Ozzano dell’Emilia (BO), Italy.
1
Abstract Since 1999, the black rat (Rattus rattus) has been eradicated from 14 Italian islands, and eradication is
ongoing on a further five islands. Most projects were funded by the European Union (EU) Life Programme. Over the
years, eradication techniques have been improved and adapted to different situations, including aerial bait distribution
on islands with large inaccessible areas, which otherwise would have relied on a manual bait distribution. A priority
list of eradications on islands, which was compiled ten years ago, has been met to a large extent, as rats have been
successfully eradicated from many islands of great importance to breeding seabirds. Despite some cases of re-invasion
occurring in early projects, advances in biosecurity measures have allowed for eradications on islands where this was
previously considered unfeasible due to a high risk of re-invasion. This paper reports on black rat eradication work
performed on Italian Mediterranean islands with small villages. We show biodiversity benefits of these programmes, but
also qualitatively address socio-economic and health impacts on local communities. Eradication projects have faced new
obstacles, due to recent changes in legislation which complicated the application of rodenticides and made it very difficult
to get permission for aerial distribution of bait on some of the priority islands.
Keywords: biosecurity, cost-effectiveness, invasive alien species, Rattus rattus, reinvasion, shearwaters
INTRODUCTION
In recent years, a growing awareness about the
importance of the threat posed by alien species on native
ecosystems has driven an increasing number of interventions
aimed at eliminating or mitigating their impacts. Much of
the effort to restore native ecosystems has been directed
towards islands, which represent ideal environments for
implementing eradication actions, because the impact of
alien species may be especially important (e.g. Manne, et
al., 1999; Baillie, et al., 2004), and their natural isolation
helps to maintain the benefits achieved.
On Italian islands, measures to eradicate rats have
had great success. Since the late 1990s, rats have been
eradicated, or locally controlled on many islands (Capizzi,
et al., 2016), with the EU Life programme providing
important financial support, making it possible to achieve
significant conservation objectives.
Rat eradications were carried out over the years on
islands with different characteristics, and experience built
up in selecting context-sensitive materials, techniques and
strategies. Indeed, activities were carried out on islands
small and large, uninhabited or with small residential
areas, flat or with very rough terrain and with significant
differences in the presence of non-target species.
Although there have been successes over the years,
some mistakes have also been made. In our opinion, a
critical review encompassing the activities so far carried
out, along with the results achieved, can help to effectively
plan future eradications.
In this paper, we review the rat eradication actions
carried out in past years as well as those currently
implemented, highlighting the progress, problems and
constraints experienced so far, and analyse the strengths
and weaknesses of the solutions adopted. Our aim is to
show that a review of past experiences can have a positive
influence on planning for future eradication attempts.
Evolution of techniques and targets
Priority list
Since resources for conservation actions are limited,
priority setting is considered a key aspect in defining
conservation strategies (Hughey, et al., 2003; Joseph, et
al., 2009), including those involving invasive alien species
management (e.g. Gallardo & Aldridge, 2013). Capizzi,
et al. (2010) established a priority list of islands for rat
eradication on Italian islands, considering the optimal
allocation of available resources. This prioritisation
considered the number of shearwater pairs and the
monetary costs of rat eradication on each island, as well
as the risk of reinvasion. To date, all the islands in the top
five (and seven in the top ten) were included in eradication
projects performed or still ongoing (Table 1). Furthermore,
recent advances in biosecurity measures have allowed the
carrying out or planning of eradications on islands that
were previously not included on the priority list because
of a high risk of reinvasion, such as Linosa (eradication
ongoing) and Ventotene (eradication ongoing).
Island size
Since our eradication projects began, the number of rat
free islands has increased considerably. This was possible
due to increased experience and confirmation that these
interventions bring substantial benefits to birds (see below).
The first eradications in 1999–2000 were carried out
on islands of a few hectares (Table 2), but since 2005 rats
have been eradicated from islands with an area of over
100 hectares (Zannone, Giannutri, Molara). Since 2012,
islands with over 1,000 hectares have also been attempted
(success declared in 2014 for Montecristo, ongoing actions
on Tavolara and Pianosa).
Field techniques
Bait delivery
In the first eradication programmes, rodenticide
baits were placed inside bait stations, at a relatively high
density (about 10 stations/ha). In subsequent eradications,
involving islands larger than 100 ha (i.e. Giannutri and
Zannone, between 2005 and 2006), bait station density
was reduced to an average of 4/ha. On Zannone, given
its relatively rough terrain, bait distribution, in some
inaccessible areas, was carried out by hand-broadcasting
from a helicopter, using rodenticide bait blocks, which
were secured inside biodegradable dispensers (sections
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 15–20. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
15
Island invasives: scaling up to meet the challenge. Ch 1A Rodents: Planning
Table 1 List of islands prioritised for rat eradication (from Capizzi, et al., 2010) and status of eradication interventions.
Crosses indicate the presence of the two shearwater species on the various islands.
Scopoli’s
shearwater
(Calonectris
diomedea)
X
X
X
1
2
3
Tavolara
Palmarola
Montecristo
4
Pianosa Group (La Scola and Pianosa)
X
5
6
7
8
9
10
Giannutri
Santa Maria Group (14 islands)
Molara
Zannone
Spargi
Soffi Group (four islands)
X
X
X
X
X
of bamboo trunk). On larger, mainly inaccessible islands,
aerial distribution was carried out on the whole island. Bait,
in the form of pellets, was distributed using helicopters,
with an automated distributor (bucket) purchased in 2008
and used by all projects since then.
Optimisation of active ingredients
In the first eradication projects (e.g. those in 1999–2000
on small islands), both bromadiolone and brodifacoum
were used, regardless of the presence of non-target
animals. In subsequent years, on larger (> 100 ha) islands
(i.e. Giannutri, Zannone and Molara), we relied solely on
brodifacoum, which was judged, on the basis of published
data, to be the most effective and the most used active
ingredient (e.g. Howald, et al., 2007; Buckle & Eason,
2015). However, when dealing with inhabited islands
with pets (e.g. Linosa and Ventotene, where eradication
is ongoing) and livestock (Pianosa, Tavolara), we chose
to perform a two-stage bait distribution, with a different
active ingredient. In the first phase (first two distributions),
when rat populations were still at a high level, a bait
containing an active ingredient less toxic for non-target
species was used (e.g. bromadiolone or difenacoum, e.g.
Capizzi & Santini, 2007; Buckle & Eason, 2015), thereby
reducing the risks of secondary poisoning for animals that
could eat dead or dying rats. The use of brodifacoum was
limited to the last two applications (second phase), when
the population of rats was expected to have been decimated
by previous baiting campaigns, and therefore the risk of a
poisoned rat (or mouse) being eaten by a non-target species
was much lower.
Biosecurity issues
Rat reinvasion following an eradication programme
is a real threat (Russell & Clout, 2005; Russell & Clout,
2007), wasting a great deal of time and monetary effort. In
recent years, rats have reinvaded some of the islands where
they had been previously eradicated (Table 3). Reinvasion
occurred as rats swam from neighbouring islands or the
mainland (maximum distance of reinvaded islands: 320 m,
average distance: 218.6 ± 102.7 m). In the case of Molara,
the hypothesis of an unsuccessful eradication was not
supported by evidence, as genetic analyses have shown
that the reinvading rats were different from the eradicated
ones (Ragionieri, et al., 2013). The distance of Molara from
other neighbouring islands and the mainland (1,400 m),
16
Yelkouan
shearwater
(Puffinus
yelkouan)
X
X
X
X
X
X
X
X
Eradication planned in 2017
Eradication planned in 2018
Eradication in 2012
Eradication in La Scola (2000), and
Pianosa (2017)
Eradication in 2005
No action
Eradication in 2009
Eradication in 2006
No action
No action
plus the simultaneous appearance of rabbits, suggests
that they have been transported by boat. However, recent
progress in the understanding of biosecurity measures, i.e. a
better understanding of rat swimming abilities as well as of
effective quarantine measures (Russell, et al., 2008; Oppel,
et al., 2011), allowed us to plan and complete eradication
programmes on islands where there is a boat service and
on islands with small villages. Therefore, in 2016, rat
eradication was achieved on Linosa, which has a small
village of about 500 people, and has just started (January
2018) on Ventotene, which has about 700 residents. If the
Ventotene rat eradication is successful, it will be the largest
inhabited island in the Mediterranean cleared of rats.
Ecological and socio-economic benefits
Benefits for shearwaters
The detrimental impact of rats on nesting shearwaters
has been well documented on several islands, both oceanic
and Mediterranean. In the Mediterranean, observed
population declines of burrowing seabirds such as
Scopoli’s shearwater (Calonectris diomedea), yelkouan
shearwater (Puffinus yelkouan), Balearic shearwater (P.
mauretanicus) and storm petrel (Hydrobates pelagicus)
was mainly attributed to alien predators, especially rats
(e.g. Penloup, et al., 1997; Martin, et al., 2000; Igual, et al.,
2006; Baccetti, et al., 2009). Detailed surveys on Italian
islands (for survey methods see Baccetti, et al., 2009)
corroborated the evidence, showing a large difference in
terms of breeding success between islands with or without
rats; the latter included both islands where rats had never
been present and where they had been eradicated (Capizzi,
et al., 2016, Fig. 1). Pooled data from both Scopoli’s
shearwater and yelkouan shearwater indicated that breeding
pairs on islands without rats had much higher breeding
success (0.78±0.17, n=15) than those breeding on islands
with rats (0.14±0.25, n=11). Rat removal also affected the
size of shearwater colonies. At La Scola, ten years after rat
eradication, the colony of Scopoli’s shearwater increased
from 60–100 pairs in 2001 to 150–250 pairs in 2010. At
Zannone, after rat eradication (2007) there was an increase
in the Scopoli’s shearwater colony from 27 pairs in 2007
to 80 pairs in 2016.
The completed rat eradications have rendered over
1,500 ha rat-free, and ongoing or planned projects will likely
increase this surface area to 4,500 ha (Fig. 2). Currently,
Capizzi, et al.: Black rats on Italian islands
Table 2 Summary table showing the Italian islands where rat eradication was completed in the period 1999–2017, and
those where the intervention is scheduled in coming months, with details on the islands, the interventions and project
details. Success (i.e. successful eradication) was established two years after the last sign of rats.
Year
Island
Region
Bait
method
bait station
Responsible
(funding)
National Park
1999
Tuscany
Tuscany
1.3
1999
Isolotto
di Porto
Ercole
Isola dei
Topi
Peraiola
bait station
National Park
bait station
National Park
successful,
reinvaded
successful
Tuscany
1
1999
Palmaiola
Tuscany
7.2
2,950
bait station
National Park
successful
1999
Tuscany
1.9
48
Tuscany
1.6
120
2001
Gemini
Alta
Gemini
Bassa
La Scola
bait station
National Park
bait station
National Park
bait station
National Park
Brodifacoum
Brodifacoum
bait station
bait station
1,400
Brodifacoum
aerial
National Park
Circeo
National Park
MPA
successful,
reinvaded
successful,
reinvaded
successful,
new incursions
(3) promptly
eradicated
successful
successful
Tuscany
1.6
242
2006
2007
Giannutri
Zannone
Tuscany
Latium
239.4
104.7
11,471
5,700
2008
Molara
Sardinia
347.9
2008
Proratora
Sardinia
4.5
200
Brodifacoum
bait station
MPA
2010
2010
Isola Piana
Isola dei
Cavalli
Sardinia
Sardinia
13.6
2.2
551
300
Brodifacoum
Brodifacoum
bait station
bait station
MPA
MPA
2012 Montecristo Tuscany 1071
2016– Linosa
Sicily
545.1
2017
29,410
43,000
Brodifacoum
Difenacoum &
brodifacoum
aerial
bait station
National Park
Sicily Region
(LIFE)
2017
Pianosa
Tuscany 1026
13,300
National Park
to be confirmed
2017
Tavolara
Sardinia
602.0
1,150
Bromadiolone bait station
& brodifacoum
Brodifacoum
aerial
2018
Palmarola
Latium
125.1
7,300
Brodifacoum
2018
Ventotene
Latium
143.6
43,000
Municipality of
Olbia (LIFE)
Latium Region
(LIFE)
Latium Region
(LIFE)
started in
autumn 2017
started in
January 2018
started in
January 2018
1999
1999
Area Distance Active
(ha)
(m)
ingredient
6.5
320
Bromadiolone,
brodifacoum
300
30
Bromadiolone,
brodifacoum
Bromadiolone,
brodifacoum
Bromadiolone,
brodifacoum
Bromadiolone,
brodifacoum
Bromadiolone,
brodifacoum
Bromadiolone,
brodifacoum
bait station
Bromadiolone bait station
& brodifacoum
Outcome
successful,
reinvaded
successful,
reinvaded in
2010
successful,
immediately
reinvaded,
eradicated
2010, reinvaded
in 2010
successful
successful,
new incursions
(2) promptly
eradicated
successful
to be confirmed
Distance = from mainland or other islands in metres. National Park = National Park of Tuscan Archipelago (LIFE).
MPA = Marine Protected Area of Tavolara – Punta Coda Cavallo
17
Island invasives: scaling up to meet the challenge. Ch 1A Rodents: Planning
Fig. 1 Boxplot showing mean and standard deviation of
breeding success (in terms of percent chick survival) of
both shearwater species on Italian islands with (n=11)
and without (n=15) rats.
15% of the Italian population of shearwater pairs (both
Calonectris diomedea and Puffinus yelkouan computed as
the geometric mean of minimum and maximum estimates,
data from Baccetti, et al., 2009, updated when necessary)
have been released from rat predation (Fig. 3). Increased
benefits to the Italian population will occur with ongoing
and planned eradications (i.e. Linosa for Scopoli’s
shearwater and Tavolara for yelkouan shearwater).
Socioeconomic and public health issues
Islands where rats have been eradicated are uninhabited
or host just a few houses. Recently, the possibility of
conducting rat eradication programmes on islands with
small villages (Linosa, 500 residents, and Ventotene,
700) also provides significant socio-economic and
health benefits for residents and tourists (see below). As
an example, in Ventotene (120 ha, 700 inhabitants, rat
eradication funded within Life PonDerat project), we ran
a preliminary survey (performed through interviews to
residents, which is still ongoing) to estimate the economic
benefits when removing rats. First, in terms of prevented
management costs, we estimated the current yearly
quantity of rodenticides used to protect crops from rat
damage at about 100 kg, corresponding to a yearly overall
cost of about €5000. Also, the municipality runs its own
pest control activities in public areas, hiring the service of
a pest control company at an annual cost of about €3000.
Second, rat eradication brings biodiversity benefits. As bait
is generally used improperly, by using the most toxic active
ingredients (usually brodifacoum) and by distributing baits
indiscriminately, the risk to non-target species is apparent.
Eradication would reduce these non-target effects. Third,
direct damage costs are prevented because a certain
amount of crop damage still occurs despite the current use
of rodenticides which would also be prevented if rats were
eradicated.
Lastly, rat eradication brings health benefits. For
example, we recorded a 15.5% prevalence of Leishmania
infantum in Rattus rattus from Montecristo, an island far
from the mainland without carnivores (except the sporadic
presence of dogs), leading us to identify rats as possible
reservoirs and vectors of this protozoan (Zanet, et al.,
2014). On inhabited islands (e.g. Ventotene and Linosa),
it is likely that rat removal will bring health benefits by
18
Fig. 2 Results of rat eradications on Italian islands in terms
of pest free area and native species recovery since 1999
in five-year intervals a) total island surface area (ha)
freed of rats, b) number of pairs of yelkouan shearwater
(Puffinus yelkouan) and of c) Scopoli’s shearwater
(Calonectris diomedea) released from rat predation. The
graphs also include eradications where the outcome
is still to be confirmed, as well as those planned in the
coming months.
reducing the impact of rodent borne diseases, although
social costs associated with rodent-borne diseases are
difficult to quantify (e.g. World Bank, 2010). On Ventotene,
the challenge is to obtain an overall estimate of the benefits
of eradication, both ecological and socio-economic
(García-Llorente, et al., 2008).
Therefore, the associated economic benefits should
also be considered when evaluating the cost-effectiveness
of these conservation efforts, as they may confer an added
value that can help with public acceptance of this type of
project.
Impact on non-target species
Conservationists, researchers and land managers can
look pragmatically at the possible loss of individual nontarget species, by comparing them with the increased
benefits to native species and ecosystems (e.g. Ogden &
Gilbert, 2009; Capizzi, et al., 2010; Gillespie & Bennett,
Capizzi, et al.: Black rats on Italian islands
Unsolved problems and lessons learnt
Authorisation and legal aspects
Limitations resulting from the application of EU
Biocide Regulation 528/2012 represent a major obstacle
to running eradication programmes, even though this
Regulation explicitly accommodates a derogation on the
use of rodenticides (Article 43), including aspects relating
to the protection of the environment. Italian authorities
interpreted the European regulation on biocides to mean
that they should only be distributed inside bait stations,
thus implicitly forbidding aerial distribution. This has led
to legal disputes during the eradication on Montecristo,
which were resolved but will cause problems for many
eradications to come. For instance, the derogation for aerial
distribution on Tavolara (which hosts the largest colony of
Puffinus yelkouan in the world) was only obtained more
than one year after the original request, thereby risking
the loss of funding and compromising the outcome of the
project.
Dealing with stakeholders
Fig. 3 Percentage of the Italian population of yelkouan
shearwater (a) and Scopoli’s shearwater (b) nesting
pairs released from rat predation.
2017). However, minimal impacts on non-target species are
often crucial to the acceptance of the project by the general
public and are a significant factor in obtaining authorisation
from public authorities. Indeed, much of the concerns of
the public and public authorities were around the impact
on non-target species, which has been demonstrated to be
almost negligible (Capizzi, et al., 2016). In a few cases, the
actual non-target impact involved species that, following
rat removal, would have become extinct anyway, i.e. a few
pairs of nocturnal raptors (barn owl, Tyto alba). We did not
observe any impact on other rat predators, such as snakes
(green whip snake, Hierophis viridiflavus and the asp viper,
Vipera aspis), or birds of prey (kestrel, Falco tinnunculus
and peregrine falcon, Falco peregrinus).
In most cases, populations of lizards (both Podarcis
sicula and P. muralis) and native geckos have increased
since rat eradication. The populations of wild or feral
ungulates (mouflons and goats, in most cases alien species
themselves) did not experience significant impacts, despite
some losses of goats on Montecristo. Finally, no impact on
pets (dogs and cats), poultry or livestock has been recorded
so far.
It is well known that communication and information
aspects are very important in projects involving the
suppression or removal of invasive species to favour
native species or ecosystems (e.g. Larson, et al., 2011;
Adriaens, et al., 2015). In the case of island communities,
the main issue is that, if not properly communicated,
actions may be perceived as an intrusion by outsiders. Onsite meetings with island inhabitants do not always receive
good feedback. In our experience, ensuring a constant
presence in the area and establishing positive relationships
with locals are paramount to raising public awareness on
relevant conservation topics, as well as gaining project
acceptance. Public approval is indeed a key factor for
rat eradication success on islands (Epanchin-Niell, et al.,
2010).
It is also vital to establish a constructive dialogue
with port authorities and ship owners, to allow boats and
harbours to be monitored, so that rats cannot be transported
with the possibility of them being distributed across the
island. This is especially important on islands served by
regular ship visits, such as those hosting small villages
(e.g. projects ongoing on Linosa and Ventotene).
Learning from failures
As mentioned above, the analysis of recolonisations
following eradications has allowed us to conclude that
islands closely neighbouring other rat-inhabited islands
present a high risk of re-invasion after a successful
eradication operation. The case of La Scola Island is
representative, with three reinvasions in about fifteen
years. The eradication of rats from the nearby (320 m)
island of Pianosa will solve the problem permanently.
Rat eradication on Molara represents a different case of
reinvasion. The island was reinvaded a few months after
an apparently successful rat eradication, but invading
rats were genetically different from the eradicated rats
(Ragionieri, et al., 2013). We strongly suspect that this
recolonisation event represents a case of sabotage, possibly
caused by the hostility of some people towards the project:
the simultaneous appearance of rabbits on the island
corroborated this hypothesis. This confirms the importance
of properly addressing community opinions (Genovesi &
Bertolino, 2001) and trying to highlight critical issues that
may otherwise compromise the outcome of the project. To
avoid the voluntary release of rats on rat-free islands, it
is crucial to implement long term biosecurity and provide
the necessary human resources for continuous awarenessraising.
19
Island invasives: scaling up to meet the challenge. Ch 1A Rodents: Planning
REFERENCES
Adriaens, T., Baert, K., Breyne, P., Casaer, J., Devisscher, S., Onkelinx,
T., Pieters, S., Stuyck, J., (2015). ‘Successful eradication of a suburban
Pallas’s squirrel Callosciurus erythraeus (Pallas 1779) (Rodentia,
Sciuridae) population in Flanders (northern Belgium)’. Biological
Invasions 17: 2517–2526.
Baccetti, N., Capizzi, D., Corbi, F., Massa, B., Nissardi, S., Spano, G. and
Sposimo, P. (2009). ‘Breeding shearwater on Italian islands: population
size, island selection and co-existence with their main alien predator’.
Rivista Italiana di Ornitologia 78: 83–99.
Baillie, J.E.M., Hilton-Taylor, C. and Stuart, S.N. (2004). 2004 IUCN Red
List of Threatened Species: A Global Species Assessment. Occasional
Paper SSC no. 62. Gland, Switzerland: IUCN.
Buckle, A.P. and Eason, C.T. (2015). ‘Control Methods: Chemical’. In:
A.P. Buckle and R.H. Smith (eds.) Rodent Pests and their Control, pp.
123–154. Wallingford, UK: CABI.
Capizzi, D. and Santini, L. (2007). I Roditori Italiani: Ecologia, Impatto
sulle Attività Umane e sugli Ecosistemi, Gestione delle Popolazioni.
Roma: Antonio Delfino Editore.
Capizzi, D., Baccetti, N. and Sposimo, P. (2010). ‘Prioritizing rat
eradication on islands by cost and effectiveness to protect nesting
seabirds’. Biological Conservation 14: 1716–1727.
Capizzi, D., Baccetti, N. and Sposimo, P. (2016). ‘Fifteen Years of Rat
Eradication on Italian Islands’. In: F.M. Angelici (ed.) Problematic
Wildlife, pp. 205–227. Switzerland: Springer International Publishing.
Epanchin-Niell, R.S., Hufford, M.B., Aslan, C.E., Sexton, J.P., Port, J.D.
and Waring, T.M. (2010). ‘Controlling invasive species in complex
social landscapes’. Frontiers in Ecology and the Environment 8: 210–
216.
Gallardo, B. and Aldridge, D.C. (2013). ‘Priority setting for invasive
species management: risk assessment of Ponto-Caspian invasive species
into Great Britain’. Ecological Applications 23: 352–364.
García-Llorente, M., Martín-López, B., González, J.A., Alcorlo, P. and
Montes, C. (2008). ‘Social perceptions of the impacts and benefits
of invasive alien species: implications for management’. Biological
Conservation 141: 2969–2983.
Genovesi, P. and Bertolino, S. (2001). ‘Human Dimension Aspects in
Invasive Alien Species Issues: The Case of the Failure of the Grey
Squirrel Eradication Project in Italy’. In: McNeely, J. (ed.) The Great
Reshuffling: Human Dimensions of Invasive Alien Species, pp. 113–119.
Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Gillespie, R. and Bennett, J. (2017). ‘Costs and benefits of rodent
eradication on Lord Howe Island, Australia’. Ecological Economics
140: 215–224.
Howald, G., Donlan, C., Galván, J.P., Russell, J.C., Parkes, J., Samaniego,
A., Wang, Y., Veitch, D., Genovesi, P., Pascal, M., Saunders, A. and
Tershy, B. (2007). ‘Invasive rodent eradication on islands’. Conservation
Biology 21: 1258–1268.
Hughey, K.F., Cullen, R. and Moran, E. (2003). ‘Integrating economics
into priority setting and evaluation in conservation management’.
Conservation Biology 17: 93–103.
Igual, J.M., Forero, M.G., Gomez, T., Orueta, J.F. and Oro, D. (2006). ‘Rat
control and breeding performance in Cory’s shearwater (Calonectris
diomedea): Effects of poisoning effort and habitat features’. Animal
Conservation 9: 59–65.
Joseph, L.N., Maloney, R.F. and Possingham, H.P. (2009). ‘Optimal
allocation of resources among threatened species: A project prioritization
protocol’. Conservation Biology 23: 328–338.
Larson, D.L., Phillips-Mao, L., Quiram, G., Sharpe, L., Stark, R., Sugita,
S. and Weiler, A. (2011). ‘A framework for sustainable invasive species
management: environmental, social, and economic objectives’. Journal
of Environmental Management 92: 14–22.
Manne, L.L., Brooks, T.M. and Pimm, S.L. (1999). ‘Relative risk of
extinction of passerine birds on continents and islands’. Nature 399:
258–261.
Martin, J.L., Thibault, J.C. and Bretagnolle, V. (2000). ‘Black rats,
island characteristics, and colonial nesting birds in the Mediterranean:
consequences of an ancient introduction’. Conservation Biology 14:
1452–1466.
Ogden, J. and Gilbert, J. (2009). ‘Prospects for the eradication of rats from
a large inhabited island: Community based ecosystem studies on Great
Barrier Island, New Zealand’. Biological Invasions 11: 1705–1717.
Oppel, S., Beaven, B.M., Bolton, M., Vickery, J. and Bodey, T.W. (2011).
‘Eradication of invasive mammals on islands inhabited by humans and
domestic animals’. Conservation Biology 25: 232–240.
20
Penloup, A., Martin, J.L., Gory, G., Brunstein, D. and Bretagnolle, V.
(1997). ‘Distribution and breeding success of pallid swift, Apus pallidus,
on Mediterranean islands: Nest predation by the roof rat, Rattus rattus,
and nest quality’. Oikos 80: 78–88.
Ragionieri, L., Cutuli, G., Sposimo, P., Spano, G., Navone, A., Capizzi,
D., Baccetti, N. and Fratini, S. (2013). ‘Establishing the eradication
unit of Molara Island: A case of study from Sardinia, Italy’. Biological
Invasions 15: 2731–2742.
Russell, J.C. and Clout, M.N. (2005). ‘Rodent Incursions on New Zealand
islands’. In: J. Parkes, M. Statham and G. Edwards (eds.) Proceedings
of the 13th Australasian Vertebrate Pest Conference, pp. 324–330.
Lincoln, New Zealand: Landcare Research.
Russell, J.C. and Clout, M.N. (2007). ‘Early Detection and Eradication of
Invading Rats’. In: W.C. Witmer, K.A. Pitt and K.A. Fagerstone (eds.)
Managing Vertebrate Invasive Species, pp. 268–272. Fort Collins, CO:
USDA/APHIS/WS, National Wildlife Research Center.
Russell, J.C., Beaven, B.M., MacKay, J.W.B., Towns, D.R. and Clout,
M.N. (2008). ‘Testing island biosecurity systems for invasive rats’.
Wildlife Research 35: 215–221.
World Bank (2010). People, Pathogens and our Planet, Volume 1:
Towards a One Health Approach for Controlling Zoonotic Diseases.
Washington: The World Bank.
Zanet, S., Sposimo, P., Trisciuoglio, A., Giannini, F., Strumia, F.
and Ferroglio, E. (2014). ‘Epidemiology of Leishmania infantum,
Toxoplasma gondii, and Neospora caninum in Rattus rattus in absence
of domestic reservoir and definitive hosts’. Veterinary Parasitology
199: 247–249.
M. Duhr, E.N. Flint, S.A. Hunter, R.V. Taylor, B. Flanders, G. Howald and D. Norwood
Duhr, M.; E.N. Flint, S.A. Hunter, R.V. Taylor, B. Flanders, G. Howald and D. Norwood. Control
of house mice preying on adult albatrosses at Midway Atoll National Wildlife Refuge
Control of house mice preying on adult albatrosses at Midway Atoll
National Wildlife Refuge
M. Duhr1, E.N. Flint2, S.A. Hunter2, R.V. Taylor3, B. Flanders4, G. Howald5 and D. Norwood2
Mid-Columbia River NWRC, 64 Maple St., Burbank, Washington 99323 USA. 2US Fish and Wildlife Service, Marine
National Monuments of the Pacific. 300 Ala Moana Blvd. Honolulu, Hawaii 96850 USA. <Beth_Flint@fws.gov>.
3
National Wildlife Refuge Association, Washington, DC 20036, USA. 4US Fish and Wildlife Service, Pacific Region One,
National Wildlife Refuges, 911 NE 11th Avenue, Portland, Oregon 97232. 5Island Conservation, 2100 Delaware Ave,
Santa Cruz, CA, 95060, USA.
1
Abstract Sand Island, Midway Atoll National Wildlife Refuge (MANWR), is home to 21% of all nesting black-footed
albatross (Phoebastria nigripes) and 47% of all nesting Laysan albatross (P. immutabilis) worldwide. During the 2015–
2016 nesting season predation and disturbance by non-native house mice (Mus musculus), here documented for the first
time, resulted in 70 abandoned nests, 42 adult birds killed and 480 wounded. In the following nesting season the affected
area increased, resulting in 242 dead adults, 1,218 injured birds and 994 abandoned nests. Mouse predation activities
triggered a mouse control response to reduce mouse densities in the affected areas using multi-catch live traps, kill
traps, and limited use of anticoagulant rodenticides in bait stations. In 2016–2017 we applied a pelleted cholecalciferol
rodenticide, AGRID3 (Bell Laboratories, Madison, WI), at a rate of 20 kg/ha in all affected areas. The purpose of this study
was to evaluate the efficacy of using AGRID3 to reduce mouse density and rate of mouse attacks on nesting albatrosses on
Sand Island. Mouse attacks decreased and mouse abundance was reduced following rodenticide applications in the plots
treated in December but changes in attack rates in the plots treated in January were not detectable and mouse abundance
increased subsequent to treatment. The plots in the December treatments were much larger than those used in January
and rainfall rate increased after December. A minimum size of treatment area may be necessary to achieve a reduction in
injury rates in albatrosses. No deleterious effects were observed in non-target organisms. The casualties resulting from
mouse predation (mostly Laysan albatross) represent a small proportion of the 360,000 pairs nesting on Sand Island.
However, the risk to adult breeding albatrosses representing such a large fraction of the global population prompted the
United States Fish & Wildlife Service to prioritise mouse control efforts.
Keywords: cholecalciferol, non-target species, Pacific, rodent, seabirds, tropical
INTRODUCTION
Midway Atoll National Wildlife Refuge (MANWR) is
home to over three million birds representing 29 species
including species of conservation concern and the largest
albatross breeding colony in the world. MANWR supports
36% of the earth’s black-footed albatross (Phoebastria
nigripes) and 73% of all Laysan albatross (P. immutabilis).
Of the three islands that make up the refuge, Sand Island is
the largest and provides habitat to approximately 360,000
breeding pairs of Laysan albatross, making it a globally
significant colony. House mice (Mus musculus) were
introduced to Sand Island more than 75 years ago and
persisted after black rats (Rattus rattus) were eradicated in
1996. Until recently, these non-native mammals appeared
to co-exist with the refuge’s large seabird populations
without harm.
This changed in December 2015 when, shortly after
the initiation of the albatross breeding season, severe
wounds were discovered on the dorsa of several incubating
albatrosses on Sand Island and images from motionsensing cameras revealed that the source of the wounds
were mice (Fig. 1). This was the first time house mice
had been observed attacking adult albatrosses and the
first documentation of mice preying on albatross in the
Northern hemisphere. House mice had not been considered
a threat to seabird populations until 2001 when they were
found preying on albatross chicks as well as other seabird
species at two sites in the Southern hemisphere (Cuthbert
& Hilton, 2004; Angel, et al., 2009; Jones & Ryan, 2009).
The discovery of attacks by mice on Sand Island
caused immediate concern for wildlife managers at
the refuge. Adult mortality has the strongest effect on
population growth rates in species such as albatrosses
with low fecundity, longevity, high age at first breeding,
and prolonged parental care. The loss of the breeding adult
is compounded by the loss of its egg or chick, and also
reduces the fecundity of its surviving mate, as it often takes
more than a year for a widowed bird to find a new mate. In
response to the attacks first discovered in December 2015,
emergency control efforts were immediately initiated at a
5 m grid resolution over attack areas using a combination
of available methods; live traps, kill traps, and difethialone
rodenticide applied in bait stations near structures.
When albatrosses returned to Sand Island in the autumn
of 2016, surveys were initiated to look for signs of mouse
attack and it quickly became clear that mice were attacking
the albatross again. Moreover, the rate at which birds
were being killed or injured suggested that the 2016–2017
outbreak might be much greater than during the previous
year. This time, however, United States Fish & Wildlife
Service (USFWS) staff had a plan and were prepared to
address the situation. Research had suggested that AGRID3
Fig. 1 Introduced house mouse attacking adult Laysan
albatross as it incubates. As captured by a Reconyx trail
camera.
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 21–25. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
21
Island invasives: scaling up to meet the challenge. Ch 1A Rodents: Planning
(Bell Laboratories, Madison, WI), a cholecalciferol
rodenticide, might provide an effective tool for reducing
the number of mice in areas where they were attacking
albatross, thereby reducing the impacts to the nesting
birds. A plan was developed for applying the rodenticide
in affected areas and also for measuring the effects of the
treatments on both mice and nesting albatross.
In this paper we describe the mouse predation on
albatross that occurred on Sand Island during the 2016–
2017 breeding season and the actions taken to abate the
threat they imposed on the albatross population there:
specifically, a broadcast application of AGRID3 in the areas
in which we observed mouse predation on albatrosses.
We also describe the monitoring that was undertaken
to measure both the direct effects of the rodenticide on
the mouse population and the indirect effects that this
treatment had on reducing albatross death, injury, and nest
abandonment.
MATERIALS AND METHODS
Study area
MANWR is located at the north-west end of the
Hawaiian Islands archipelago, 1,930 km from Honolulu,
Hawaii at 28.208° N; -177.379° W. One of the oldest
atoll formations in the world, MANWR consists of three
islands within an 8 km diameter fringing reef. MANWR is
classified as a tropical wet/dry savannah with an average
annual rainfall of 1,104 mm (43.5 in). MANWR has had
a relatively continuous human presence since 1904 when
a station was built to support the construction of a transPacific telegraph cable. From 1941 until 1997, Midway
Atoll was used by the United States Military during which
time both black rats and house mice were introduced. As
a consequence, the atoll’s ecosystems are highly altered.
In 2015 there were 190 species of plants observed, 24
(13%) native and 166 (87%) non-native (Starr & Starr,
2015). The largest, and only, mouse infested island is Sand
at 460 ha. MANWR currently supports a resident human
community of 50 people along with an operational runway,
Henderson Airfield. In 1988 the natural habitats of Midway
Atoll began to be managed as part of the National Wildlife
Refuge system. Its conservation importance is reflected in
its designation as a UNESCO World Heritage Site and its
inclusion within the Papahānaumokuākea Marine National
Monument.
Baiting methods
During 2016–2017, AGRID3 (Bell Laboratories,
Madison, WI), a cholecalciferol rodenticide, was handbroadcast in all affected areas to reduce mouse populations
more effectively and with less disturbance to other wildlife
species compared to trapping. AGRID3 pellets contain
0.075% cholecalciferol (non-anticoagulant), which acts by
disrupting calcium (Ca) homeostasis through increasing
Ca absorption from the small intestine, mobilisation of Ca
from the bones into the blood stream, and decreasing Ca
excretion by the kidneys (Marshall, 1984). Cholecalciferol
has been proven to be toxic and effective at controlling
rodents, yet relatively safe to non-target species when used
according to label specifications. Due to cholecalciferol’s
unique mode of action, target specificity, no taste aversion,
and delayed toxic effect, it has been successfully used in
commensal and agriculture field rodent control situations
(Hix, et al., 2012). These attributes make it ideal for use as
an interim control measure in the event that eradication is
subsequently preferred and approved. The registered use of
AGRID3 in the United States has only been for agriculture
purposes in the past. The USFWS collaborated with Bell
Laboratories, Inc. to develop a supplemental label to be
attached to AGRID3 Pelleted Bait (EPA REG. NO. 1245522
117-3240). This supplemental label specifically for use by
USFWS to control house mice on MANWR was approved
by the Environmental Protection Agency for use in a
wildland setting.
We hand-broadcast AGRID3 pelleted bait along a 5 m
grid (one application within each 25 m2 square grid cell)
over every mouse attack area on Sand Island, as well as a
10 m buffer zone on the periphery of the area, on December
17–18, 2016. Previous experimental bait uptake trials
using the protocol described in Pott, et al., (2015), in which
we applied placebo bait at 40 kg/ha, marked pellets, and
measured pellets taken over a four-day period, led to the
selection of 20 kg/ha as an effective application rate under
average conditions and 35 kg/ha when mouse density
was very high. Following bait application, we surveyed
treatment areas to document any sick or injured non-target
species or instances of non-target species foraging on bait
pellets. We repeated the application at the same rate of 20
kg/ha on 20 January 2017. Over the course of the season
from 17 December 2016 to 20 January 2017 we applied
721 kg of AGRID3 to the treatment areas. Areas receiving
only a single application included the control plot and
impact areas identified after the December application
such as Plots 4 and 5. Each application took approximately
440 person-hours to complete.
Non-targets
In order to reduce house mouse predation on incubating
albatrosses while minimising the effects (mortality and
disturbance) to non-target species, including Laysan ducks
and migratory shorebirds, managers treated albatross attack
areas where dead adults or abandoned nests were found on
Sand Island, MANWR, with AGRID3. AGRID3 was chosen
specifically because of its minimal potential effects on nontarget species, specifically endangered Laysan ducks (Anas
laysanensis; listed under the United States Endangered
Species Act of 1973) and shorebirds which are protected
under the Migratory Bird Treaty Act, particularly bristlethighed curlews (Numenius tahitiensis), Pacific golden
plovers (Pluvialis fulva), and ruddy turnstones (Arenaria
interpres). These species were present in large numbers on
Sand Island, MANWR, during the mouse attacks and are
known to have ingested rodenticide pellets or insects that
have consumed bait at other sites where rodent eradication
has been implemented. Eason, et al. (2000) documented
that mallard ducks fed cholecalciferol at a rate of 2,000
milligrams/kilogram were not affected and concluded that
ducks would have to consume 2,000 g (4.4 lbs) of bait with
this concentration to receive a lethal dose. Smaller Laysan
ducks may consume some bait; however, it is unlikely the
ducks would consume enough to cause injury and would
need to ingest more than twice their body weight in pellets
to experience lethal effects.
Study design and monitoring methods
Starting in December 2016, when most albatrosses had
laid their eggs, observers trained to detect mouse-injured
albatrosses again searched for, documented, and mapped
birds showing signs of mouse attack as well as areas that
had an unusually high occurrence of abandoned eggs in
nest cups across Sand Island. To avoid double counting,
they marked dead adult albatrosses. Nests belonging to
injured birds (typically bite wounds, sometimes resulting
in severe infection) and abandoned eggs were also marked
every three days in the intensive monitoring area (Plot
1). Once the majority of mouse attack areas had been
identified, three baiting plots (Plots 1 [16,493 m2], 2
[15,119 m2], and 3 [11,740 m2]) and a control [6,031 m2]
that was not treated with rodenticide were established and
monitored for changes in mouse abundance in all plots
prior to rodenticide applications on 17 December 2016
Duhr, et al.: Mice preying on albatrosses
and for two weeks afterwards. Two additional baiting
plots (Plots 4 [1,900 m2] and 5 [4,725 m2]) were added
later and monitored for dead adults and abandoned eggs
one day before and once at six days and once at 10 days
after a second bait application that began on 18 January
2017. The plots were all of different sizes because we
chose entire discrete areas in which dead albatrosses and
abandoned nests were found to label as attack areas. The
control area was smaller than the treatment plots because
our priority was to implement a management action as
quickly as possible in as much of the colony as possible.
General surveillance for signs of mouse attacks continued
after hatching in early February and throughout the rest of
the chick rearing period.
All nests in Plot 1 were monitored to determine
reproductive success, defined as number of nests with
an incubating adult present at the beginning of February
divided by the total number of nests with eggs present at
the start of the study. The reproductive success in Plot 1 was
compared to data from plots unaffected by mouse predation
that were part of a long-term albatross demography project
being conducted at MANWR for the same time interval.
We measured mouse relative abundance in all five plots
and the control area two days before rodenticide treatment,
one day a week later and one day two weeks after
application. We used six multi-catch mouse traps (Trapper
24/7 Bell Laboratories) per treatment area, baited with
peanut butter, and summed the number of mice captured
over one night for each plot. The traps were centred within
the plot ca.10 m from each other. To detect any change in
number of mice at each plot, we conducted a one-tailed,
paired t-test comparing the mean number of captures prior
to bait application with the number of captures two weeks
post-treatment (α = 0.05). In addition, for Plots 1, 2, 3 and
the control area we walked a 150 m transect and counted all
mice seen within 2.5 m of the path on either side between
7:30 and 10:00 p.m. the night immediately before the bait
application and then one night one week after broadcast
and one night two weeks after the broadcast.
We used weather data measured daily at Henderson
Airfield weather station located on Sand Island and
available from the U.S. National Climate Data Center,
<(https://www7.ncdc.noaa.gov/CDO/cdopoemain.cmd?d
atasetabbv=DS3505&countryabbv=&georegionabbv=&r
esolution=40>) to evaluate fluctuations in mouse relative
abundance over time in the context of rainfall and aid in
our interpretation of results.
Fig. 3 Areas treated with AGRID3 at a rate of 20 kg/ha
December 2016 through January 2017.
RESULTS
Over the course of the 2015–2016 breeding season, mice
killed at least 42 adult albatross, wounded an additional 480
birds, and resulted in 70 abandoned nests in three distinct
areas, totalling 1.65 ha of Sand Island. During the breeding
season of 2016–17 mouse predation was first observed
on 4 December, 2016. Numbers of injured and dead adult
albatross and abandoned nests increased dramatically
in comparison with the previous breeding season. The
number of affected areas in the colony increased from three
to 50 and the total affected area increased from 1.65 ha to
11 ha (Fig. 2). Albatrosses nest on all of the 460 ha of Sand
Island except where they are excluded by active runway
paving or structures, so the area affected is still a relatively
small proportion of all the albatrosses at Midway. All areas
where albatross mortality was detected in 2015–2016 also
had mouse predation in 2016–2017. By mid-February
there were 242 dead adults, 1218 injured birds, and 994
abandoned nests. This represented a 7-fold increase in
mortality, more than double the rate of injury and a more
than 10-fold rate of nest abandonment compared to the
previous year. The majority of birds found injured and dead
were Laysan albatrosses; few black-footed albatrosses were
affected. Six carcasses recovered fresh from the area were
sent to the USGS Wildlife Health Laboratory in Honolulu
in January 2016. Analysis of the specimens revealed that
the birds were in excellent body condition with no cause of
death evident other than the large wounds on their necks,
backs or flanks. Study of the wound sites confirmed the
rodent bites occurred before death.
There were no confirmed instances of mouse predation
after February 6, 2017, about the time that most eggs started
hatching. Most identified mouse attack areas were baited
twice before predation stopped in February (Fig. 3). There
were no observations of any non-target organism such as
shorebirds or Laysan ducks interacting with bait pellets in
the field or being found sick or dead in the baited areas.
Fig. 2 Areas in which mouse attacks (dead adults,
wounded adults, abandoned eggs) were detected in the
2016–2017 albatross breeding season.
The number of newly deceased adults and abandoned
nests diminished after both bait applications in Plot 1 (Fig.
4) where we were able to conduct more intensive mortality
and nest abandonment monitoring every three days. In
contrast, during December, the number of abandoned
eggs more than doubled in the control area from 10 to 23
but no dead adults were recorded in that area. In January,
Plot 4 showed a decrease in the number of abandoned
nests after the AGRID3 application but Plot 5 continued
to have relatively steady counts of newly dead adults and
abandoned nests (Fig. 5). Reproductive success (number of
eggs in early February / number of eggs in mid-December)
23
Island invasives: scaling up to meet the challenge. Ch 1A Rodents: Planning
Fig. 4 Absolute counts of new detections of abandoned
eggs and dead adults surveyed approximately every
three days in Plot 1 throughout the incubation period of
breeding Laysan albatrosses at Midway.
Fig. 7 Total mice counted at night on a 150 m transect
(2.5 m to each side) prior to and one and two weeks
after application of cholecalciferol bait in Plots 1–3 and a
control site on Sand Island during December 2016.
after treatment (Plot 4 t(5) = -0.99 P = 0.18; Plot 5 t(5) =
-1.66 P = 0.08). Mouse detections on the 150 m transect in
Plots 1, 2, and 3 showed a decline after the application of
AGRID3 while detections remained much the same in the
control plot (Fig. 7).
DISCUSSION
Fig. 5 Count of new detections of abandoned eggs and
dead adults in Plots 4 and 5 immediately before and one
week and two weeks after baiting.
The exposure of a non-negligible proportion of the
world’s Laysan and black-footed albatrosses to a threat
of adult mortality stimulated the management team at
MANWR to seek short-term and long-term solutions.
The application of a pelleted cholecalciferol rodenticide,
AGRID3, in a wildland setting at MANWR, where
many non-target species are present, shows promise as
a management tool to limit house mouse predation on
breeding seabirds without causing harm to the non-target
shorebird and duck species that inhabit Sand Island.
AGRID3 measurably reduced mouse predation on nesting
albatross in the areas where injured and dead albatrosses
and abandoned nests were being detected.
While this study was limited in scope and sample size
due to the prioritisation of rodent management for the
purposes of protecting nesting albatrosses, the larger plots
studied during the December application of rodenticide
showed decreases in the attacks by mice on albatross as
well as some reduction in mouse abundance. The results
from the January trial were less promising, showing
an increase in mouse density and ambiguous effects on
albatross mortality and nest abandonment counts.
Fig. 6 Number of mice trapped in 6 multi-catch traps per
plot, one day before and one and two weeks subsequent
to applications of cholecalciferol rodent bait in Plots 1–5,
and a control site at Sand Island during December 2016
and January 2017. Only 1 application was done in areas
4 and 5.
in Plot 1 was six percent lower than in the unaffected longterm demography plots.
After the December rodenticide application, the number
of mice trapped in Plots 1, 2, and 3 dropped (Fig. 6) (Plot 1
t(5) = 2.46 P = 0.03; Plot 2 t(5) = 0.8 P = 0.23; Plot 3 t(5) =
2.18 P = 0.04). Over the same time period the control site
showed an increase in mice trapped (t(5) = -2.63 P = 0.02).
Trapping in Plots 4 and 5, done a month later in January,
showed a different pattern with mouse numbers increasing
24
There were two differences between the December and
January trials that might explain the contrasting outcomes.
First, the plots baited in December were much larger in
area than the plots baited in January. In a food-limited
environment, mice may have been attracted by the bait
into the smaller plots elevating the mouse density thus
offsetting mortality and mouse population reduction. In an
experimental application of cholecalciferol over a much
larger area of 100 ha in New Zealand Hix, et al. (2012)
observed a 100% reduction of mouse numbers. Second,
rainfall increased dramatically over the two months of
the study. The increase in rain between December and
January might have increased the amount of natural rodent
foods within the study area while also leading to higher
rates of pellet degradation due to the moister conditions,
thus reducing bait availability. There was no control plot
established in the January trials so changes in mouse
behaviour or abundance cannot be evaluated but Plot 1
continued to show a decrease in mouse attacks throughout
the January trial period leading to the possible conclusion
that the results in Plots 4 and 5 were due to the smaller plot
Duhr, et al.: Mice preying on albatrosses
size. During future efforts to control rodent populations
in targeted areas using a broadcast of rodenticide, control
areas of at least 1.2 ha should be considered to ensure
sufficient coverage to compensate for edge effects.
REFERENCES
The decision to apply AGRID3 prior to the albatross
breeding season in any particular year may be informed by
the likelihood that conditions will trigger mouse predation.
Hypotheses about the conditions on Sand Island that may
have triggered the emergence of house mouse attacks
include population fluctuations of mice and a shift in mouse
behaviour due to habitat changes and food availability.
Golden crownbeard, Verbesina encelioides, an introduced
sunflower-relative, was once dominant across the island
with coverage now reduced to less than one percent due to
control measures ongoing since 2011. We have no evidence
that Verbesina is consumed by mice, and it is considered a
poisonous plant to ungulates (Keeler, et al., 1992) and is
allelopathic, thus inhibiting all other vegetation (Inderjit,
et al., 2000). Verbesina distribution and density was much
reduced several years before mice were documented killing
albatrosses at Midway. Changes in seasonality of rainfall
patterns observed during the 2015–16 and 2016–17 El Niño
event may have shifted the timing of normal population
fluctuations in the mouse population of Sand Island, in
which drying conditions reduce forage and subsequently
cause mass-starvation. In 2015–16 and 2016–17 this
crash occurred just as albatrosses began the vulnerable
incubation period when the adult birds are reluctant to leave
their eggs. Rodent populations are well known to fluctuate
with rainfall (Jaksic, et al., 1997) and climate change may
increase the frequency of El Niño–Southern Oscillation
events (Timmermann, et al., 1999), exacerbating the risk
to albatrosses in the future. The question of whether there
was cultural transmission of albatross predation behaviour
in the mice at Sand Island remains open. During 2016–17
the behaviour arose almost simultaneously over much of
the island so it seems unlikely.
Cuthbert, R. and Hilton, G. (2004). ‘Introduced house mice Mus musculus:
A significant predator of threatened and endemic birds on Gough Island,
South Atlantic Ocean?’ Biological Conservation 117: 483–489.
Preparations for a proposed mouse eradication attempt
at Sand Island, MANWR, are underway and the proposed
toxicant is brodifacoum. AGRID3, being a cholecalciferolbased rodenticide may be advantageous for control
operations prior to a possible eradication to reduce the
chance of mice developing aversion or resistance to the
type of bait products and toxicants that might be used in an
actual eradication operation.
Timmermann, A., Oberhuber, J., Bacher, A., Esch, M., Latif, M. and
Roeckner, E. (1999). ‘Increased El Niño frequency in a climate model
forced by future greenhouse warming.’ Nature 398: 694–697.
Angel, A., Wanless, R.M. and Cooper, J. (2009). ‘Review of impacts of
the introduced house mouse on islands in the Southern Ocean: Are mice
equivalent to rats?’ Biological Invasions 11: 1743–1754.
Eason, C.T., Wickstrom, M., Henderson, R., Milne, L. and Arthur, D.
(2000). ‘Non-target and secondary poisoning risks associated with
cholecalciferol’. New Zealand Plant Protection 53: 299–304.
Hix, S., Aylett, P. Shapiro, L., MacMorran, D., Eason, C.T. Sam, S., Ross,
J.G., Miller, A., Ogilvie, S.C. (2012). ‘Low-dose cholecalciferol bait
for possum and rodent control.’ New Zealand Journal of Agricultural
Research 55: 207–215.
Inderjit, Asakawa, C. and Dakshini K.M.M. (2000). ‘Allelopathic
potential of Verbesina encelioides root leachate in soil.’ Canadian
Journal of Botany 77(10): 1419–1424.
Jaksic, F.M., Silva, S.I., Meserve, P.L. and Gutierrez, J.R. (1997). ‘A
long-term study of vertebrate predator responses to an El Niño (ENSO)
disturbance in western South America’. Oikos 78: 341–354.
Jones, M.G.W. and Ryan, P.G. (2009). ‘Evidence of mouse attacks on
albatross chicks on sub-Antarctic Marion Island’. Antarctic Science 22:
39–42.
Keeler, R.F., Baker, D.C., Panter, K.E. 1992. ‘Concentration of galegine
in Verbesina encelioides and Galega oficinalis and the toxic and
pathologic effects induced by the plants.’ Journal of Environmental
Pathology, Toxicology and Oncology 11(2): 11–17
Marshall, E.F. (1984). ‘Cholecalciferol: A Unique Toxicant for Rodent
Control’. In: D.O. Clark (ed.) Proceedings of the Eleventh Vertebrate
Pest Conference, pp. 95–98. Davis, CA: University of California.
Pott, M., Wegmann A.S. Griffiths, R., Samaniego-Herrera, A., Cuthbert,
R., Brooke, M. de L., Pitt, W.C., Berentsen, A.R., Holmes, N.D.,
Howald, G.R., Ramos-Rendón, K. and Russell, J.C. (2015). ‘Improving
the odds: Assessing bait availability before rodent eradications to aid
in selecting bait application rates.’ Biological Conservation 185: DOI:
10.1016/j.biocon.2014.09.049.
Starr, F. and Starr, K. 2015. Botanical Survey of Midway Atoll. Unpublished
Report. Prepared for United States Fish and Wildlife Service. Midway
Atoll National Wildlife Refuge.
ACKNOWLEDGEMENTS
We are grateful to the managers, staff, and volunteers
of Papahānaumokuākea Marine National Monument and
Midway Atoll National Wildlife Refuge including Matt
Brown, Robert Peyton, Dan Clark, Carmen Antaky, Nasser
Hasan, Taylor Smith, Mike Marxen, Katrina Scheiner,
Stephanie Kolar, Ann Humphrey, David Dow, Kristina
McOmber, Wieteke Holthuijzen, Preecha Songserm, Surat,
Surasak Fakkaew, Chatpong Upara, Ek Anan in Uthen,
Keely Hassett, Kiah Walker, Aisha Rickli-Rahman, Eric
Baker, Beth Wolff, Michelle Smith, Naira DeGarcia, and
Victoria Taylor for providing support and assistance at
every step. Thanks to Craig Riekena and Amy Breunig at
Bell Labs for generous donation of supplies and equipment
to allow us to respond on very short notice, USDA-APHIS
Wildlife Services for excellent advice, The National Fish
and Wildlife Foundation (NFWF) for generous support, US
Environmental Protection Agency for rigorous and helpful
counsel and Thierry Work of the USGS National Wildlife
Health Laboratory for quick diagnoses. The careful reading
and fine suggestions from two anonymous reviewers were
much appreciated.
25
G.A. Harper, P. Carr and H. Pitman
Harper, G.A.; P. Carr and H. Pitman. Eradicating black rats from the Chagos – working towards the whole archipelago
Eradicating black rats from the Chagos – working towards
the whole archipelago
G.A. Harper1, P. Carr2 and H. Pitman2
Biodiversity Restoration Specialists, P.O. Box 65, Murchison 7053, New Zealand. <biodivrestoration@gmail.com>.
2
Chagos Conservation Trust, 23 The Avenue, Sandy, Bedfordshire, UK SG19 1ER.
1
Abstract The Chagos Archipelago comprises some 58 islands covering 5,000 ha in the centre of the Indian Ocean. Black
rats (Rattus rattus) were introduced about 230 years ago and have likely had a severe impact on the native terrestrial
fauna, which is dominated by seabirds and land crabs. Most of the archipelago’s terrestrial land mass is vegetated with
old coconut plantations, with over 75% of the native forest cleared for coconut from 26 of the largest islands. Likely as a
result of this colonisation and clearance, at least 30 islands have rats present (95.3% of the Chagos landmass) along with
feral cats (Felis catus) on 62%, which suppresses the recovery of native fauna and flora. Efforts at rat eradication include
the failed attempt on Eagle Island (252 ha) in the northern Chagos Archipelago in 2006 and the recent success of a groundbased eradication on Île Vache Marine in 2014, where two applications of brodifacoum poison were hand-spread at a rate
of 18 kg/ha. Two islets on the nearby Salomon atoll were also cleared of black rats during the same operation with single
bait applications. The 2014 operation was successful on what are regarded as difficult islands for rat eradication, being
‘wet’ tropical islands with land crabs and coconut plantations present, and has engendered confidence to proceed with
additional rat eradications on other northern Chagos islands.
Keywords: atoll, Birgus, Chagos, eradication, hand-broadcast, Rattus, seabirds, tropical
INTRODUCTION
Invasive species have caused 75% of terrestrial
vertebrate extinctions on islands (McCreless, et al., 2016)
and of these species’ rats are probably the most pervasive,
having been introduced to more than 85% of oceanic
islands and archipelagos (Harper & Bunbury, 2015). Rats
have been responsible for some 40–60% of all bird and
reptile extinctions (Howald, et al., 2007). Rats prey upon
and compete with animals and can consume all parts of
plants, which disrupts ecosystem function and can cause
direct or indirect cascades of collapse, through interruption
of pollination and nutrient pathways, seed predation, and in
some cases leading to forest collapse (Towns, et al., 2006;
Athens, 2009; Towns, 2009; Hilton & Cuthbert, 2010).
Black rats (Rattus rattus) have been present on the
Chagos Archipelago, in the mid-Indian Ocean, since the
late 1700s when the archipelago was settled (WenbanSmith & Carter, 2016). Diego Garcia is in the southern
Chagos Archipelago and is the largest (~2,900 ha) and only
inhabited island, with a transient population associated with
a military base. It has rats and cats (Felis catus) present
and there are no current plans for rat eradication. In the
northern Chagos Archipelago (~2,100 ha total combined
area), 26 of the 55 islands are known or suspected to
have black rats present (Carr & Harper, 2015). These ratinfested islands comprise some 1,700 ha in combined total
land area or some 47% of the islands in the group. Only
4.7% of the entire Chagos terrestrial space is regarded as
mammalian predator free and seabird population density is
approximately 20 times higher on rat-free islands (Hilton
& Cuthbert, 2010). (Fig. 1).
Low-lying, remote and geologically young (49 mMYA,
Duncan & Hargraves, 1990), the Chagos Archipelago has
not had the speciation that has developed on similarly
isolated elevated archipelagos such as Hawaii and the
Mascarene Islands. The atolls of the Chagos Archipelago
are largely formed from marine sand deposits with some
raised rock formations. Many islands had their native
forest removed during settlement and replaced with a
dense monoculture of coconut palms (Cocos nucifera). As
several seabird species preferentially nest in native trees,
this destruction of nesting habitat was probably the first
major impact on the previously large seabird colonies
that existed there (Carr, 2013). This was followed by
Fig. 1 Map of the Chagos Archipelago showing location of
islands mentioned in the text.
direct persecution by man and other introduced predators;
principally this was rats, cats, dogs and pigs. By the start
of the 1900s, the vast seabird colonies now indicated by
guano deposits had disappeared (Carr, 2011).
In:
26C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 26–30. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Harper, et al.: Eradicating rats from Chagos
On less anthropogenic-impacted islands the architecture
of the native oceanic rain forest allows arboreal nesting by
lesser noddy (Anous tenuirostris) and red-footed booby
(Sula sula), whereas the open areas are used by species
such as brown booby (Sula leucogaster), brown noddy
(Anous stolidus), sooty tern (Onychoprion fuscatus)
and the tropical shearwater (Puffinus bailloni). Two
introduced birds, the domestic chicken (Gallus gallus),
and Madagascan fody (Foudia madagascariensis) are
the only land birds resident in the northern Chagos, the
former found on only a few islands. Fodies are found on
most of the vegetated islands. Land crabs are the dominant
invertebrates, with the coconut crab (Birgus latro) being
the most obvious. Smaller hermit crab species, the
burrowing land crab (Cardisoma carniflex) and other land
crab species are present (Stoddart, 1971a; PC pers. obs.).
There is a reviving population of green turtles (Chelonia
mydas) and hawksbill turtles (Eretmochelys imbricata) that
nest on some islands (Mortimer & Day, 2009). No native
mammals, including bats, exist on the islands.
Previous rat eradication attempts in the Chagos
Rat eradication planning
Rat eradications on tropical islands have a higher
failure rate than temperate islands for a variety of reasons,
including the presence of coconuts, land crabs as bait
competitors, and on ‘wet’ tropical islands, in particular
(Russell & Holmes, 2015; Holmes, et al., 2015). Groundbased operations also have had a higher failure rate than
aerial bait applications but are usually cheaper to undertake
on small islands, with less logistical and technical input
required. Hence, for the rat eradications on the northern
Chagos a ground-based eradication was planned for cost
and logistical reasons but needed to be very cognisant of
the risk factors associated with the islands. A successful
outcome for the eradication operation in the face of these
impediments would promote confidence in the technique
for tropical islands with similar characteristics.
The Chagos Conservation Trust is championing the
eradication of rats from the archipelago, to provide an
environment for populations of existing native species
to recover and to restore the ecosystem to a state prior to
rat invasion (<https://chagos-trust.org/about/vision-andmission>). This endeavour is in concert with The British
Indian Ocean Territory Interim Conservation Management
Framework of 2014 (<https://biot.gov.io/biot-interimconservation-management-framework-september-2014.
pdf>), which has an overarching vision: “To maintain and,
where possible, enhance the biodiversity and ecological
integrity of the British Indian Ocean Territory (BIOT)”.
All eradication attempts require comprehensive
planning before implementation and this was particularly
true for a rat eradication programme on a highly isolated
tropical island, which presented a novel suite of problems.
Invasive mammal eradication work in the Chagos
Archipelago faces both logistical and ecological challenges
due to the archipelago’s remoteness and inaccessibility,
along with the wet climate and the vegetation composition
with the significant component of coconut ‘chaos’. Île
Vache Marine was selected for the initial rat eradication
because it was: deemed a realistic and manageable size for
a start-up operation; the risk of reinvasion was considered
negligible due to its distance from other islands; there were
no susceptible non-target species; it nestled in amongst
five confirmed or proposed IUCN classified Important Bird
Areas (Carr, 2011); there was some anecdotal evidence that
shearwaters had once bred on the island and, if successful,
the probability of re-colonisation by marine avifauna was
likely.
Île Vache Marine (12.4 ha, 2 m elevation) is situated
on the southern rim of the Peros Banhos atoll (05°25› S,
71°49› E, Fig. 1). It is a typical tropical low-lying oceanic
coralline island and the vegetation comprises a shoreline
perimeter of Scaevola taccada on the exposed southern
coast with introduced coconut and the occasional Guettarda
speciosa and Morinda citrifolia on the coast facing the
atoll. The mean annual rainfall for the Peros Banhos
atoll (data from 1950–1966) is approximately 4,000 mm,
distributed bi-modally, with a slightly drier period through
the austral winter (Stoddart, 1971b). Île Vache Marine was
never inhabited, but was visited until 1974 for coconut
harvesting. The plantation workers would have come
from Île du Coin (126 ha), some 6 km distant, the former
plantation headquarters and likely source of rats. In 2014
there were very limited numbers of seabirds present.
There was an attempt to eradicate black rats from Eagle
Island (252 ha) in 2006. A team of 11 established 2,864
bait stations on a 30 m × 30 m grid of cut tracks starting
in early February. The bait stations were loaded with
Talon™ wax blocks (0.05 g/kg brodifacoum with bitrex)
that was maintained in the stations until the team departed
in late April (Meier, 2006). Later checks revealed that the
operation had failed.
The Île Vache Marine eradication served two purposes.
Firstly, it was an opportunity to undertake a rat eradication
operation, albeit small, as proof that the method could be
successful in the northern Chagos islands and engender
confidence in the technique as a management tool for
biodiversity gains in the region. Secondly, it added Île
Vache Marine to a string of rat-free islands in eastern
Peros Banhos, which were situated amongst Important
Bird Areas and within an area designated as a Strict Nature
Reserve under BIOT Law.
As an adjunct to the planned eradications, an additional
bait trial was carried out on Diego Garcia in order to
measure bait-take by rats at a measured rat population
density and refine bait application rates for future rat
eradications on Chagos atolls (Harper & Carr, 2015).
METHODS
Île Vache Marine rat eradication, Peros Banhos atoll
Parallel lines were cut at 25 m intervals across the
island in June 2014. This was undertaken by volunteers
from the British Forces stationed on Diego Garcia. The
timing was important, in that it needed to be done long
enough before the operation so that any disturbance did
not affect rat behaviour but not too early as re-growth was
rapid, particularly in S. taccada thickets.
August was chosen as the month for bait application
as it was one of the driest months of the year (Stoddart,
1971b) and when a vessel was available. The eradication
operation staff assembled the equipment and supplies
in Diego Garcia on 31 July 2014 prior to departure on 1
August. The team, GH, PC and members of the British
Forces on Diego Garcia, landed on Île Vache Marine early
on 2 August to allow passage over the coral reef at high
tide. The cut lines were checked and, where required, were
either re-cut or additional lines slotted in between existing
lines. Sites for bait throwing were marked at 25 m along
the cut lines and black plastic bait stations (Protecta LP,
Bell Labs, USA) laid at these sites for post-broadcast bait
deployment. The bait stations were raised 40 mm off the
ground with wooden blocks to reduce interference by
hermit crabs. By the end of the first day there was a 25 m
× 25 m grid of 154 sites across the entire island. The island
27
Island invasives: scaling up to meet the challenge. Ch 1A Rodents: Planning
size was also reconfirmed at 12 ha by walking the coast of
the island with a GPS unit (Garmin 62S).
Bait application trials and eradications on other Indian
Ocean islands showed that bait could be spread at a rate
of >15 kg/ha and be available to all rats for four nights
(Merton, et al., 2002, Harper & van Dinther, 2014). Bait
was spread on Île Vache Marine on 3–4 August. Pollard
pellet bait (Bell Labs 25W) was hand spread at a rate of
18 kg/ha by GH and PC. This involved hand-throwing bait
at each of the grid sites. Bait (280 g) was thrown in four
directions at right angles to each other, such that it reached
about 10–12 m, along with 280 g spread at the throwing
point. Bait spreading by the two operators began at each
end of the island and lines were traversed such that the
operators were converging on each other. Bait coverage
was almost completed on the first day, except for a strip
of about 2 ha in the centre of the island. This was covered
the next morning and a little additional bait was spread
above the high tide mark around the coast of the island
where hermit crabs were abundant. All the equipment and
empty poison bait containers were removed by the end of
the morning. The team departed for Diego Garcia shortly
thereafter.
A second bait application was undertaken 11 days later.
This was to ensure all rats had access to bait, particularly
if breeding was occurring and suckling mothers or young
animals may have been missed in the first bait application.
The island was revisited on 14–15 August and poison bait
pellets (Pestoff 20R) were hand laid at a rate of 18 kg/ha.
Differing bait types were used for the two applications
as ship rats have been observed with distinct preferences
for one or other bait, thus circumventing possible bait
avoidance (Harper & van Dinther, 2014). Several recently
dead rats were located during the second bait application,
suggesting rats had readily consumed the poison bait laid
in the first application. The bait stations were also then
loaded with wax-based poison baits (Ditrac™ 0.005%
w/w brodifacoum, Bell Labs) at a rate of three bait blocks
(150 g) secured inside each station. This was to ensure
that if heavy rain degraded the bait post-departure, or any
rats missed the hand-laid bait, then poison bait was still
available for several weeks after the operation. The team
departed Île Vache Marine on 15 August at midday. It did
not rain during either of the bait deployments.
Bird counts on Île Vache Marine had been undertaken
by PC since 2009. Counts in 2014 revealed fewer than
five pairs of brown noddy and white tern (Gygis alba)
and ten pairs of great crested terns (Thalasseus bergii),
were breeding on the island. About 15 pairs of the one
introduced passerine, the Madagascar fody (Foudia
madagascariensis) were present.
In April 2015, the bait stations were removed by PC
and a Connect Chagos graduate (a Zoological Society/CCT
project with funding from the UK FCO) during a different
expedition.
The eradication phase of Îles du Sel and Jacobin,
Salomon Islands atoll
Additional poison bait intended as a contingency for
the Île Vache Marine operation was deployed on two
islets, Îles du Sel (2.2 ha) and Île Jacobin (1.6 ha), in the
Salomon atoll, some 25 km east of the Peros Banhos.
These two islands were selected for their small size and
their relative isolation from other islands. This meant there
was a lower probability of re-invasion by rats than other
islands in the area and a single application of the remaining
bait was deemed practical. On arrival, a quick survey was
carried out immediately before each operation to assess
the likelihood of success. Both islands were dominated
by coconut, with varying amounts of native forest present,
with few other factors that would limit the probability of
success, as identified by Holmes, et al. (2015). Of note was
the lack of large seeds or seedlings of native trees.
Bait was deployed on Île du Sel and Île Jacobin on
16 August 2014. The islands were circumnavigated and
waypoints marked at 25 m intervals on each side of the
islands using a GPS unit (Garmin 62S). The operators (GH
& PC) then walked from the first waypoint on one side to
the corresponding waypoint on the opposite side of each
island without cutting the vegetation. Pellet bait (Pestoff
20R) was broadcast at 25 m intervals, using the same
method as on Île Vache Marine. Bait was spread at a higher
rate of 20 kg/ha on Île du Sel and 25 kg/ha on Île Jacobin
as it was a single application. The difference in application
rate was due to slightly more bait remaining after the first
island was treated.
There were opportunities for post-eradication
monitoring on Île Vache Marine by PC as part of other
expeditions. The first check was made seven months later
in April 2015 with a Connect Chagos graduate during
daylight, and during an overnight stay, and no sign of
rats was seen. A second daytime check was made by PC
in February 2016 and again no rat sign was recorded but
signs of vegetative recovery were noted (Table 1). An
opportunity for both GH and PC to undertake a more
comprehensive survey of the island became available over
9–10 April 2017, when 45 rat snap-traps and wax tags were
deployed over a 24-hour period. In addition, coconuts were
cut open and placed on the ground near the campsite and
searches were made for gnawed seeds/coconuts, rat tracks/
caches etc. Additional searches were conducted at night by
torchlight to detect rats.
During the same expedition, surveys were made at Île
du Sel and Île Jacobin on 15 April and 15 rat detection
devices (snap traps, wax tags, secured portions of coconut
flesh) were deployed overnight on each island. Searches
Table 1 Initial checks of Île Vache Marine for rat sign.
Date
24–
25/03/2015
Event
Rodent survey including:
a. 50 × snap-traps deployed overnight
b. Check for rat gnawing on fallen fruit and flowers
c. Check for rat excrement
d. Daytime visual inspection of island
e. Nocturnal inspection of island (overnight stay)
Results
No sign of rat presence
Responsible
P Carr
C Narina
J Schlayer
09/02/2016
Rodent survey including:
a. Check for rat gnawing on fallen fruit and flowers
b. Check for rat excrement
c. Daytime visual inspection of island
No sign of rat presence.
Obvious signs of native tree
seed germination especially
Guittarda speciosa and
extensive flowering of
Scaevola taccada
P Carr
28
Harper, et al.: Eradicating rats from Chagos
were made for signs of rats similar to the operation on Île
Vache Marine. The islands were revisited the following
morning and detection devices recovered and further
searches made.
Diego Garcia bait trial
Two 1 ha plots were set out 200 m apart in disused
coconut plantation forest some 2 km west of the small
township on western Diego Garcia. The plots were
divided into a 5 × 5 grid at 25 m intervals. Within the
plots an internal trapping grid of 15 Victor snap-traps was
established on an interval of 25 m × 12.5 m. The internal
grid was centrally located so that there was a 25 m buffer
from the plot perimeter.
Poison bait (Pestoff 25R pellets, Animal Control
Products, NZ) was hand-spread on both 1 ha plots at a rate
of 15 kg/ha on 7 August 2014. The bait had been dyed with
Rhodamine-B, which fluoresces under UV light. After one
night to allow rats to access the bait the snap-traps were
baited with coconut and peanut butter and set. Trapped
rats were collected morning and evening for the next
three days. The rats were dissected and their gut cavities
examined under UV light for evidence that the dyed bait
had been consumed.
To give a simple estimate of rat population density, the
number of rats caught was divided by the effective trapping
area (ETA). To estimate ETA for rats, a boundary strip was
added to the edge of the trapping grids (Dice, 1938). The
width of the boundary strip was set by adding the average
radius (15 m) of a home range of ship rats from mangrove
forest on Aldabra Atoll and forest on Juan de Nova and
Europa (Harper, et al., 2015, Ringler, et al., 2014).
RESULTS
Rat eradications in northern Chagos Islands
None of the various indicators used to detect rats during
the overnight stay on Île Vache Marine on 9–10 April
2017 showed that rats remained on the island. Prior to the
eradication rats had been seen on every previous inspection
and were easily trapped both diurnally and nocturnally.
Moreover, there had been an increase in breeding pairs
of seabird species for pre- and post-eradication counts,
including a significant increase in numbers of white tern
(T1= -2.32, d.f. = 6, p = 0.03), which are vulnerable to rats,
and great crested terns (T1= -4.73, d.f. = 3, p = 0.009).
Similarly, none of the indicators for detecting rats on
Îles du Sel and Jacobin showed sign of any rats. Many
seeds of the large native tree Intsia bijuga had germinated
and there was a carpet of 300 mm high seedlings on the
forest floor of both islands, along with many untouched
seeds. These large seeds appear to be a favoured food of
rats, as the seeds and seedlings are rarely found on ratinfested islands.
Diego Garcia bait trial
Sixty rats were removed from traps over the three days;
30 from each plot. There was significant interference with,
and removal of, trapped rats by land crabs so this is highly
likely to be a minimum number of rats trapped. Of the 60
rats, 59 (98.3%) had eaten dyed bait. The one rat that had
not consumed bait was an adult female that was trapped
in the first morning after the bait application, so bait had
been available for a little over 36 hours. Some bait was still
present on the last day of trapping.
Of the trapped rats, only two were juveniles (both
female) and there was a slight sex bias towards males
(32:28). Several adult male rats were in poor condition,
whilst some rats were in good condition with substantial
amounts of mesenteric fat. Of the 26 adult female rats
trapped, two were pregnant.
The trapping grids within the bait grids were 25 m in
diameter and adding a 15 m boundary strip gave a total
radius of the ETA of 40 m, for an area of 0.5 ha. At least 30
rats were caught on each trapping grid, which translates to
a minimum population density of 60 rats/ha.
DISCUSSION
Rats were eradicated on three small ‘wet tropical’ islands
in the northern Chagos with two hand-spread application
rates of 18 kg/ha each on the larger Île Vache Marine and
single applications of 20 and 25 kg/ha respectively, on the
smaller islets Îles du Sel and Jacobin.
Of particular interest was the success of the rat
eradications on the very small islets, considering that
only one bait application, albeit at a higher initial rate but
cumulatively less than on Île Vache Marine, was made on
each. Best practice suggests two applications, although it
is generally acknowledged that the second application acts
as an insurance policy against unforeseen confounding
factors, such as heavy rain ruining bait (Keitt, et al., 2015),
and because rats can breed year-round in the wet tropics
(Russell & Holmes, 2015). In this case the small size of
the islands, selection of the driest period of year and well
planned rapid bait deployment by a small team is likely
to have assisted with operational success as the factors
associated with eradication failure on tropical islands were
reduced (Holmes, et al., 2015).
Of crucial importance were the parallel and well-cut
lines cut in the thick vegetation on Île Vache Marine, such
that there were no gaps in bait coverage due to converging
tracks. The bait application took 1.5 days, at a rate of about
5 ha/person/day. The bait applications began at both ends
of the island simultaneously and a gap in bait coverage was
left for one night in both cases, which did not affect the
operational success. It is not known whether the bait was
degraded by any heavy rain in the days immediately after
bait deployment as the team left the area shortly after both
applications. Although there were several land crab species,
including hermit crabs, present on the island, coconut crabs
that can outcompete rats for bait were absent.
It can be concluded that rats can be eradicated from
small Chagos Archipelago islands with a minimum toxic
bait application of 20 kg/ha, and the trials on Diego Garcia
indicate that a 15 kg/ha application rate is too low. This
suggests that on similar small islands at least, single
applications of poison can successfully remove rats and
should be considered in appropriate circumstances. A
single bait application has advantages in reduced logistics,
cost and possible impact on the environment. Where
possible, further bait trials will be undertaken on islands in
the northern Chagos Archipelago, to gain more confidence
with the amount of bait and bait presentation required.
These trials are of particular importance on islands largely
dominated by coconut crabs, and with burrowing crab
species present (Holmes, et al., 2015; GH & PC, pers.
obs.) and where mangrove or Pemphis acidula is present at
periodically flooded sites (Harper, et al., 2015).
This operation has provided evidence that rats can be
eradicated from small wet tropical islands that contain
large populations of land crabs and coconut forest that has
previously been deemed difficult to achieve (Holmes, et
al., 2015). We demonstrate that careful assessment and
planning prior to the operation can result in a successful
outcome (Keitt, et al., 2015). Given the success of groundbased rat eradication operations on the three small islands
in the Chagos Archipelago, an eradication is being planned
29
Island invasives: scaling up to meet the challenge. Ch 1A Rodents: Planning
for larger islands in the near future, such as Île Yéyé (61
ha), which is the only remaining rat-infested island in
the eastern Peros Banhos Strict Nature Reserve. If this is
successful a larger operation to eradicate rats from all of
the northern Chagos Archipelago is likely to be pursued.
McCreless, E.E., Huff, D.D. Croll, D.A., Tershy, B.R., Spatz, D.R.,
Holmes, N.D., Butchart S.H.M. and Wilcox, C. (2016). ‘Past and
estimated future impact of invasive alien mammals on insular threatened
vertebrate populations’. Nature Communications 7: 12488.
ACKNOWLEDGEMENTS
Merton, D.V., Climo, G., Laboudallon, V., Robert, S. and Mander, C.
(2002). ‘Alien mammal eradication and quarantine on inhabited islands
in the Seychelles’. In: C.R. Veitch and M.N. Clout (eds.) Turning the
tide: the eradication of invasive species, pp. 182–198. Occasional Paper
SSC no. 28. IUCN SSC Invasive Species Specialist Group. IUCN,
Gland, Switzerland and Cambridge, UK .
The Chagos Conservation Trust and Dr Grant Harper
would like to sincerely thank the UK Government who
generously funded this project through the Darwin
Initiative. This project would not have been possible
without the help and support of several organisations
and persons including the British Indian Ocean Territory
Administration, London, UK; the BIOT Commissioner’s
Representative and British Forces, Diego Garcia; the Base
Operating Service Contractor (BOSC) on Diego Garcia;
the Captains and crews of BIOT Patrol Vessels - Pacific
Marlin and Grampian Frontier; Royal Botanical Gardens
Kew and Royal Society for the Protection of Birds. ACP
(NZ) generously donated non-toxic bait and rodenticide
and Bell Labs (USA) generously donated rodenticide. Two
anonymous reviewers provided valuable comments which
improved the paper.
REFERENCES
Athens, J.S. (2009). ‘Rattus exulans and the catastrophic disappearance of
Hawai’i’s native lowland forest’. Biological Invasions 11: 1489–1501.
Carr, P. (2011). A guide to the birds of the British Indian Ocean Territory.
Information Press, Oxford, UK.
Carr, P. (2013). ‘Factors Influencing Breeding Island Selection of Redfooted Booby Sula sula (Linn. 1766) in the Chagos Archipelago, Central
Indian Ocean, and the Implications for Future Island Management
Plans’. MSc thesis. UK: Warwick University.
Carr, P. and Harper G.A. (2015). ‘The distribution of ship rat Rattus rattus
in the Chagos Archipelago’. Chagos News 47: 20–31.
Dice, L.R. (1938). ‘Some census methods for mammals’. Journal of
Wildlife Management 2: 119–130.
Duncan, R.A. and Hargraves, R.B. (1990). ‘40AR/39AR Geochronology
of Basement Rocks from the Mascarene Plateau, the Chagos Bank, and
the Maldives’. In: R. A. Duncan, J. Backman, L.C. Peterson, et al. (eds.)
Proceedings of the Ocean Drilling Program, Scientific Results Vol. 115,
pp. 43–51. Texas: College Station.
Harper, G.A. and van Dinther, M. (2014). Bait and Trap Trials for
Invasive Rats and Cats on Aldabra Atoll and Implications for Their
Eradication. Unpublished report for the Seychelles Islands Foundation,
Republic of Seychelles.
Harper, G.A., van Dinther, M. and Bunbury, N. (2015). ‘The response
of black rats (Rattus rattus) to evergreen and seasonally arid habitats:
informing eradication planning on a tropical island’. Biological
Conservation 185: 66–74.
Harper, G.A. and Bunbury, N. (2015). ‘Invasive rats on tropical islands:
Their population biology and impacts on native species’. Global
Ecology and Conservation 3: 607–627.
Harper, G.A. and Carr, P. (2015). ‘An initial trial to determine an effective
rat bait application rate at Diego Garcia, British Indian Ocean Territory’.
Chagos News 47: 17-19.
Hilton, G.M. and Cuthbert, R.J. (2010). ‘The catastrophic impact of
invasive mammalian predators on birds of the UK Overseas Territories:
A review and synthesis’. Ibis 152: 443–458.
Holmes, N. D., Griffiths,R., Pott, M., Alifano, A., Will, D., Wegmann, A.S.
and Russell, J.C. (2015). ‘Factors associated with rodent eradication
failure’. Biological Conservation 185: 8–16.
Howald, G., C. J. Donlan, C.J., Galván, J.P., Russell, J.C., Parkes, J.,
Samaniego, A., Wang, Y., Veitch, D., Genovesi, P., Pascal, M., Saunders,
A. and Tershy, B. (2007). ‘Invasive rodent eradication on islands’.
Conservation Biology 21: 1258–1268.
Keitt, B., Griffiths, R., Boudjelas, S., Broome, K., Cranwell, S., Millett,
J., Pitt, W. and Samaniego-Herrera, A. (2015). ‘Best practice guidelines
for rat eradication on tropical islands’. Biological Conservation 185:
17–26.
30
Meier, G.G. (2006). ‘Rat eradication on Eagle Island’. Chagos News 28:
2–4.
Mortimer, J.A. and Day. M. (2009). ‘Sea Turtle Populations and Habitats
in the Chagos Archipelago’. In: C.R.C. Sheppard and M.R.D. Seaward
(eds.) Ecology of the Chagos Archipelago. Westbury, England: Linnean
Society Occasional Publications 2.
Ringler, D., Russell, J., Jaeger, A., Pinet, P., Bastien, M. and Le Corre, M.
(2014). ‘Invasive rat space use on tropical islands: Implications for bait
broadcast’. Basic Applied Ecology 15: 179–186.
Russell, J.C. and Holmes, N.D. (2015). ‘Tropical island conservation: Rat
eradication for species recovery’. Biological Conservation 185: 1–7.
Stoddart, D.R. (1971a). ‘Terrestrial fauna of Diego Garcia and other
Chagos atolls’. Atoll Research Bulletin 149: 163–170.
Stoddart, D.R. (1971b). ‘Rainfall on Indian Ocean coral islands’. Atoll
Research Bulletin 147: 1–21.
Towns, D.R., Atkinson I.A.E. and C.H. Daugherty, C.H. (2006). ‘Have
the harmful effects of introduced rats on islands been exaggerated?’
Biological Invasions 8: 863–891.
Towns, D.R. (2009). ‘Eradications as reverse invasions: lessons from
Pacific rat (Rattus exulans) removals on New Zealand islands’.
Biological Invasions 11: 1719–1733.
Wenban-Smith, N. and Carter, M. (2016). Chagos: A History; Exploration,
Exploitation, Expulsion. London: Chagos Conservation Trust.
G. Maggs, M.A.C. Nicoll, N. Zuël, D.J. Murrell, J.G. Ewen, C. Ferrière, V. Tatayah, C.G. Jones and K. Norris
Maggs, G.; M.A.C. Nicoll, N. Zuël, D.J. Murrell, J.G. Ewen, C. Ferrière, V. Tatayah, C.G. Jones and K. Norris. Bridging the researchmanagement gap: using knowledge exchange and stakeholder engagement to aid decision-making in invasive rat management
Bridging the research-management gap: using knowledge
exchange and stakeholder engagement to aid
decision-making in invasive rat management
G. Maggs1,2, M.A.C. Nicoll1, N. Zuël3, D.J. Murrell2, J.G. Ewen1, C. Ferrière3, V. Tatayah3, C.G. Jones3,4 and K. Norris1
Institute of Zoology, Zoological Society of London, Regent’s Park, London. <gbmaggs@hotmail.co.uk>. 2Centre for
Biodiversity and Environmental Research, Department of Genetics, Evolution and Environment,University College
London, Gower Street, London. 3Mauritian Wildlife Foundation, Grannum Road, Vacoas, Mauritius.4Durrell Wildlife
Conservation Trust, Les Augrès Manor, La Profonde Rue, Trinity, Jersey, JE3 5BP, Channel Islands.
1
Abstract The world is facing a biodiversity crisis. Nowhere is that more apparent than on oceanic islands where invasive
species are a major threat for island biodiversity. Rats are one of the most detrimental of these and have been the target
of numerous eradication programmes; a well-established conservation tool for island systems. For at-risk native species
inhabiting large, populated islands, where rat eradication is not an option, control of rat populations has been conducted but
this requires continuous management and therefore its long-term viability (and that of the at-risk native species which the
project aims to protect) can be uncertain. Large-scale rat management areas or ‘mainland islands’ have been successfully
developed in New Zealand. However, large-scale management is a long-term investment with huge financial implications
and committing to such an investment can be met with reluctance. This reluctance, and its subsequent hindrance to
decision-making, can be caused by uncertainty relating to species conservation outcomes, and the multiple objectives
of stakeholders. We address the issue of uncertainty and the importance of communication between all stakeholder
parties in relation to the Mauritius olive white-eye (Zosterops chloronothos), a critically endangered passerine endemic
to Mauritius and highly threatened by invasive rats. Specifically, we illustrate how the combination of scientific research
and communication, knowledge exchange, and stakeholder workshops, can address some of the barriers of decisionmaking, helping to bridge the research-management gap, and enable the timely expansion of existing rat management for
the benefit of this highly threatened bird.
Keywords: mainland islands, Mauritius, rat control, uncertainty, Zosterops chloronothos
INTRODUCTION
The world is facing a biodiversity crisis and nowhere
is that more apparent than on oceanic islands where
invasive species are a major threat (Jones & Merton, 2012;
Rodrigues, et al., 2014). Recent research has identified
islands as conservation priority areas for evolutionary
distinct and globally endangered (EDGE) species,
increasing the importance of conservation for island
endemics from areas such as Hawaii, New Zealand, the
Mascarenes and the West Indies where there are high
extinction rates (Diamond, 1989; Jetz, et al., 2014). A
major cause of extinction for island birds has been invasive
species and rats are the most detrimental; having reached
around 90% of all islands they have been identified as a
massive threat to ecosystems (Atkinson, 1985; Towns, et
al., 2006; Blackburn, et al., 2014).
The eradication of invasive rats from islands is a
well-established conservation tool with 474 successful
eradications of Rattus rattus and R.norvegicus (black rat
and brown rat) between 1951 and 2014 (Towns & Broome,
2003; DIISE, 2015). However, for species inhabiting large,
populated islands, where eradication is not an option,
localised rat control has to be conducted. However, this is
not a long-term solution for many species of conservation
interest as the areas of control can be too small to create
viable populations and rat reinvasion rates can be too high.
An alternative are large-scale rat management areas or
‘mainland islands’ which have been successfully developed
in New Zealand (Saunders & Norton, 2001; Butler, et al.,
2014). However, large-scale management is a long-term
investment with huge financial implications and in a world
of limited resources and accountability, committing to
such an investment can be met with reluctance (Cullen, et
al., 2001; Burns, et al., 2012; McCarthy, 2014; Smith, et
al., 2015). This reluctance, caused by uncertainty, could
hinder decision-making and result in projects maintaining
inadequate small-scale management which does not ensure
species survival.
Here we address this issue of outcome uncertainty
and the importance of communication between scientists,
project managers and stakeholders concerning the
Mauritius olive white-eye (Zosterops chloronothos), a
critically endangered passerine endemic to Mauritius and
highly threatened by invasive rats (Maggs, et al., 2015;
Birdlife International, 2016). The olive white-eye is part of
an ancient Indian Ocean white-eye lineage and is in the top
10% of the EDGE bird species list based on their high level
of endemism and evolutionary distinctiveness (Warren, et
al., 2006; Jetz, et al., 2014). Research has identified rats
(black and brown) as a major limiting factor for olive whiteeye, preying on nests and causing an estimated annual
population decline of 14%; however, rat management
can mitigate this threat and ensure population persistence
(Maggs, et al., 2015). Based on these findings, small-scale
management has been implemented over remnant olive
white-eye breeding territories around the Combo region of
the Black River Gorges National Park (BRGNP), Mauritius
(Fig. 1; Ferrière, et al., 2016). However, small-scale rat
management is not adequate enough to enable olive whiteeye population viability in the long-term, highlighting the
need for large-scale management in the form of a mainland
island (Maggs, 2017).
Here we illustrate how a collaborative approach to
conservation management can aid decision-making through
communication between scientists, managers, and project
stakeholders which can facilitate scaling up small-scale rat
control to the implementation of a mainland island. For
highly threatened species, such as the olive white-eye, this
approach ensures the timely implementation of evidencebased decisions and bridges the gap between research and
management.
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 31–35. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
31
Island invasives: scaling up to meet the challenge. Ch 1A Rodents: Planning
knowledge exchange between experts across New Zealand
and project managers in Mauritius. Grey literature and
expert knowledge were gathered, identifying potential
management techniques and the demands and practicalities
involved which aided scientific research.
The sites visited across New Zealand varied in
management type and size but all targeted invasive rat
species (black and brown). Meetings with the experts and
managers were standardised by discussing the same topics,
these included:
Fig. 1 Mauritius, illustrating the location of the Combo
region (black rectangle) within the Black River Gorges
National Park (BRGNP).
METHODS
To combat uncertainty, two tools were used; knowledge
exchange and stakeholder workshops, in combination with
scientific research, to break down some of the barriers to
decision-making.
Knowledge exchange
When scaling-up invasive species management there
are many logistical and financial considerations. For
conservation programmes which have never established
such large-scale management, accounting for these
considerations and identifying limitations is difficult.
Methods and costings of mainland islands have been
published (Clapperton & Day, 2001; Gillies, 2002; Gillies,
et al., 2006; Scofield, et al., 2011; Burns, et al., 2012;
Norbury, et al., 2014; Carter, et al., 2016), but detailed
information regularly remains in undocumented individual
staff experiences or when data is gathered it remains in
inaccessible forms and grey literature. This compounds
information inaccessibility resulting in personnel within
programmes making decisions based on limited experience
rather than evidence (Sutherland, et al., 2004; Pullin &
Knight, 2005; Kapos, et al., 2008). Gaining first-hand
experience can enable a nuanced understanding of both
short and long-term management, which for robust and
realistic costing is vital.
To identify the considerations which should be made
and gain first-hand information a knowledge exchange
was conducted with rat control/eradication experts and
conservation managers in the field across New Zealand
during April–May 2015. These individuals were identified
either through the Hihi Recovery Group, which works
closely with numerous mainland island managers, or
identifying people through published literature. Using
a ‘boundary organisation’ approach (Cook, et al., 2013;
Cvitanovic, et al., 2015) scientific researchers facilitated
32
● Management history. Have other management
techniques been previously used, if so, what was the
scale of the management and why did it change?
● Identifying mainland island area. What process was
used to identify locations, what were the constraints
and benefits considered, how were topography and
river courses tackled and what was the conservation
objective of the mainland island?
● Management technique. What rat management
technique is currently used, over what area, and how
long has it been in place, is there a buffer zone, how
many staff and volunteers work on the site, have
additional techniques been trialled and if so what
were the outcomes?
● Maintenance. How often are the traps/stations/fence
checked or re-baited, how long does this take and
how many staff members does this require, what
maintenance demands are there, how often does
equipment need replacing and how do weather
conditions impact the management and work load?
● Management efficiency. Is rat abundance or presence
monitored in the management site, if so, what is the
rate of rat incursions or rat abundance and is there
a response protocol and if so how quickly is this
implemented?
Stakeholder workshops
Improving knowledge exchange between decision–
makers and scientists is fundamental to support sustainable
evidence-based management. However, despite evidence
being available, in some cases decisions can still remain
hindered due to multiple objectives from a mix of
stakeholders with differing priorities, values or conflicting
interests (Conroy, et al., 2002). Science can help overcome
these obstacles by providing tools to inform decisions and
aid stakeholders to make informed trade-offs if required.
An approach termed ‘interdependency’ recognises that
all participants in knowledge exchange can contribute,
emphasising the need for a two-way exchange between
scientists and decision-makers (Contandriopoulos, et
al., 2010; Cvitanovic, et al., 2015). This can increase
understanding and stakeholder communication through
access to the best scientific information, enabling
science-based decision-making (Meek, et al., 2015). This
process supports collaboration and bridges the researchimplementation gap (Knight, et al., 2008), but requires the
roles of participants to be outlined from the start to ensure
clarity throughout the workshop process; identifying
expert advisors, decision-makers or workshop facilitators
to mediate between stakeholders.
To ensure collaboration between scientific researchers
and decision-makers and avoid conflicting interests,
a stakeholder workshop was held in the case of the
olive white-eye. During this workshop there were three
main objectives to be considered by decision-makers
when tackling development from small-scale localised
management to a large-scale mainland island: should
a mainland island be established, what size it should be
to enable population viability and management cost-
Maggs, et al.: Bridging the research-management gap
effectiveness. The workshop was facilitated by scientific
researchers, from the Zoological Society of London
(ZSL) and University College London (UCL), who
provided expert advice on these three objectives; this was
accompanied by field staff providing first-hand information
on the status of the species and the current management
in place from the Mauritian Wildlife Foundation (MWF)
(Ferrière, et al., 2016).
Scientific research on the olive white-eye has
successfully developed decision-making tools identifying
the mainland island area required to ensure population
persistence and management cost-effectiveness (Maggs,
2017). These decision-making tools outline scenarios and
assist in identifying informed, evidence-based management
for the remnant olive white-eye population, ensuring
population persistence and clear financial and logistical
requirements over 50 years (see Maggs, 2017 for details).
Using these tools, discussions were held between the
expert advisors (scientific researchers and field staff) and
the key decision-makers (project managers, organisation
directors, project funders and government officials) where
the scientific evidence was discussed, expert opinion
shared and questions raised through open dialogue and in a
transparent environment.
RESULTS
Knowledge exchange
In total, over four weeks, 30 individuals participated
in the knowledge exchange including managers and
volunteers from eight mainland island sites and experts
from additional conservation companies, central
government (Department of Conservation) and specialist
groups across New Zealand (Fig. 2). The rat management
techniques identified across these sites and discussed
included trapping, ground-based poisoning, self-resetting
traps and predator-proof fencing. The information gathered
through the knowledge exchange was vital for the detailed
long-term budgeting of a mainland island in Mauritius
Fig. 2 The distribution of the mainland islands visited during
a knowledge exchange in April and May 2015 and the
organisations who participated: Hihi Recovery Group,
Biodiversity Restoration Specialists, (1) Shakespear
Open Sanctuary (Auckland Council), (2) Tawharanui
Open Sanctuary (Auckland Council), (3) Sanctuary
Mountain Maungatautari, (4) Boundary Stream Mainland
Island (Department of Conservation), (5) Rotokare Scenic
Reserve Trust, (6) Bushy Park Sanctuary, (7) Zealandia,
(8) Rotoiti Nature Recovery Project (Department of
Conservation).
under each of the four management techniques, providing
detail into the equipment and materials required and labour
demands. This first-hand information was combined with
existing literature and fed directly into scientific research
conducting cost-effectiveness analysis for the four rat
management techniques, accounting for the costs over
50 years. By accurately budgeting each management
technique over 50 years the long-term cost-effectiveness of
the four rat management techniques against olive white-eye
population quasi-extinction risk were robustly illustrated;
which acts as the effectiveness score of the rat management
techniques (Table 1; see Maggs, 2017 for full details).
Stakeholder workshop
Eighteen delegates from six organisations participated
in the stakeholder workshop; these included project
management (MWF), organisation directors (MWF and
Durrell Wildlife Conservation Trust), scientific researchers
(ZSL and UCL), project funders (Chester Zoo) and
government officials (National Parks and Conservation
Service).
The olive white-eye is a priority species for
conservation in Mauritius and it was decided, based on
the scientific findings presented, that a mainland island
should be established, aiming for the minimum area
identified by Maggs (2017) of 275 ha; required at a low
population density to ensure a 99% chance of population
persistence over 50 years. Using the cost-effectiveness
analysis conducted by Maggs (2017), and presented at
the workshop, the rat management technique decided
upon was Goodnature®A24 self-resetting traps based
on their cost-effectiveness, specifically, their low labour
requirements and competitive financial costs (Maggs,
2017). Although a relatively new technique, their longterm costs, maintenance and replacements, were accounted
for based on manufacturer recommendations; the same
long-term costs were accounted for all of the techniques
discussed.
Trapping was considered too labour intensive even
though it was highly cost-effective when considering
equipment costs alone. Poison was ruled out based on the
potential environmental impacts and the overall high cost
of poison and associated labour. Predator-proof fencing
was not considered as an option based on the huge initial
setup cost and the long-term financial commitment, also
the habitat loss associated with installing a predator-proof
fence (at least 8m of forest would need to be cleared both
sides of the fence to prevent mammals jumping over (Day,
2004); with highly threatened flora species within the
BRGNP this cannot be justified at this time). Fencing is
the most cost-effective technique when creating a mainland
island over vast areas and could maintain zero rat densities,
which the other techniques cannot achieve, but complete
rat removal is not required for olive white-eye viability,
merely reduced rat densities. The techniques combined
were not discussed but could be in an additional option to
consider in the future.
As well as the rat management technique it was also
identified that the mainland island would have to take a
‘multi-species/multi-threat’ approach, targeting a number
of invasive species until the impact of individual species is
known in order to avoid the ‘surprise factor’ of secondary
unexpected and undesired results (Alterio, et al., 1999;
Saunders & Norton, 2001; Caut, et al., 2009; Carter, et al.,
2016). This would involve targeting small Indian mongoose
(Urva auropunctata), feral cats (Felis domesticus) and
potentially crab eating macaques (Macaca fascicularis)
as well as rats. This level of predator control would also
benefit other highly threatened endemic species such as the
Mauritius cuckoo-shrike (Coracina typica), echo parakeet
33
Island invasives: scaling up to meet the challenge. Ch 1A Rodents: Planning
Table 1 The minimum area required for a mainland island to ensure a 99% chance of population persistence for the
Mauritius olive white-eye over 50 years, the total cost of establishing and running a mainland island over 50 years,
the establishment costs alone and the average annual costs; comparing trapping, ground based poisoning, selfresetting traps and predator-proof fencing (Maggs, 2017).
Area (ha)
Trapping
Poisoning
Self-resetting traps
Predator-proof fencing
275
300
275
275
(Psittacula eques) and Mauritius pink pigeon (Nesoenas
mayeri), which are found within the same regions.
Finally, it was suggested that, if possible, the site of
a mainland island should be combined with existing
conservation management areas (CMAs), which have
been established on mainland Mauritius in the BRGNP to
protect native vegetation communities by removing exotic
flora (Cheke & Hume, 2008). The control of rats would
encourage habitat regeneration and resources could be
combined for both invasive fauna and flora control.
DISCUSSION
This case study aimed to illustrate how a collaborative
approach to conservation management, through
knowledge exchange and stakeholder workshops, can aid
communication and decision-making. In this case, it was
used to guide the timely expansion of rat management
from existing small-scale control (32 ha) to a mainland
island (275 ha), relatively quickly and effectively, which
is vital for highly threatened and declining species, such as
the olive white-eye.
A mainland island has never been established in
Mauritius. The rat management techniques used for
the olive white-eye have been limited to localised snaptrapping and ground-based poisoning (Maggs, et al.,
2015). In the past, feasibility studies have been conducted
for various techniques, including predator-proof fencing,
but taking the step from localised to landscape scale
management was not taken due to resource limitations and
long-term financial and logistical uncertainty (Day, 2004).
Here we have tackled the barriers of logistical and
financial uncertainty and decision-making through a
‘co-production’ approach with full cooperation between
scientific researchers and decision-makers (Cvitanovic,
et al., 2015; van Kerkhoff & Lebel, 2015). Conducting
knowledge exchange allowed project managers to gain
first-hand information and fill knowledge gaps from
leading experts in the field of invasive species management.
Incorporating this into a robust analysis of the financial
and logistical requirements of a mainland island helped to
minimise uncertainty, justify expenditure and identify the
long-term financial support required from funders (Maggs,
2017). A stakeholder workshop then allowed scientific
research to be fed directly back to all involved, successfully
highlighting project priorities and enabling all participants
to come to a unified decision on future management goals
for the olive white-eye; guiding science-based conservation
while maintaining transparency among stakeholders.
Through this collaborative approach, in just three
years, long-term management goals have been identified
to establish the first mainland island in Mauritius to protect
the olive white-eye and ensure long-term population
34
Total Cost
(£ millions)
2.9
7.9
3.8
5.7
Establishment
Costs (£)
186,700
40,925
130,315
1,766,472
Annual
costs (£)
37,908
157,913
37,505
80,196
viability. Implementation of a mainland island within the
national park has started in the Brise Fer CMA with the
introduction of olive white-eye planned for 2021 if rat
management can maintain adequately low rat densities
over a prolonged period. The area of the mainland island
will be increased with growth in capacity, aiming to reach
the full mainland island size (275 ha) within 5–10 years.
Without these processes, project decisions could have
taken years longer to reach the same point if field trials were
required (to test all potential rat management techniques),
accurate long-term financial requirements were not known,
open discussion was not had or scientific research was not
fed back to decision-makers; delays which would have
detrimental impacts on highly threatened and declining
species like the olive white-eye.
The methods discussed here address ways to approach
existing challenges, reduce uncertainty and enable
evidence-based decision-making. The approaches taken,
although case-specific, provide methods for researchers
and managers to adopt and apply to different scenarios
depending on the decision-making barriers and uncertainty
being faced; bridging both the knowledge-action boundary
and the research-management divide (Roux, et al., 2006;
Cook, et al., 2013), which is rarely achieved in conservation.
ACKNOWLEDGEMENTS
The Mauritius olive white-eye recovery project is
conducted by the Mauritian Wildlife Foundation in
conjunction with the National Parks and Conservation
Service (Government of Mauritius). We would like to thank
all those who participated in the knowledge exchange –
The Hihi Recovery Group: John Ewen, Lynn Adams, Troy
Makan, Doug Armstrong, Allan Anderson, Kevin Parker,
Julie Panfylova, John Perrott, Neil Anderson, Ian Fraser,
John Stewart, Vix Franks, Anna Santure, Kate Richardson,
Morag Fordham, Gen Spargo and Nick Fisentzidis; Peter
Abbott, Jenny Long, Graeme Elliott, Jason Malham and
Ruth Malham from the Department of Conservation; Matt
Maitland from Auckland Council; Ron Millar and Matt
Cook from Sanctuary Mountain Maungatautari; Allan
Anderson from Bushy Park Sanctuary; Simon Collins
from Rotokare Scenic Reserve Trust; Raewyn Empson
and Russ Drewry from Zealandia; and Grant Harper from
Biodiversity Restoration Specialists. We would like to thank
the Mauritius olive white-eye project partner organisations
for participating in the stakeholder workshop – Mauritian
Wildlife Foundation, Zoological Society of London,
University College London, Chester Zoo, Durrell Wildlife
Conservation Trust and National Parks and Conservation
Service. We would finally like to thank the anonymous
reviewers for their comments and contributions to earlier
drafts. Gwen Maggs was funded by Chester Zoo (UK) and
Zoological Society of London.
Maggs, et al.: Bridging the research-management gap
REFERENCES
Alterio, N., Moller, H. and Brown, K. (1999). ‘Trappability and densities
of stoats (Mustela erminea) and ship rats (Rattus rattus) in a South
Island Nothofagus forest, New Zealand’. New Zealand Journal of
Ecology 23: 95–100.
Atkinson, I.A.E. (1985). ‘The Spread of Commensal Species of Rattus
to Oceanic Islands and their Effects on Island Avifaunas’. In: P.J.
Moors (ed.) Conservation of Island Birds, pp. 35–81. ICBP Technical
Publication 3. Cambridge, UK: International Council for Bird
Preservation (ICBP).
Birdlife International (2016). The IUCN Red List of Threatened Species
2016. <www.iucnrelist.org/details/22714005/0>.
Blackburn, T.M., Essl, F., Evans, T., Hulme, P.E., Jeschke, J.M., Kühn,
I., Kumschick, S., Marková, Z., Mrugała, A., Nentwig, W., Pergl, J.,
Pyšek, P., Rabitsch, W., Ricciardi, A., Richardson, D.M., Sendek,
A., Vilà, M., Wilson, J.R.U., Winter, M., Genovesi, P. and Bacher, S.
(2014). ‘A unified classification of alien species based on the magnitude
of their environmental impacts’. PLOS Biology 12: 1–11.
Burns, B., Innes, J. and Day, T. (2012). ‘The Use and Potential of Pestproof Fencing for Ecosystem Restoration and Fauna Conservation
in New Zealand’. In: M. Somers and M. Hayward (eds.) Fencing for
Conservation: Restriction of Evolutionary Potential or a Riposte to
Threatening Processes? pp. 1–320. Springer Science+Business Media.
Butler, D., Lindsay, T. and Hunt, J. (2014). Paradise Saved. Auckland,
NZ: Random House New Zealand.
Carter, A., Barr, S., Bond, C., Paske, G., Peters, D. and van Dam, R.
(2016). ‘Controlling sympatric pest mammal populations in New
Zealand with self-resetting, toxicant-free traps: a promising tool for
invasive species management’. Biological Invasions 18: 1723–1736.
Caut, S., Angulo, E. and Courchamp, F. (2009). ‘Avoiding surprise
effects on Surprise Island: Alien species control in a multitrophic level
perspective’. Biological Invasions 11: 1689–1703.
Jetz, W., Thomas, G.H., Joy, J.B., Redding, D.W., Hartmann, K. and
Mooers, A.O. (2014). ‘Global distribution and conservation of
evolutionary distinctness in birds’. Current Biology 24: 919–930.
Jones, C.G. and Merton, D.V. (2012). ‘A Tale of Two Islands: The Rescue
and Recovery of Endemic Birds in New Zealand and Mauritius’.
In: E.G. Ewen, D.P. Armstrong, K.A. Parker and P.J. Seddon (eds.)
Reintroduction Biology: Integrating Science and Management, pp.
33–72. Wiley-Blackwell.
Kapos, V., Balmford, A., Aveling, R., Bubb, P., Carey, P., Entwistle, A.,
Hopkins, J., Mulliken, T., Safford, R., Stattersfield, A., Walpole, M. and
Manica, A. (2008). ‘Calibrating conservation: new tools for measuring
success’. Conservation Letters 1: 155–164.
Knight, A.T., Cowling, R.M., Rouget, M., Balmford, A., Lombard, A.T.
and Campbell, B.M. (2008). ‘Knowing but not doing: Selecting priority
conservation areas and the research-implementation gap’. Conservation
Biology 22: 610–617.
Maggs, G., Nicoll, M., Zuël, N., White, P.J.C., Winfield, E., Poongavanan,
S., Tatayah, V., Jones, C.G. and Norris, K. (2015). ‘Rattus management
is essential for population persistence in a critically endangered
passerine: Combining small-scale field experiments and population
modelling’. Biological Conservation 191: 274–281.
Maggs, G. (2017). ‘The Ecology and Conservation of Wild and
Reintroduced Populations of the Critically Endangered Mauritius
Olive White-eye (Zosterops chloronothos). PhD thesis: London, UK:
University College London.
McCarthy, M.A. (2014). ‘Contending with uncertainty in conservation
management decisions’. Annals of the New York Academy of Sciences
1322: 77–91.
Cheke, A. and Hume, J. (2008). ‘Lost Land of the Dodo’. United
Kingdom: T & AD Poyser.
Meek, M.H., Wells, C., Tomalty, K.M., Ashander, J., Cole, E.M., Gille,
D.A., Putman, B.J., Rose, J.P., Savoca, M.S., Yamane, L., Hull, J.M.,
Rogers, D.L., Rosenblum, E.B., Shogren, J.F., Swaisgood, R.R. and
May, B. (2015). ‘Fear of failure in conservation: The problem and
potential solutions to aid conservation of extremely small populations’.
Biological Conservation 184: 209–217.
Clapperton, B.K. and Day, T.D. (2001). ‘Cost-effectiveness of exclusion
fencing for stoat and other pest control compared with conventional
control’. DOC Science Internal Series 14: 5–19.
Norbury, G., Hutcheon, A., Reardon, J. and Daigneault, A. (2014). ‘Pest
fencing or pest trapping: A bio-economic analysis of cost-effectiveness’.
Austral Ecology 39: 795–807.
Conroy, D.E., Willow, J.P. and Metzler, J.N. (2002). ‘Multidimensional
fear of failure measurement: The performance failure appraisal
inventory’. Journal of Applied Sport Psychology 14: 76–90.
Pullin, A.S. and Knight, T.M. (2005). ‘Assessing conservation
management’s evidence base: A survey of management-plan compilers
in the United Kingdom and Australia’. Conservation Biology 19: 1989–
1996.
Contandriopoulos, D., Lemire, M., Denis, J.-L. and Tremblay, E. (2010).
‘Knowledge exchange processes in organizations and policy arenas: A
narrative systematic review of the literature’. The Milbank Quarterly
8: 444–483.
Cook, C.N., Mascia, M.B., Schwartz, M.W., Possingham, H.P. and
Fuller, R.A. (2013). ‘Achieving conservation science that bridges the
knowledge-action boundary’. Conservation Biology 27: 669–678.
Cullen, R., Fairburn, G.A. and Hughey, K.F.D. (2001). ‘Measuring the
productivity of threatened-species programs’. Ecology Economics 39:
53–66.
Cvitanovic, C., Hobday, A.J., van Kerkhoff, L., Wilson, S.K., Dobbs, K.
and Marshall, N.A. (2015). ‘Improving knowledge exchange among
scientists and decision-makers to facilitate the adaptive governance of
marine resources: A review of knowledge and research needs’. Ocean
and Coastal Management 112: 25–35.
Day, T. (2004). Feasibility of pest proof fencing in Mauritius. Cambridge,
New Zealand: XcluderTM Pest Proof Fencing Company.
Diamond, J. (1989). ‘Overview of Recent Extinctions’. In: D. Western
and M. Pearl (eds.) Conservation for the Twenty-First Century, pp.
37–41. New York: Oxford University Press.
DIISE. (2015). The Database of Island Invasive Species Eradications,
developed by Island Conservation, Coastal Conservation Action
Laboratory UCSC, IUCN SSC Invasive Species Specialist Group,
University of Auckland and Landcare Research New Zealand. <http://
diise.islandconservation.org/>.
Ferrière, C., Page, J., Zuël, N., Tatayah, V. and Jones, C. (2016). Mauritius
Olive White-eye Recovery Programme Annual Report 2015–16. Vacoas,
Mauritius.
Gillies, C. (2002). ‘Managing Rodents on the New Zealand Mainland –
What Options are Currently Available?’. DOC Scientific Internal Series
47: 5–20.
Gillies, C., Styche, A., Bradfield, P., Chalmers, K., Leach, M., Murphy,
E., Ward-Smith, T. and Warne, R. (2006). ‘Diphacinone Bait for
Ground Control of Rats on Mainland Conservation Land’. Science for
Conservation 270: 5–20.
Rodrigues, A.S.L., Brooks, T.M., Butchart, S.H.M., Chanson, J., Cox, N.,
Hoffmann, M. and Stuart, S.N. (2014). ‘Spatially explicit trends in the
global conservation status of vertebrates’. PLOS ONE 9: 1–18.
Roux, D.J., Rogers, K.H., Biggs, H.C., Ashton, P.J. and Sergeant, A.
(2006). ‘Bridging the science-management divide: Moving from
unidirectional knowledge transfer to knowledge interfacing and
sharing’. Ecology and Society 11(1): 4.
Saunders, A. and Norton, D.A. (2001). ‘Ecological restoration at mainland
islands in New Zealand’. Biological Conservation 99:109–119.
Scofield, R.P., Cullen, R. and Wang, M. (2011). ‘Are predator-proof
fences the answer to New Zealand’s terrestrial faunal biodiversity
crisis?’ New Zealand Journal of Ecology 35: 312–317.
Smith, M., Wallace, K., Lewis, L. and Wagner, C. (2015). ‘A structured
elicitation method to identify key direct risk factors for the management
of natural resources’. Heliyon 1: e00043.
Sutherland, W.J., Pullin, A.S., Dolman, P.M. and Knight, T.M. (2004).
‘The need for evidence-based conservation’. Trends in Ecology and
Evolution 19: 305–308.
Towns, D.R. and Broome, K.G. (2003). ‘From small Maria to massive
Campbell: Forty years of rat eradications from New Zealand islands’.
New Zealand Journal of Zoology 30: 377–398.
Towns, D.R., Atkinson, I.A.E. and Daugherty, C.H. (2006). ‘Have
the harmful effects of introduced rats on islands been exaggerated?’.
Biological Invasions 8: 863–891.
van Kerkhoff, L.E. and Lebel, L. (2015). ‘Coproductive capacities:
Rethinking science-governance relations in a diverse world’. Ecology
and Society 20: 14.
Warren, B.H., Bermingham, E., Prys-Jones, R.P. and Thébaud, C. (2006).
‘Immigration, species radiation and extinction in a highly diverse
songbird lineage: White-eyes on Indian Ocean islands’. Molecular
Ecology 15: 3769–3786.
35
J.P. Parkes
Parkes, J.P. Timing aerial baiting for rodent eradications on cool temperate islands: mice on Marion Island
Timing aerial baiting for rodent eradications on cool temperate islands:
mice on Marion Island
J.P. Parkes
Kurahaupo Consulting, 2 Ashdale Lane, Christchurch, New Zealand. <john.parkes1080@gmail.com>.
Abstract Aerial baiting from helicopters with a bait-sowing bucket and GPS to ensure coverage with anticoagulant
toxins in cereal-based baits can reliably eradicate rodents on islands. Current best practice for temperate islands is to bait
in winter when the rodents are not breeding, rodent numbers are lowest so competition for toxic baits is lowest, natural
food is likely to be scarce, and many non-target species are absent from the island. However, short winter day lengths
at high latitudes restrict the time helicopters can fly and poor weather in winter may increase risks of failure. This paper
notes precedents from cool temperate islands where baiting was not conducted in winter and then uses the extensive data
on mice on Marion Island to explore whether current recommendations for winter baiting based on breeding and natural
food availability are important risk factors in determining time of year to bait. Marion Island mice do not breed between
early May and late September, mouse densities reach a maximum in May and minimum in November, but the biomass
of main natural food (invertebrates) does not fluctuate greatly over the year. This means the per capita food availability is
least in autumn and increases through winter to most in spring and summer. The weight of the stomach contents of mice
is also highest in winter. Based on this per capita food parameter, mice are likely to be most hungry between about March
and May suggesting baiting would be more effective in this period (perhaps towards the end of it when breeding stops)
than in the more traditional winter season.
Keywords: food availability, rodent abundance, seasonality
INTRODUCTION
Successful attempts to eradicate one or more (up
to four) species of rodent by sowing toxic baits from
an aircraft have been made on at least 166 islands in 13
countries (DIISE, 2018; J. Parkes, unpubl. data) since
the first use of this method in 1985 against Norway rats
(Rattus norvegicus) on Whale Island (143 ha) in New
Zealand (Imber, et al., 2000). Most of these islands are at
latitudes with temperate climates (n = 96) biased by the
large sample from New Zealand, or in tropical latitudes (n
= 64) biased by those in the Montebello Group of islands
in Western Australia. Few islands are at latitudes with cold
climates similar to Marion Island (n = 6). Aerial baiting
is currently the only practical option to eradicate rodents
on large or topographically difficult islands and has a
high success rate when modern best practice is followed
(Parkes, et al., 2011; Parkes, 2016). The cost of operational
failure is high, especially for large, remote islands, both in
the money invested (Holmes, et al., 2016) and if failures
discourage risk-averse funders from attempting further
projects. Therefore, careful planning and application of
best practice based on precedence and analysis of the
particular constraints and risks for each project is essential.
Pest eradications achieved by reduction of the target
population to zero by a sequence of removal events (e.g. by
shooting, trapping or by deployment and re-baiting of bait
stations) provide information (e.g. catch per unit effort, kill
locations, trends in rates of bait-take across seasons and
years) as the population is reduced (e.g. Thomas & Taylor,
2002). Under this strategy, the ‘start rules’ are not critical
as managers can (and should) adapt actions as information
accrues during the project, e.g. to allow a change in tactics
to account for animals that might avoid one control method
(Parkes, et al., 2010). In these projects knowing when to
stop and declare success is the more critical issue – at least
in terms of efficiency and risk management (Ramsey, et
al., 2011).
In contrast, the use of aerial baiting provides little
information on likely success or failure from the control
itself, other than bait coverage if GPS technology is
used. Under this strategy everything has to ‘go right on
the day’ and ‘start rules’ with meticulous planning are
critical (Cromarty, et al., 2002; Springer, 2016). One
key ‘start rule’ is to identify the optimal time of year (or
at least avoid sub-optimal times) to conduct the baiting.
Broome, et al. (2014) suggest that winter to early spring is
the preferred time of year to aerially bait rodents on New
Zealand’s temperate islands because it is supposed that (a)
rodents are often not breeding and so young individuals
that might not be exposed to bait because of possible lack
of dependence or subordinate behaviours are at lowest
levels of abundance, (b) rodent densities are likely to be
lowest and so competition for baits least, (c) natural foods
are likely to be least abundant, the rodents most hungry and
so 100% are likely to eat the baits, and (d) some potential
non-target animals such as seabirds are not present in this
season. Most rodent eradication projects have followed this
advice by baiting in winter for temperate islands. However,
these factors are not always mutually independent (least
food and fewest rodents), and other factors (weather or
logistics) may constrain decisions. The parameter around
food availability we are really seeking is the time of year
when there is least per capita food, which may or may not
be when there is least food or fewest rodents and may or
may not be what drives any breeding season. Managers are
probably wise to stick with precedence and bait in winter
or early spring (or during dry seasons in the tropics) in
the absence of any data on the seasonality of food, rodent
dynamics or breeding seasonality.
However, for a variety of reasons a few rodent
eradications on cool temperate islands using aerial baiting
have been conducted in the summer. Mice (Mus musculus)
were eradicated on Enderby Island (710 ha at 50⁰S) in
January 1993 because that was when the primary target
species, the rabbits (Oryctolagus cuniculus), were not
breeding (Torr, 2002). Norway rats and house mice were
eradicated on the subantarctic island of South Georgia
(103,000 ha and 4,900 ha, respectively, at 54⁰S) between
late February and late May (mostly in March–April) in
phases between 2011, 2013 and 2015 because weather
conditions and persistent snow cover made a winter
operation impossible (Anon., 2016; Martin & Richardson,
2017). Timing and other operational details of aerial baiting
on several islands in the French Southern Territories appear
to have been determined by the timing of the supply ship,
the Marion Dufrense. Rabbits and ship rats (Rattus rattus),
but not mice were eradicated from Saint Paul Island (900 ha
at 38⁰S) in January–February (Micol & Jouventin, 2002).
Attempts to eradicate rodents from some of the islands in
the Golfe du Morbihan in the Kerguelen group (49⁰S) have
In:
36C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 36–39. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Parkes: Timing baiting for mice on Marion Island
been made during summer months when the supply ship
visits the region. Ship rats and mice were eradicated from
Île Château (220 ha) (Anon., 2006) and ship rats but not
mice from Île Australia (2337 ha). Attempts to eradicate
mice on Île Stoll (60 ha) and ship rats and mice on Île
Moules (500 ha) failed (Anon., 2006; DIISE, 2018).
So, maybe we are unnecessarily constraining ourselves
to times of year with the worst weather and shortest days
on islands at high latitudes by baiting in winter. This paper
explores this seasonality question by describing the process
used to inform decision-makers of a proposed eradication
of mice on Marion Island, a place where the long history of
research by South African scientists has provided most of
the information to answer the question.
RESULTS AND DISCUSSION
Marion Island
Marion Island (29,000 ha) and the adjacent Prince
Edward Island are South Africa’s only offshore islands.
They lie on the sub-Antarctic convergence at 46⁰54′S in
the south Indian Ocean. Apart from a meteorological
station on Marion Island, the islands are uninhabited.
Marion is an active volcano rising to 1,230 m a.s.l. (Fig.
1). The climate is cool, wet and temperate with only a few
degrees seasonal variation between coldest and warmest
months (mean annual temperature is 6.4⁰C and mean
annual precipitation is about 200 cm). The physical and
biotic characters of Marion Island are described in detail in
Chown & Froneman (2008) and the impacts and history of
the introduced flora and fauna by Angel & Cooper (2011)
and Greve, et al. (2017).
Mice were introduced accidentally some 200 years
ago, probably with sealers, and are having a significant
impact on the native biota (Angel & Cooper, 2011; Dilley,
et al., 2016) such that the South African government is
considering whether they might be eradicated (Parkes,
2016). Cats (Felis catus) were introduced in 1948 in an
attempt to control mice at the meteorological station but
soon spread as feral animals over the island, killing mice
as primary prey and an estimated 450,000 seabirds per year
(Dilley, et al., 2017). The cats were eradicated between
1977 and 1991 (Bester, et al., 2002).
Breeding season
Mice can breed all year if high quality food is available,
e.g. during beech (Nothofagus spp.) mast events in New
Zealand winters (Ruscoe, et al., 2005). However, mice
Fig. 1 Vegetated lava (foreground) and swamp habitat
(middle background), Marion Island (Photo by John
Parkes, April 2016).
Fig. 2 Monthly pregnancy rates of adult mice, Marion Island
in 1991/92 (after Matthewson, et al. (1994) black bars)
and 1992/93 (after Avenant & Smith (2004) grey bars).
have a distinct breeding season on Marion Island with
no pregnant animals present between early May and late
September (Fig. 2). However, this is not a universal rule on
all cool temperate islands as 16% of mice sampled during
August/September 2012 on Steeple Jason Island (51⁰S in
the Falkland Islands) were pregnant (Rexer-Huber, et al.,
2013).
Density of mice and competition for bait
This breeding season is reflected in the monthly
abundance of mice on Marion Island with increasing
numbers from the start of breeding in late spring and
declining numbers once breeding ends in late autumn (Fig.
3), resulting in lowest densities (at the favoured habitats)
at the start of the breeding season (43/ha) in spring and
highest (242/ha) in early winter before the decline (Avenant
& Smith, 2004).
Baiting during low rodent densities is recommended by
Broome, et al. (2014) in part to ensure there are plenty of
baits such that all mice, irrespective of their social status,
have access to baits. Bait sowing rates in high-density
rodent populations of 8 kg/ha in an initial sowing followed
by a second sowing of 6 kg/ha about eight days later would
result in 7,000 baits/ha – or even in the highest density
mouse habitats of Marion Island of 23 baits per mouse.
This seems more than adequate to overcome any potential
between-mouse competition given each bait contains a
lethal dose.
Fig. 3 Seasonal abundance of mice (minimum number
known to be alive/ha) averaged across three main
habitat types, Marion Island (after Ferreira, et al., 2006).
37
Island invasives: scaling up to meet the challenge. Ch 1A Rodents: Planning
Fig. 5 Monthly changes in the weight of stomach contents
adjusted for body weights of Marion Island mice (after
Matthewson, et al. (1994) reported in Smith, et al.
(2002)).
Fig. 4 Seasonal biomass of main invertebrate prey of
mice (after Gleeson & van Rensburg (1982)). Total
invertebrate biomass (top solid line), weevils (dotted
line), moth larvae (dashed line), spiders (lower broken
line).
Seasonal variation in per capita natural food
The decades of detailed research conducted on Marion
Island (Chown & Froneman (2008) have included studies
on the seasonal diet of mice and on the seasonal biomass
of their prey. Invertebrates form the bulk of mice diet
(depending on habitat) with the larvae and adults of the
flightless keystone moth (Pringleophaga marioni) (between
13 and 64% by volume) and weevils (Ectemnorhinus spp.)
(between 11 and 32% by volume) being the most important,
followed by earthworms (Microscolex kerguelarum)
(between 1 and 9% by volume). Plant material, mostly
grass and sedge seeds was important, between 16 and 48%
by volume (Smith, et al., 2002).
There appears to be only small seasonal variation in
the abundance of the main invertebrate fauna favoured
by mice (Fig. 4) and Avenant & Smith (2004) found no
significant summer–winter differences in invertebrate
biomass in the habitat most favoured by mice – apart from
spiders which were actually more abundant in winter.
The preferred prey for mice, larvae, pupae and adults of
Pringleophaga marioni, has a long-life cycle of between
two and five years (Haupt, et al., 2014) so the absence of
seasonal fluctuations is not unexpected given also the small
seasonal differences in climate on Marion Island (le Roux
& McGeoch, 2008).
by holding individuals in safe captivity (Rexer-Huber &
Parker, 2011), but risks to the latter have to be accepted
(e.g. as on Macquarie Island; Parkes, 2016; Springer, 2016)
or avoided by baiting when the birds are least common on
the island. Marion Island has only two terrestrial birds
at risk – the kelp gull (Larus dominicanus) and lesser
sheathbill (Chionis minor) while among the 26 nesting
seabird species only three (sub-Antarctic skua (Catharacta
antarctica), southern giant petrel (Macronectes giganteus)
and northern giant petrel (M. hallii) are at low to modest
risk if the baiting was conducted in mid-winter (Parkes,
2016; Springer, 2016).
CONCLUSIONS
Optimal timing of aerial baiting on Marion Island
depends on whether the non-breeding season is more or
less important than the period with minimum per capita
food availability for the mice. Neither hypothesis has
been tested. If the latter is most important then a March–
May baiting is indicated, but if the former then a May–
September baiting is indicated – May at least being a month
of overlap. Of course, an earlier timing in late summer is
better than a later one in winter, when short days, snow and
gales limit flying time.
Seasonal absence of non-target species
It is not clear whether the lack of large changes in
seasonal abundance and biomass of invertebrates seen on
Marion Island is normal for all cool temperate islands.
Most studies on other islands lack the year-round data
on changes in invertebrate biomass available for Marion
Island. However, mice on other cool temperate islands
also show a lack of strong seasonality in the occurrence
of invertebrates (the bulk of their diet) in their diet, e.g.
on Macquarie Island (Copson, 1986) and Île Guillou (Le
Roux et al., 2002). This suggests the multi-year life cycles
of the invertebrate species on Marion Island may also apply
on similar islands and the per capita food supply depends
on seasonal changes in mouse density rather than on food
abundance. Therefore, mice are likely to be hungriest
when they are at maximum densities and not during the
winter when they are likely to be least hungry and perhaps
less likely to eat artificial food such as baits. An aerial
baiting project between March and May is indicated on
this condition. Of course, the other considerations mooted
by Broome, et al. (2014) might constrain such a choice, as
might weather, day length, logistics of ship and helicopter
availability as with other projects noted in the introduction.
Most cool temperate islands have a mix of permanent
resident bird and seasonally present nesting or moulting
seabirds. Unacceptable risks to the former from toxic
baiting and secondary poisoning have to be mitigated, e.g.
However, there are several
comparisons between mouse and
derived across several studies over
may not be a problem except that
The absence of strong seasonal changes in invertebrate
prey abundance mean that there is least food per mouse
when mouse density is at a maximum, i.e. between March
and May, and most food per mouse over winter and spring.
For example, the per capita food availability is an order
of magnitude lower in early winter when mice are at
maximum densities than in early summer when they are at
lowest densities. The weight of stomach contents of mice
also increases during winter to reflect this (Fig. 5), and
mice begin to scavenge or prey upon other mice in autumn
and winter (Smith, et al., 2002).
38
caveats. First, the
food abundance are
several decades. This
the whole ecosystem
Parkes: Timing baiting for mice on Marion Island
around mice on Marion Island is highly dynamic. Second,
the biomass of invertebrates has collapsed by about 90%
since the mid-1970s (Table 1 in Parkes, 2016 and references
therein), despite which mouse densities have increased
(between 1990 and 2002; Ferreira, et al. (2006) and well
after cats were eradicated; the climate is warming (le Roux
& McGeoch, 2008); and mice are switching their primary
prey from moths to weevils and earthworms (Chown &
Smith, 1993) and learning to eat albatross chicks (Dilley,
et al., 2016).
Finally, if natural food availability is a problem
limiting bait acceptance by rodents (i.e. the proportion of
a population that eat the bait) as suspected for some recent
failures on tropical islands (Parkes, et al., 2011; Keitt, et
al., 2014), and such food competition cannot be predicted
or avoided, then one solution is to increase the palatability
of the bait relative to natural foods by adding lures or
attractants.
ACKNOWLEDGEMENTS
I thank the South African Department of Environmental
Affairs, the South African National Antarctic Programme
and Birdlife South Africa for support during my visit to
Marion Island. I also thank Ross Wanless and John Cooper
and the many researchers, mostly ornithologists, for
discussions on the problem when I was on the island and
the many previous researchers on whose work I have relied
to collate the argument for this paper.
REFERENCES
Angel, A. and Cooper, J. (2011). A review of the Impacts of the House
Mouse Mus musculus on sub-Antarctic Marion Island, Prince Edward
Islands. Report to the Prince Edward Islands Management Committee,
South African National Antarctic Programme. Core Initiatives.
Anon. (2006). Status of Measures Implemented by France in the French
Southern and Antarctic lands (TAAF) Relating to the Conservation of
Petrels and Albatrosses. Agreement on the Conservation of Albatrosses
and Petrels. Second meeting of advisory committee, Brasilia, Brazil
5–8 June 2006.
Anon. (2016). The South Georgia Habitat Restoration Project: Final
Report. Dundee, Scotland: South Georgia Heritage Trust.
Avenant, N.L. and Smith, V.R. (2004). ‘Seasonal changes in age class
structure and reproductive status of house mice on Marion island (subAntarctic)’. Polar Biology 27: 99–111.
Bester, M.N., Bloomer, J.P., van Aarde, R.J., Erasmus, B.H., van
Rensburg, P.J.J., Skinner, J.D., Howell, P.G. and Naude, T.W. (2002).
‘A review of the successful eradication of feral cats from sun-Antarctic
Marion Island, southern Indian ocean’. South African Journal of Wildlife
Research 32: 65–73.
Broome, K., Cox, A., Golding, C., Cromarty, P., Bell, P. and McClelland
P. (2014). Rat Eradication Using Aerial Baiting. Current Agreed Best
Practice used in New Zealand (Version 3.0). Wellington, New Zealand:
New Zealand Department of Conservation.
Chown, S.L. and Smith, V.R. (1993). ‘Climate change and the short-term
impact of feral house mice at the sub-Antarctic Prince Edward Islands’.
Oecologia 96: 508–516.
Chown, S.L. and Froneman, P.W. (eds.) (2008). The Prince Edward
Islands. Land-sea Interactions in a Changing Ecosystem. Stellenbosch,
South Africa: Sun Press.
Copson, G.R. (1986). ‘The diet of the introduced rodents Mus musculus
L. and Rattus rattus L. on subantarctic Macquarie Island’. Australian
Wildlife Research 13: 441–445.
Cromarty, P.L., Broome, K.G., Cox, A., Empson, R.A., Hutchinson,
W.M. and McFadden I. (2002). ‘Eradication planning for invasive
alien animal species on islands – the approach developed by the New
Zealand Department of Conservation’. In: C.R. Veitch and M.N. Clout
(eds.). Turning the tide: the eradication of invasive species, pp. 85–91.
Occasional Paper SSC no. 28. IUCN SSC Invasive Species Specialist
Group. IUCN, Gland, Switzerland and Cambridge, UK.
DIISE (2018). ‘The Database of Island Invasive Species Eradications,
developed by Island Conservation, Coastal Conservation Action
Laboratory UCSC, IUCN SSC Invasive Species Specialist Group’.
University of Auckland and Landcare Research New Zealand. <http://
diise.islandconservation.org>.
Dilley, B.J., Schoombie, S., Schoombie, J., and Ryan, P.G. (2016).
‘“Scalping” of albatross fledglings by introduced mice spreads rapidly
at Marion Island’. Antarctic Science 28: 73–80.
Dilley, B.J., Schramm, M. and Ryan, P.G. (2017). ‘Modest increases in
densities of burrow-nesting petrels following the removal of cats (Felis
catus) from Marion Island’. Polar Biology 40: 625–637.
Ferreira, S.M., van Aarde, R.J. and Wassenaar, T.D. (2006). ‘Demographic
responses of house mice to density and temperature on sub-Antarctic
Marion Island’. Polar Biology 30: 83–94.
Gleeson, J.P. and van Rensburg, P.J.J. (1982). ‘Feeding ecology of the
house mouse Mus musculus on Marion Island’. South African Journal
of Antarctic Research 12: 34–39.
Greve, M., Mathakutha, R., Steyn, C. and Chown, S.L. (2017). ‘Terrestrial
invasions on sub-Antarctic Marion and Prince Edward Islands’. Bothalia
47(2): a2143.https://doi.org/10.4102/abc.v4712.2143.
Haupt, T.M., Crafford, J.E. and Chown, S.L. (2014). ‘Solving the puzzle
of Pringleophaga – threatened, keystone detritivores in the subAntarctic’. Insect Conservation and Diversity 7: 308–313.
Holmes, N.D., Campbell, K.J., Keitt, B.S., Griffiths, R., Beek, J., Donlan,
C.J. and Broome, K. (2016). ‘Correction: reporting costs for invasive
vertebrate eradications’. Biological Invasions 18: 2801–2807.
Imber, M., Harrison, M. and Harrison, J. (2000). ‘Interactions between
petrels, rats and rabbits on Whale Island, and effects of rat and rabbit
eradication’. New Zealand Journal of Ecology 24: 153–160.
Keitt, B., Griffiths, R., Boudjelas, S., Broome, K., Cranwell, S., Millett,
J., Pitt, W. and Samaniego-Herrera, A. (2014). ‘Best practice guidelines
for rat eradication on tropical islands’. Biological Conservation 185:
17–26.
Le Roux, V., Chapuis, J-L., Frenot, Y. and Vernon, P. (2002). ‘Diet of the
house mouse (Mus musculus) on Guillou Island, Kerguelen archipelago,
Subantarctic’. Polar Biology 25: 49–57.
le Roux, P.C. and McGeoch, M.A. (2008). ‘Changes in climate extremes,
variability and signature on sub-Antarctic Marion Island’. Climate
Change 86: 309–329.
Martin, A.R., and Richardson, M.G. (2017). ‘Rodent eradication scaled
up: clearing rats and mice from South Georgia.’ Oryx: 1-9. doi.
org/10.1017/S003060531700028X
Matthewson, D.C., van Aarde, R.J. and Skinner, J.D. (1994). ‘Population
biology of house mice (Mus musculus L.) on sub-Antarctic Marion
Island’. South African Journal of Zoology 29: 99–106.
Micol, T. and Jouventin, P. (2002). ‘Eradication of rats and rabbits from
Saint-Paul Island, French Southern Territories’. In: C.R. Veitch and
M.N. Clout (eds.). Turning the tide: the eradication of invasive species,
pp. 199–205. Occasional Paper SSC no. 28. IUCN SSC Invasive Species
Specialist Group. IUCN, Gland, Switzerland and Cambridge, UK.
Parkes, J.P., Ramsey, D.S.L., Macdonald, N., Walker, K., McKnight,
S., Cohen, B.S. and Morrison, S.A. (2010). ‘Rapid eradication of
feral pigs (Sus scrofa) from Santa Cruz Island, California’. Biological
Conservation 143: 634–641.
Parkes, J., Fisher, P. and Forrester, G. (2011). ‘Diagnosing the cause of
failure to eradicate introduced rodents on islands: brodifacoum versus
diphacinone and method of bait delivery’. Conservation Evidence 8:
100–106.
Parkes, J. (2016). ‘Eradication of House Mice Mus musculus from
Marion Island: A Review of Feasibility, Constraints and Risks’. In:
R.M. Wanless (ed.) Birdlife South Africa Occasional Report Series 1.
Johannesburg, South Africa: Birdlife South Africa.
Ramsey, D.S.L., Parkes, J.P., Will, D., Hanson, C.C. and Campbell, K.J.
(2011). ‘Quantifying the success of feral cat eradication, San Nicolas
Island, California’. New Zealand Journal of Ecology 35: 163–173.
Rexer-Huber, K. and Parker, G. (2011). Captive Husbandry of the Gough
Island Bunting and Moorhen. RSPB Research Report No. 42. Sandy,
UK: Royal Society for the Protection of Birds.
Rexer-Huber, K., Parker, G.C., Reeves, M., Stanworth, A.J. and Cuthbert,
R.J. (2013). ‘Winter ecology of house mice and the prospects for their
eradication from Steeple Jason (Falkland Islands)’. Polar Biology36:
1791–1797.
Ruscoe, W.A., Elkington, J.S., Choquenot, D. and Allen, R.B. (2005).
‘Predation of beech seed by mice: effects of numerical and functional
responses’. Journal of Animal Ecology 74: 1005–1019.
SGHT (2016). The South Georgia Habitat Restoration Project: Final
Report. Dundee, UK: South Georgia Heritage Trust.
Smith, V.R., Avenant, N.K. and Chown, S.L. (2002). ‘The diet and impact
of house mice on a sub-Antarctic island’. Polar Biology 25: 703–715.
Springer, K. (2016). ‘Methodology and challenges of a complex multispecies eradication in the sub-Antarctic and immediate effects of
invasive species removal’. New Zealand Journal of Ecology 40: 273–
278.
Thomas, B.W. and Taylor, R.H. (2002). ‘A History of Ground-based
Rodent Eradication Techniques Developed in New Zealand, 1959–
1993’. In: C.R. Veitch and M.N. Clout (eds.). Turning the tide: the
eradication of invasive species, pp. 301–310. Occasional Paper SSC
no. 28. IUCN SSC Invasive Species Specialist Group. IUCN, Gland,
Switzerland and Cambridge, UK.
Torr, N. (2002). ‘Eradication of Rabbits and Mice from Subantarctic
Enderby and Rose Islands’. In: C.R. Veitch and M.N Clout (eds.).
Turning the tide: the eradication of invasive species, pp. 319–328.
Occasional Paper SSC no. 28. IUCN SSC Invasive Species Specialist
Group. IUCN, Gland, Switzerland and Cambridge, UK.
39
Preston, G.R.; B.J. Dilley, J. Cooper, J. Beaumont, L.F. Chauke, S. L. Chown, N. Devanunthan, M. Dopolo, L. Fikizolo, J. Heine, S. Henderson, C.A. Jacobs, F. Johnson, J. Kelly, A.B. Makhado, C. Marais,
J. Maroga, M. Mayekiso, G. McClelland, J. Mphepya, D. Muir, N. Ngcaba, N. Ngcobo, J.P. Parkes, F. Paulsen, S. Schoombie, K. Springer, C. Stringer, H. Valentine, R.M. Wanless and P.G. Ryan. South
Africa works towards eradicating introduced house mice from sub-Antarctic Marion Island: the largest island yet attempted for mice
South Africa works towards eradicating introduced house mice from
sub-Antarctic Marion Island: the largest island yet attempted for mice
G.R. Preston1, B.J. Dilley2, J. Cooper3, J. Beaumont1, L.F. Chauke1, S. L. Chown4, N. Devanunthan1, M. Dopolo1,
L. Fikizolo1, J. Heine5, S. Henderson6, C.A. Jacobs1, F. Johnson6, J. Kelly7, A.B. Makhado1, C. Marais1, J. Maroga6,
M. Mayekiso1, G. McClelland2, J. Mphepya1, D. Muir1, N. Ngcaba1, N. Ngcobo1, J.P. Parkes8, F. Paulsen10,
S. Schoombie2, K. Springer7, C. Stringer7, H. Valentine1, R.M. Wanless2,10 and P.G. Ryan2
Department of Environmental Affairs, Private Bag X4390, Cape Town 8000, South Africa. <GPreston@environment.
gov.za>. 2FitzPatrick Institute of African Ornithology, DST-NRF Centre of Excellence, University of Cape Town,
Rondebosch 7701, South Africa. 3Department of Botany and Zoology, Stellenbosch University, Pvt Bag X1, Matieland
7602, South Africa. 4School of Biological Sciences, Monash University, Melbourne, Victoria 3800, Australia. 5Johan
Heine, Kishugu, 1 Club Street, Old Nelspruit Airport, Nelspruit 1200, South Africa. 6Department of Public Works,
266 Pretorius Street, Centre Walk Building, Pretoria 0001, South Africa. 7Royal Society for the Protection of Birds,
Sandy, Bedfordshire, UK. 8Kurahaupo Consulting, 2 Ashdale Lane, Strowan, Christchurch 8052, New Zealand.
9
Mamont Foundation, St Julian’s Court, St Julian’s Avenue, St Peter Port, GY1 6AX, Guernsey. 10Seabird Conservation
Programme, Birdlife South Africa, PO Box 7119, Cape Town 8012, South Africa.
1
Abstract House mice (Mus musculus) were introduced to South Africa’s sub-Antarctic Marion Island, the larger of the
two Prince Edward Islands, by sealers in the early 19th century. Over the last two centuries they have greatly reduced the
abundance of native invertebrates. Domestic cats (Felis catus) taken to the island in 1948 to control mice at the South
African weather station soon turned feral, killing large numbers of breeding seabirds. An eradication programme finally
removed cats from the island by 1991, in what is still the largest island area cleared of cats at 290 km2. Removal of the
cats, coupled with the warmer and drier climate on the island over the last half century, has seen increasing densities of
mice accumulating each summer. As resources run out in late summer, the mice seek alternative food sources. Marion is
home to globally important seabird populations and since the early 2000s mice have resorted to attacking seabird chicks.
Since 2015 c. 5% of summer-breeding albatross fledglings have been killed each year, as well as some winter-breeding
petrel and albatross chicks. As a Special Nature Reserve, the Prince Edward Islands are afforded the highest degree of
protection under South African environmental legislation. A recent feasibility plan suggests that mice can be eradicated
using aerial baiting. The South African Department of Environmental Affairs is planning to mount an eradication attempt
in the winter of 2021, following a partnership with the Royal Society for the Protection of Birds to eradicate mice on
Gough Island in the winter of 2020. The eradication programme on Marion Island will be spearheaded by the South
African Working for Water programme – Africa’s biggest conservation programme focusing on the control of invasive
species –which is already driving eradication projects against nine other invasive species on Marion Island.
Keywords: albatross, climate change, eradication, Felis catus, invasive species, Mus musculus, petrel, predation
G.R. Preston, B.J. Dilley, J. Cooper, J. Beaumont, L.F. Chauke, S. L. Chown, N. Devanunthan, M. Dopolo, L. Fikizolo, J. Heine, S. Henderson, C.A. Jacobs, F. Johnson, J. Kelly, A.B. Makhado, C.
Marais, J. Maroga, M. Mayekiso, G. McClelland, J. Mphepya, D. Muir, N. Ngcaba, N. Ngcobo, J.P. Parkes, F. Paulsen, S. Schoombie, K. Springer, C. Stringer, H. Valentine, R.M. Wanless and P.G. Ryan
INTRODUCTION
In the late 18th and early 19th century humans
travelled far and wide in the southern oceans to exploit
marine wildlife (Trathan & Reid, 2009). An unfortunate
consequence of this travel was the deliberate or incidental
introduction of alien animal and plant species to distant
environments, causing extensive changes in biological
communities (Mooney & Cleland, 2001). The effects of
invasive species on biodiversity have been described as
“immense, insidious and usually irreversible” (IUCN,
2000). Island ecosystems are highly susceptible to change
and introduced species are the main cause of species
extinctions on islands (Manne, et al., 1999; Chapin, et al.,
2000).
Many seabirds nest on isolated islands that lack land
mammals and consequently one of the major threats to
oceanic seabird species is the introduction of mammalian
predators such as rats (Rattus spp.), domestic cats (Felis
catus) and house mice (Mus musculus) onto their breeding
islands (Croxall, et al., 2012). The devastating impact of
rats on seabird populations breeding on oceanic islands
has been well documented (Atkinson, 1985; Jones, et
al., 2008). However, mice have been introduced to even
more oceanic islands than have rats and, although their
impacts on sub-Antarctic island biota are legion (Angel,
et al., 2009), until recently they were considered to have
little impact on seabird populations (Wanless, et al., 2007;
Jones, et al., 2008).
Sub-Antarctic Marion Island (290 km2, 46°54´S,
37°45´E) is the larger of the two South African Prince
Edward Islands which lie c.2,300 km south-east of
Cape Town in the south-western Indian Ocean (Fig. 1).
As a Special Nature Reserve, established in 1995, the
Prince Edward Islands are afforded the highest degree of
protection under South African environmental legislation
(de Villiers & Cooper, 2008). They also have been a
Wetland of International Importance in terms of the Ramsar
Convention since 2007 (de Villiers, et al., 2011) and are
surrounded by a large (180,000 km²) Marine Protected
Area, declared in 2013, that reaches in places to the edges
of South Africa’s Exclusive Economic Zone (Lombard, et
al., 2007; Nel & Omardien, 2008).A revised management
plan adopted in 2014 guides and controls activities at the
island group, including biosecurity protocols to avoid alien
introductions (DST-NRF Centre of Excellence for Invasion
Biology, 2014).
The Prince Edward Islands currently support breeding
populations of 29 species of birds, all but two of which
probably breed on Marion Island (Ryan & Bester, 2008;
Peter Ryan, FitzPatrick Institute unpubl. data; Table 1).
Eight bird species of the order Procellariiformes that breed
on Marion are listed by the International Agreement on the
Conservation of Albatrosses and Petrels, to which South
Africa is a founding signatory (Cooper, et al., 2006).These
four albatross and four petrel (Macronectes and Procellaria)
In:
40C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 40–46. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Preston, et al.: Eradicating mice from Marion Island
common than at nearby predator-free Prince Edward Island.
A sustained eradication programme that commenced in the
mid-1970s had finally eradicated cats from the island by
1991 (Bester, et al., 2002), in what until recently was the
largest island area cleared of cats.
We give an overview of the adverse impacts of mice
on Marion Island’s biota and ecosystem and discuss the
mouse eradication attempt planned for the austral winter
of 2020.
A syndrome of adverse factors
House mice have been present on Marion Island
for nearly 200 years (Berry, et al., 1978), significantly
disrupting terrestrial ecosystem functioning (Chown &
Smith, 1993). The mice may be seen as part of a syndrome
of interacting factors (Parkes, 2016) having adverse
impacts on native invertebrates, plants and seabirds (e.g.
Phiri, et al., 2009; Angel & Cooper, 2011). The mice have
changed the state of Marion Island’s ecosystems compared
with the near-pristine condition of neighbouring Prince
Edward Island (45 km2, see Fig. 1).
Fig. 1 South Africa’s Prince Edward Islands (46°54´S,
37°45´E) lie 2,300 km south-east of Cape Town in
the south-western Indian Ocean. Marion Island has
introduced house mice, but Prince Edward Island, 22 km
to the north-east, remains free of introduced mammals.
species are at risk at sea to bycatch by commercial fisheries,
especially longlining, and are considered threatened
or near threatened at regional (Taylor, et al., 2015) and
global (BirdLife International, 2017) levels. Marion Island
supports about 25% of the world’s breeding population of
wandering albatrosses Diomedea exulans (globally and
regionally Vulnerable), 12% of sooty (Phoebetria fusca)
and 7% of grey-headed (Thalassarche chrysostoma)
albatrosses (both globally and regionally Endangered)
and smaller percentages of light-mantled albatrosses (P.
palpebrata) (globally and regionally Near Threatened)
and grey petrels (Procellaria cinerea) (globally Near
Threatened and regionally Vulnerable).
We show that these five ACAP-listed species, along
with the regionally Near Threatened great-winged
petrel (Pterodroma macroptera), are at serious risk to
predation from introduced mice on Marion Island. Mice
were accidentally introduced to Marion Island during
the sealing era sometime before 1818 and were the sole
introduced mammal until 1948 when five domestic cats
were introduced to control mice at the newly-established
weather station (Watkins & Cooper, 1986). However, even
in the 1950s, little was known about the potential harmful
effects of invasive species on islands. Rand (1954) was the
biologist on the Eighth South African Expedition to Marion
Island over 1951/52 and noted how “a few domestic cats
have gone feral and prey on the smaller petrels or mice
that are widespread over the coastal plain” (p. 178) and
“mice often burrow into the [albatross] nest cone but do
no appreciable damage” (p. 189). Unfortunately, the cats
preferred to eat the island’s native birds, especially the
burrow-nesting petrels, and by the 1970s more than 2,000
cats were killing some 450,000 birds each year (van Aarde,
1980). As a result, at least one species, the common diving
petrel (Pelecanoides urinatrix), disappeared from the
island and all the other burrowing petrels became far less
For more than 30 years the burrowing petrel
populations on Marion Island were impacted by cats (top
predators) and mice (mesopredators). Whereas mice target
eggs and chicks (Fugler, et al., 1987; Dilley, et al., 2015;
Dilley, et al., 2018), reducing reproductive success, cat
predation was far more detrimental because they killed
chicks and adults, affecting both reproduction and adult
survival (Le Corre, 2008). Removal of the top predator
on Marion Island has benefited adult survival but may
have triggered a ‘mesopredator release’ effect (Zavaleta,
et al., 2001; Le Corre, 2008), whereby mouse numbers
expanded, increasing their impact on petrel populations
(Rayner, et al., 2007). The dramatic decrease in burrowing
petrel populations at Marion Island caused by the cats is
again presumed to have adversely affected key ecological
processes driven by burrowing petrels such as soil turnover and marine nutrient imports (Caut, et al., 2012).
Mouse numbers cycle seasonally on Marion Island,
linked partly to changes in the abundance of invertebrates
and seeds. Mouse densities are highest in autumn, when
breeding ceases, and are lowest in early summer, before
breeding resumes. Invertebrate biomass also changes
seasonally, but not to the same extent, so that the per capita
food supply (from macro invertebrates as the primary food
of the mice) was estimated to be 3.4 kg/ha and 3.6 kg/ha
in early summer but only 0.4 kg/ha and 0.2 kg/ha in early
winter in the Biotic and Mire habitats favoured by mice,
respectively (Avenant & Smith, 2003).
Peak mouse densities occur in April–May, and have
increased between 1990 and 2008, driven in part by a
warmer, drier climate (Ferreira, et al., 2006; le Roux &
McGeoch, 2008; McClelland, et al., 2018). By comparison,
invertebrate biomass has decreased >80% since the late
1970s (McClelland, et al., 2018). Since 2015, there has
been a marked increase in the frequency of mice attacking
surface-breeding seabird chicks (Dilley, et al., 2016a) and
if invertebrate biomass continues to decline, the impact of
mouse predation on Marion’s seabird chicks is likely to
become even more serious.
Overview of mice attacking seabirds at Marion Island
The first signs of mouse attacks on seabirds at Marion
Island were recorded in 2003, when wandering albatross
chicks were observed with rump wounds typical of those
inflicted by mice on Tristan albatross (D. dabbenena)
chicks on Gough Island (Jones & Ryan, 2010; Table 2).
The first attacks on summer-breeding albatross chicks were
41
Island invasives: scaling up to meet the challenge. Ch 1A Rodents: Planning
Table 1 Estimated risk of local extirpation of bird species currently known or thought to breed on Marion Island if the mice
are not eradicated.
Grey-backed storm petrel Garrodia nereis*
?1
Known or
considered
vulnerable to
predation
yes
Black-bellied storm petrel Fregetta tropica*
?1
yes
Species
Estimated numbers
of breeding
pairs
8002
<52
5,0002
1,0001
50–1002
14,0002
3003
1,4653
7,9001
1,8001
1,0001
150,0002
145,0004
5,0001
24,0005
251
501
1,7501
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
uncertain
Northern giant petrel Macronectes halli
4001
uncertain
Crozet shag Leucocarbo melanogenis
270
uncertain
Brown skua Catharacta antarctica
300
uncertain
Kelp gull Larus dominicanus
100
uncertain
Lesser sheathbill Chionis minor
700
uncertain
Grey petrel Procellaria cinerea
Cape petrel Daption capense
Kerguelen petrel Lugensa brevirostris
South Georgian diving petrel Pelecanoides georgicus
Common diving petrel Pelecanoides urinatrix
Great-winged petrel Pterodroma macroptera
Light-mantled albatross Phoebetria palpebrata
Sooty albatross Phoebetria fusca
Grey-headed albatross Thalassarche chrysostoma
Wandering albatross Diomedea exulans
Fairy prion Pachyptila turtur
Salvin’s prion Pachyptila salvini
Blue petrel Halobaena caerulea
Soft-plumaged petrel Pterodroma mollis
White-chinned petrel Procellaria aequinoctialis
Antarctic tern Sterna vittata
Kerguelen tern Sterna virgata
Southern giant petrel Macronectes giganteus
King penguin Aptenodytes patagonicus
Gentoo penguin Pygoscelis papua
Macaroni penguin Eudyptes chrysolophus
Southern rockhopper penguin Eudyptes chrysocome
1
6
1
1
220,0001
no
9001
no
370,0001
no
67,0001
no
Estimated years
to local
extirpation
possibly locally
extirpated
possibly locally
extirpated
30
30
50
50
50
50–100
50–100
50–100
50–100
50–100
50–100
50–100
50–100
50–100
50–100
50–100
50–100
*Current breeding not proven but suspected
Data sources: 1Ryan & Bester (2008); 2FitzPatrick unpubl. data; 3Schoombie et al., (2016); 4Dilley et al., (2017); 5Ryan et al., (2012);
6
Ryan et al., (2009)
recorded in April 2009 when sooty albatross fledglings
at two colonies were found ‘scalped’ with raw, bleeding
crowns and necks (Jones & Ryan, 2010; Fig. 2). Mice were
suspected of being responsible for these wounds (Jones
& Ryan, 2010), even though summer-breeding albatross
chicks are seldom attacked by mice on Gough Island
(Cuthbert, et al., 2013). Another sooty albatross fledgling
was attacked in 2010 at one of the colonies where scalpings
occurred in 2009 (Ben Dilley, FitzPatrick Institute unpubl.
data), but no further attacks were recorded until 2015,
when mice attacked large chicks of all three albatross
species that fledge in autumn: grey-headed (Fig. 3), sooty
and light-mantled albatrosses (Dilley, et al., 2016a, Table
2). Filming at night confirmed that mice were responsible
42
for these wounds, with most affected chicks dying within a
few days of being attacked (Dilley, et al., 2016a). Attacks
started independently in small pockets all around the
island’s 70 km coastline, separated by distances hundreds
of times greater than mouse home ranges (Wanless, et al.,
2008; Dilley, et al., 2016a; Fig. 2). In 2015, three of the
six mouse-injured wandering albatross chicks had head
wounds (‘scalpings’, see Fig. 4). In 2016, 2017 and 2018
mouse attacks continued on summer-breeding albatross
fledglings, indicating that the sudden increase in 2015 was
not a one-off event.
With cats having been eradicated from Marion Island
by 1991, we expected burrowing petrel populations to
Preston, et al.: Eradicating mice from Marion Island
Fig. 2 Marion Island showing the locations of albatross
breeding colonies (sooty albatross = dark grey shade;
light-mantled albatross = light grey shade; all greyheaded albatross colonies are along Grey-headed
Albatross Ridge and Rook’s Peninsula) and mouseattack sites from 2009–2015 (adapted from Dilley, et al.,
2016a). Contour lines indicate 100 m.
have recovered by two decades later. Initial indications
were positive; following the removal of cats there were
marked increases in the breeding success of burrowing
petrels, especially great-winged petrels that breed in winter
when cat predation pressure was most severe (Cooper &
Fourie, 1991; Cooper, et al., 1995). However, the postcat recovery of burrowing petrel numbers on Marion
has been much slower than anticipated, especially for
smaller species (Dilley, et al., 2016b). Of the nine species
of burrowing petrels breeding on Marion Island, the two
smallest species (black-bellied (Fregetta tropica) and
grey-backed storm petrels (Garrodia nereis)) are now very
uncommon and are likely locally extirpated on the island
due to mice (Dilley, et al., 2016b). Recent evidence from a
repeat survey of burrow densities (Dilley, et al., 2016b) and
from analyses of brown skua Catharacta antarctica prey
remains (Cerfonteyn & Ryan, 2016) both suggest there has
been little or no recovery of burrowing petrel populations
at Marion since cats were eradicated.
Predation by mice is the most likely explanation for the
limited recovery of Marion’s petrel populations (Dilley,
et al., 2016b). Recent evidence from breeding success
studies shows that mice are suppressing the recovery of
burrowing petrel populations, especially those that breed
in winter, through predation on eggs and chicks (Dilley,
et al., 2018). Winter breeders had lower breeding success
than did summer breeders, with most fatalities being of
small chicks <14 days old. Mice were filmed attacking
and killing chicks of two winter-breeding species:
Fig. 3 Grey-headed albatross chicks showing distinctive
‘scalping’ wounds inflicted by mice on Marion Island in
2015 (photo Ben Dilley).
grey petrel (three of 18 nests filmed; <https://youtu.be/
Og1d6a2cmXQ>) and great-winged petrel (one of 19;
<https://youtu.be/D9vPoFsjvgs>, Dilley, et al., 2018).
Grey petrel chicks, which had the highest mortality rate,
hatch in early winter when mouse densities are still fairly
high, but food availability is low, resulting in the lowest
seasonal per capita food availability for mice (Dilley, et
al., 2018). Most grey petrel mortalities occurred when
chicks were <7 days old, and were likely due to mouse
predation (Dilley, et al.,2018).
We conclude that mice are currently suppressing
the recovery of burrowing petrel populations on Marion
Island, especially those that breed in winter, through
predation on eggs and chicks. The widespread increase
in mouse predation on albatross chicks at Marion in 2015
is also a cause for concern. Left uncontrolled, it is feared
that 18 of the 28 species breeding on Marion Island may
be vulnerable to local extirpation (see Table 1), should the
mice not be eradicated.
PLAN OF ACTION
The Prince Edward Islands are recognised as a Special
Nature Reserve, which affords the highest degree of
protection under South African environmental legislation,
and the islands’ management plan aims to eradicate alien
plants and animals as far as possible (DST-NRF Centre of
Excellence for Invasion Biology, 2014). As summarised
above, the structure of Marion Island’s terrestrial ecosystem
has been radically transformed by introduced mice, which
are now threatening the island’s globally important seabird
Table 2 Summary of mouse attacks on surface-nesting seabirds breeding on Marion Island (from Dilley, et al.,
2016a and FitzPatrick Institute unpubl. data).
Species
Wandering albatross
Sooty albatross
Light-mantled albatross
Grey-headed albatross
Year of first attack
2003
2009
2015
2015
Maximum number attacked % of annual production
6
0.8%
45
4.3%
1
4.0%
102
4.6%
43
Island invasives: scaling up to meet the challenge. Ch 1A Rodents: Planning
Fig. 4 A Wandering albatross chick being scalped by a
mouse on Marion Island in the winter of 2015 (photo
Stefan Schoombie).
populations. Given the island’s importance as a breeding
site for threatened albatrosses and other seabird species
that are being killed by mice, there is an urgent need to
eradicate mice from the island. A detailed feasibility plan
(Parkes, 2016) suggests that mice can be eradicated using
aerial baiting. This follows the now well-established
approach of using helicopters fitted with GPS guidance
systems and under slung bait-distribution buckets to spread
brodifacoum-laced pellets across the entire island over a
relatively short period, to ensure that all rodents have
access to the poison bait. Such operations, pioneered on
New Zealand’s offshore islands, have a good track record
in recent years with 21 of 22 operations around the world
targeting mice being successful in the last decade (DIISE,
2015). However, the operation on Marion Island will be
an order of magnitude larger than any previous island
eradication targeting mice only (cf. Springer, 2016; Martin
& Richardson, 2017). This will require the deployment of
poison bait with a high level of accuracy given the small
home ranges of mice relative to rats (Parkes, 2016).
The South African Department of Environmental
Affairs is planning to mount an eradication attempt
on Marion Island in the austral winter of 2021. This is
timed to follow a planned eradication of mice on Gough
Island led by the United Kingdom’s Royal Society for the
Protection of Birds in the winter of 2020. Gough Island,
part of the UK Overseas Territory of St Helena, Ascension
and Tristan da Cunha, is one of the world’s most important
seabird breeding islands. It is the site where mice were first
appreciated to pose a significant risk to breeding seabirds
(Cuthbert & Hilton 2004; Wanless, et al., 2007), and
experiences very high levels of chick mortality in several
species, including the Tristan albatross (globally Critically
Endangered), Atlantic petrel (Pterodroma incerta)
(Endangered) and Macgillivray’s prion (Pachyptila
macgillivrayi) (Endangered) (Davies, et al., 2015; Dilley,
et al., 2015). Despite these impacts, the island still supports
some 12 million breeding seabirds of 22 species and is
regarded as a top-priority island for rodent eradication
world-wide (Hilton & Cuthbert, 2010).
At 65 km2, Gough will be the largest island where an
eradication has been attempted targeting mice alone (mice
were eradicated from 129 km2 Macquarie Island (Australia),
but they occurred at lower densities than on Marion due to
the presence of black rats (R. rattus) on the island (Springer
2016;
<http://www.parks.tas.gov.au/?base=13013>).
Planning for the Gough Island eradication has involved
more than a decade of research to ensure the highest
probability of success (e.g. Angel & Cooper, 2006; Brown,
2007; Parkes, 2008; Wanless, et al., 2009; Cuthbert, et
44
al., 2011a; Cuthbert, et al., 2011b; Cuthbert, et al., 2014;
Cuthbert, et al., 2016). At 290 km2, Marion Island is almost
five times larger than Gough Island, but the terrain is less
rugged, and the presence of a largely un-vegetated interior
above 800 m with few, if any, mice in winter makes an
eradication attempt at Marion less challenging in some
regards (Parkes, 2016). The intention is to commence the
operation during early winter, when mouse numbers are
falling due to lack of food and cold conditions, increasing the
likelihood of all animals consuming bait (see Parkes, 2019,
for further details on the crucial decision of ‘when to bait’
on Marion). Mice also cease breeding on Marion from late
May to August, reducing the chances of semi-independent
young in the den failing to encounter bait (Parkes, 2016).
Winter also coincides with the period of lowest numbers of
brown skuas and giant petrels (Macronectes spp.) present
on the island, which might be killed accidentally by either
primary or secondary poisoning.
Mitigation plans will be needed to reduce the impacts
on resident scavenging species (Wanless, et al., 2010).
At this stage, the intention is to keep approximately 100
lesser sheathbills (Chionis minor) in captivity during the
eradication attempt, given the moderate level of mortality of
snowy sheathbills (C. albus) during the rodent eradication
at South Georgia (Martin & Richardson, 2017). The Prince
Edward Islands are home to an endemic subspecies of
sheathbill C. m. marionensis, but nearby Prince Edward
Island houses a substantial population of this subspecies and
could be used to re-establish birds on Marion Island. Kelp
gulls (Larus dominicanus) also are resident scavengers at
Marion Island, but they may be less susceptible to nontarget poisoning (Martin & Richardson, 2017). Given the
small population size (Table 1) and difficulty of catching
and maintaining captive birds, there is currently no plan
to mitigate impacts on this species. Gulls are thought to
move freely between Marion and Prince Edward Island,
so immigration should aid the recovery of the Marion
population after the eradication.
The eradication on Marion Island was stimulated by
the donation of US$100,000 and the three helicopters
used in the South Georgia rodent eradication by the
Mamont Foundation to the South African Department of
Environmental Affairs in early 2017. South Africa has
a weather station on Gough Island, and will assist this
eradication effort through the provision of accommodation
(including possible refurbishments on the island), the
hosting of the eradication team from its Cape Town
harbour, and assistance with transportation. In return,
the equipment used and expertise developed during the
Gough eradication will be transferred to South Africa for
use in the planned Marion eradication. The programme
on Marion Island will be spearheaded by the Department
of Environmental Affairs’ Working for Water programme
– Africa’s biggest conservation programme focusing
on the control of invasive species. Working for Water
is already managing eradication projects against eight
invasive vascular plant species on Marion Island, and the
possible eradication of one introduced invertebrate, the
rough woodlouse (Porcellio scaber), is being assessed (D.
Muir). The South African Government is budgeting for
this programme (with an initial budget of about US$2.2
million). It will seek to raise co-funding, including through
a crowd-funding initiative being led by BirdLife South
Africa, a non-governmental organisation.
Eradicating rodents from islands is an effective, longterm conservation management action, provided robust
biosecurity measures are put in place to minimise the
likelihood of any reintroductions. The South African
National Antarctic Programme has imposed stringent
quarantine measures on all vessels and materials going to
the Prince Edward Islands (and Gough Island) since the
Preston, et al.: Eradicating mice from Marion Island
early 1990s. These include fumigation of the resupply
vessel prior to each voyage, use of rat guards on all
hawsers when in harbour, placement of rodenticide baits
at strategic points throughout the ship, and inspection of
all cargo before being opened ashore (DST-NRF Centre of
Excellence for Invasion Biology, 2014).
Cooper, J., Baker, G.B., Double, M.C., Gales, R., Papworth, W., Tasker,
M. and Waugh, S.M. (2006). ‘The Agreement on the Conservation
of Albatrosses and Petrels: Rationale, history, progress and the way
forward’. Marine Ornithology 34: 1–5.
ACKNOWLEDGEMENTS
Cuthbert, R. and Hilton, G. (2004). ‘Introduced house mice Mus musculus:
a significant predator of threatened and endemic birds on Gough Island,
South Atlantic Ocean?’ Biological Conservation 117: 483–489.
Logistical and financial support for scientific research
on Marion Island was provided by the South African
Department of Environmental Affairs through the South
African National Antarctic Programme (SANAP), the
National Research Foundation, the University of Cape Town
and the Agreement on the Conservation of Albatrosses and
Petrels. The support of BirdLife International, BirdLife
South Africa, the Mamont Foundation, the Royal Society
for the Protection of Birds and the FitzPatrick Institute of
African Ornithology at the University of Cape Town, is
gratefully acknowledged.
REFERENCES
Angel, A., Wanless, R.M. and Cooper, J. (2009). ‘Review of the impacts
of house mice on islands’. Biological Invasions 11: 1743–1754.
Angel, A. and Cooper, J. (2006). A Review of the Impacts of Introduced
Rodents on the Islands of Tristan da Cunha and Gough. Sandy, UK:
Royal Society for the Protection of Birds.
Angel, A. and Cooper, J. (2011). A Review of the Impacts of the House
Mouse Mus musculus on Sub-Antarctic Marion Island, Prince Edward
Islands. Report to the Prince Edward Islands Management Committee,
South African National Antarctic Programme.
Atkinson, I.A.E. (1985). ‘The Spread of Commensal Species of Rattus
to Oceanic Islands and their Effects on Island Avifaunas’. In: P.J.
Moors (ed.) Conservation of Island Birds, pp. 35–81. ICBP Technical
Publication 3. Cambridge, UK: International Council for Bird
Preservation.
Croxall, J.P., Butchart, S.H., Lascelles, B., Stattersfield, A.J., Sullivan,
B., Symes, A. and Taylor, P. (2012). ‘Seabird conservation status,
threats and priority actions: a global assessment’. Bird Conservation
International 22: 1–34.
Cuthbert, R.J., Visser, P., Louw, H. and Ryan, P.G. (2011a). ‘Palatability
and efficacy of rodent baits for eradicating house mice (Mus musculus)
from Gough Island, Tristan da Cunha’. Wildlife Research 38: 196–203.
Cuthbert, R.J., Parker, G., Louw, H., Rexer-Huber, K., Parker, G. and
Ryan, P.G. (2011b). ‘Preparations for the eradication of mice from
Gough Island: results of bait acceptance trials above ground and around
cave systems’. In: C.R. Veitch, M.N. Clout and D.R. Towns (eds.)
Island invasives: eradication and management, pp. 47–50. Occasional
Paper SSC no. 42. Gland, Switzerland: IUCN and Auckland, New
Zealand: CBB.
Cuthbert, R.J., Louw, H., Parker, G., Rexer-Huber, K. and Visser,
P. (2013). ‘Observations of mice predation on dark-mantled sooty
albatross and Atlantic yellow-nosed albatross chicks at Gough Island’.
Antarctic Science 25: 763–766.
Cuthbert, R.J., Broome, K., Bradley, J. and Ryan, P.G. (2014). ‘Evaluating
the effectiveness of aerial baiting operations for rodent eradications on
cliffs on Gough Island, Tristan da Cunha’. Conservation Evidence 11:
25–28.
Cuthbert, R.J., Wanless, R.M., Angel, A., Burlé, M.-H., Hilton, G.M.,
Louw, H., Visser, P., Wilson, J. and Ryan, P.G. (2016). ‘Drivers of
predatory behaviour and extreme body size in house mice Mus musculus
on Gough Island’. Journal of Mammalogy 97: 533–544.
Davies, D., Dilley, B.J., Bond, A.L., Cuthbert, R.J. and Ryan, P.G. (2015).
‘Trends and tactics of mouse predation on Tristan Albatross Diomedea
dabbenena chicks at Gough Island, South Atlantic Ocean’. Avian
Conservation and Ecology 10(1): 5 (doi: 10.5751/ACE-00738-100105).
Avenant, N.L. and Smith, V.R. (2003). ‘The microenvironment of house
mice on Marion Island (sub-Antarctic)’. Polar Biology 26: 129–141.
de Villiers, M.S. and Cooper, J. (2008). ‘Conservation and Management’.
In: S.L Chown and P.W. Froneman (eds). The Prince Edward Islands:
Land–Sea Interactions in a Changing Ecosystem, pp. 113–131.
Stellenbosch, South Africa: SunPress.
Berry, R.J., Peters, J. and van Aarde, R.J. (1978). ‘Sub-antarctic house
mice: colonization, survival and selection’. Journal of Zoology, London
184: 127–141.
de Villiers, M.S., Chown, S.L. and Cooper, J. (2011). Prince Edward
Islands Conservation Handbook. Keeping your Footprint at the Islands
Small. Stellenbosch, South Africa: SunPress.
Bester, M.N., Bloomer, J.P., van Aarde. R.J., Erasmus, B.H., van
Rensburg, P.J.J., Skinner, J.D., Howell, P.G. and Naude, T.W. (2002).
‘A review of the successful eradication of feral cats from sub-Antarctic
Marion Island, southern Indian Ocean’. South African Journal of
Wildlife Research 32: 65–73.
Dilley, B.J., Davies, D., Bond, A.L. and Ryan, P.G. (2015). ‘Effects of
mouse predation on burrowing petrel chicks at Gough Island’. Antarctic
Science 27: 543–553.
BirdLife International (2017). ‘IUCN Red List for Birds’. <http://www.
birdlife.org>. Accessed 3 August 2017.
Brown, D. (2007). A Feasibility Study for the Eradication of Rodents from
Tristan da Cunha. Sandy, UK: Royal Society for the Protection of Birds.
Caut, S., Angulo, E., Pisanu, B., Ruffino, L., Faulquier, L., Lorvelec, O.,
Chapuis, J.-L., Pascal, M., Vidal, E. and Courchamp, F. (2012). ‘Seabird
modulations of isotopic nitrogen on islands’. PLOS ONE 7: e39125.
Cerfonteyn, M. and Ryan, P.G. (2016). ‘Have burrowing petrels recovered
on Marion Island two decades after cats were eradicated? Evidence
from sub-Antarctic skua prey remains’. Antarctic Science 28: 51–57.
Chapin, F.S.; Zavaleta, E.S., Viner, V.T., Naylor, R.L., Vitousek, P.M.,
Sala, O.E., Reynolds, H.L., Hooper, D.U., Mack, M., Diaz, S.E., Hobbie,
S.E. and Lavorel, S. (2000). ‘Consequences of changing biodiversity’.
Nature 405: 234–242.
Chown, S.L. and Smith, V.R. (1993). ‘Climate change and the short-term
impact of feral house mice at the sub-Antarctic Prince Edward Islands’.
Oecologia 96: 508–516.
Cooper, J. and Fourie, A. (1991). ‘Improved breeding success of greatwinged petrels Pterodroma macroptera following control of feral
cats Felis catus at sub-Antarctic Marion Island’. Bird Conservation
International 1: 171–175.
Cooper, J., Marais, A.v.N., Bloomer, J.P. and Bester, M.N. (1995). ‘A
success story: breeding of burrowing petrels (Procellariidae) before
and after the extinction of feral cats Felis catus at sub-Antarctic Marion
Island’. Marine Ornithology 23: 33–37.
Dilley, B.J., Schoombie, S., Schoombie, J. and Ryan, P.G. (2016a).
‘‘Scalping’ of albatross fledglings by introduced mice spreads rapidly at
Marion Island’. Antarctic Science 28: 73–80.
Dilley, B.J., Schramm, M. and Ryan, P.G. (2016b). ‘Modest increases in
densities of burrow-nesting petrels following the removal of cats Felis
catus from Marion Island’. Polar Biology 40: 25–637.
Dilley, B.J., Davies, D., Schramm, M., Connan, M. and Ryan, P.G. (2017).
‘The distribution and abundance of Blue Petrels (Halobaena caerulea)
breeding at subantarctic Marion Island, Emu– Austral Ornithology 117:
222–232.
Dilley, B.J., Schoombie, S., Stevens, K., Davies, D., Perold, V., Osborne,
A., Schoombie, J., Brink, C.W., Carpenter-Kling, T. and Ryan, P.G.
(2018). ‘Mouse predation affects breeding success of burrow-nesting
petrels at sub-Antarctic Marion Island’. Antarctic Science 30: 93–104.
DIISE (2015). ‘The Database of Island Invasive Species Eradications,
developed by Island Conservation, Coastal Conservation Action
Laboratory UCSC, IUCN SSC Invasive Species Specialist Group,
University of Auckland and Landcare Research New Zealand’. <http://
diise.islandconservation.org>. Accessed 3 August 2017.
DST-NRF Centre of Excellence for Invasion Biology (2014). Prince
Edward Islands Management Plan Version 0.2. Pretoria, South Africa:
Department of Environmental Affairs.
Ferreira, S.M., van Aarde, R.J. and Wassenaar, T.D. (2006). ‘Demographic
responses of house mice to density and temperature on sub-Antarctic
Marion Island’. Polar Biology 30: 83–94.
Fugler, S.R., Hunter, S., Newton, I.P. and Steele, W.K. (1987). ‘Breeding
biology of blue petrels Halobaena caerulea at the Prince Edward
Islands’. Emu 87: 103–110.
45
Island invasives: scaling up to meet the challenge. Ch 1A Rodents: Planning
Hilton, G.M. and Cuthbert, R.J. (2010). ‘The catastrophic impact of
invasive mammalian predators on birds of the UK Overseas Territories:
a review and synthesis’. Ibis 152: 443–458.
IUCN (2000). Guidelines for the Prevention of Biodiversity Loss Due to
Biological Invasion. Occasional Paper SSC no. 62. Gland, Switzerland:
IUCN.
Jones, H.P., Tershy, B.R., Zavaleta, E.S., Croll, D.A., Keitt, B.S.,
Finkelstein, M.E. and Howald, G.R. (2008). ‘Severity of the effects of
invasive rats on seabirds: a global review’. Conservation Biology 22:
16–26.
Jones, M.G.W. and Ryan, P.G. (2010). ‘Evidence of mouse attacks on
albatross chicks on sub-Antarctic Marion Island’. Antarctic Science 22:
39–42.
Le Corre, M. (2008). ‘Conservation biology: cats, rats and seabirds’.
Nature 451: 134–135.
le Roux, P.C. and McGeoch, M.A. (2008). ‘Changes in climate extremes,
variability and signature on sub-Antarctic Marion Island’. Climatic
Change 86: 309–329.
Lombard, A.T., Reyers, B., Schonegevel. L.Y., Cooper, J., Smith-Adao,
L.B., Nel, D.C., Froneman, P.W., Ansorge, I.J., Bester, M.N., Tosch,
C.A., Strauss, T., Akkers, T., Gon, O., Leslie. R.W. and Chown, S.L.
(2007). ‘Conserving pattern and process in the Southern Ocean:
designing a Marine Protected Area for the Prince Edward Islands’.
Antarctic Science 19: 39–54.
Schoombie, S., Crawford, R.J.M., Makhado, A.B., Dyer, B.M. and Ryan,
P.G. (2016). ‘Recent population trends of sooty and light-mantled
albatrosses breeding on Marion Island’. African Journal of Marine
Science 38:119–127.
Springer, K. (2016). ‘Methodology and challenges of a complex multispecies eradication in the sub-Antarctic and immediate effects of
invasive species removal’. New Zealand Journal of Ecology 40: 273–
278.
Taylor, M.R., Peacock, F. and Wanless, R.M. (2015). The 2015 ESKOM
Red Data Book of Birds of South Africa, Lesotho and Swaziland.
Dunkeld West, South Africa: BirdLife South Africa.
Trathan, P.N. and Reid, K. (2009). ‘Exploitation of the marine ecosystem
in the sub-Antarctic: historical impacts and current consequences’.
Papers and Proceedings of the Royal Society of Tasmania 143: 9–14.
van Aarde, R.J. (1980). ‘The diet and feeding behaviour of feral cats, Felis
catus at Marion Island’. South African Journal of Wildlife Research 10:
123–128.
Wanless, R.M., Angel, A., Cuthbert, R.J., Hilton, G.M. and Ryan, P.G.
(2007). ‘Can predation by invasive mice drive seabird extinctions?’
Biology Letters 3: 241–244.
Wanless, R.M., Fisher, P., Cooper, J., Parkes. J. and Ryan, P.G. (2008).
‘Bait acceptance by house mice: an island field trial’. Wildlife Research
35: 806–811.
McClelland, G.T.W., Altwegg, R., van Aarde, R.J., Ferreira, S., Burger,
A.E. and Chown, S.L. (2018). ‘Climate change leads to increasing
population density and impacts of a key island invader’. Ecological
Applications 28: 212–224.
Wanless, R.M., Ryan, P.G., Altwegg, R., Angel, A., Cooper, J., Cuthbert,
R. and Hilton, G.M. (2009). ‘From both sides: dire demographic
consequences of carnivorous mice and longlining for the critically
endangered Tristan albatrosses on Gough Island’. Biological
Conservation 142: 1710–1718.
Manne, L.L., Brooks, T.M. and Pimm, S.L. (1999). ‘Relative risk of
extinction of passerine birds on continents and islands’. Nature 399:
258–261.
Wanless, R.M., Cooper, J., Slabber, M.J., and Ryan, P.G. (2010). ‘Risk
assessment of birds foraging terrestrially at Marion and Gough Islands
to primary and secondary poisoning’. Wildlife Research 37: 524–530.
Martin, A.R. and Richardson, M.G. (2017). Rodent eradication scaled
up: clearing rats and mice from South Georgia. Oryx: doi:10.1017/
S003060531700028X.
Watkins, B.P. and Cooper, J. (1986). ‘Introduction, present status and
control of alien species at the Prince Edward Islands, sub-Antarctic’.
South African Journal of Antarctic Research 16: 86–94.
Mooney, H.A. and Cleland, E.E. (2001). ‘The evolutionary impact of
invasive species’. Proceedings of the National Academy of Sciences
USA 98: 5446–5451.
Zavaleta, E.S., Hobbs, R.J. and Mooney, H.A. (2001). ‘Viewing invasive
species removal in a whole-ecosystem context’. Trends in Ecology and
Evolution 16: 454–459.
Nel, D. and Omardien, A. (eds). (2008). Towards the Development of
a Marine Protected Area at the Prince Edward Islands. WWF South
Africa Report Series 2008/Marine/001. South Africa: WWF Sanlam
Living Waters Partnership.
Parkes, J. (2008). A Feasibility Study for the Eradication of House Mice
from Gough Island. RSPB Research Report No.34. Sandy, UK: Royal
Society for the Protection of Birds.
Parkes, J. (2016). ‘Eradication of House Mice Mus musculus from Marion
Island: A Review of Feasibility, Constraints and Risks’. In: R.M.
Wanless (ed). BirdLife South Africa Occasional Report Series No. 1.
Johannesburg, South Africa: BirdLife South Africa.
Parkes, J.P. (2019). ‘Timing aerial baiting for rodent eradications on cool
temperate islands: mice on Marion Island’. In: C.R. Veitch, M.N. Clout,
A.R. Martin, J.C. Russell and C.J. West (eds.) Island invasives: scaling
up to meet the challenge, pp. 36–39. Occasional Paper SSC no. 62.
Gland, Switzerland: IUCN.
Phiri, E.E., McGeoch, M.A. and Chown, S.L. (2009). ‘Spatial variation
in structural damage to a keystone plant species in the sub-Antarctic:
interactions between Azorella selago and invasive house mice’.
Antarctic Science 21: 189–196.
Rand, R.W. (1954). ‘Notes on the birds of Marion Island’. Ibis 96: 173–
206.
Rayner, M.J., Hauber, M.E., Imber, M.J., Stamp, R.K. and Clout, M.N.
(2007). ‘Spatial heterogeneity of mesopredator release within an oceanic
island system’. Proceedings of the National Academy of Sciences USA
104: 20862–20865.
Ryan, P.G. and Bester, M.N. (2008). ‘Pelagic Predators’. In: S.L. Chown
and P.W. Froneman (eds). The Prince Edward Islands: Land–Sea
Interactions in a Changing Ecosystem, pp. 121–164. Stellenbosch,
South Africa: Sun Press.
Ryan, P.G., Whittington, P.A. and Crawford, R.J.M. (2009). ‘A tale of
two islands: contrasting fortunes for sub-Antarctic skuas at the Prince
Edward Islands’. African Journal of Marine Science 31: 431–437.
Ryan, P.G., Dilley, B.J. and Jones, M.G.W. (2012). ‘The distribution
and abundance of white-chinned petrels (Procellaria aequinoctialis)
breeding at the sub-Antarctic Prince Edward Islands’. Polar Biology
35: 1851–1859.
46
E. Rojas-Mayoral, F.A. Méndez-Sánchez, B. Rojas-Mayoral and A. Aguirre-Muñoz
Rojas-Mayoral, E.; F.A. Méndez-Sánchez, B. Rojas-Mayoral and A. Aguirre-Muñoz. Improving the efficiency of
aerial rodent eradications by means of the numerical estimation of rodenticide density
Improving the efficiency of aerial rodent eradications by means of the
numerical estimation of rodenticide density
E. Rojas-Mayoral, F.A. Méndez-Sánchez, B. Rojas-Mayoral and A. Aguirre-Muñoz
Grupo de Ecología y Conservación de Islas, A.C. Ave. Moctezuma 836, Zona Centro C.P. 22800, Ensenada,
Baja California. <evaristo.rojas@islas.org.mx>.
Abstract Invasive rodents are present on approximately 80% of the world’s islands and constitute one of the most serious
threats to island biodiversity and ecosystem functioning. The eradication of rodents is central to island conservation
efforts and the aerial broadcast of rodenticide bait is the preferred dispersal method. To improve the efficiency of rodent
eradication campaigns, the generation of accurate and real-time bait density maps is required. Creating maps to estimate
the spatial dispersion of bait on the ground has been carried out using traditional GIS methodologies, which are based
on limiting assumptions and are time intensive. To improve accuracy and expedite the evaluation of aerial operations,
we developed an algorithm for the numerical estimation of rodenticide density (NERD). The NERD algorithm performs
calculations with increased accuracy, displaying results almost in real-time. NERD describes the relationship between bait
density, the mass flow rate of rodenticide through the bait bucket, and helicopter speed and produces maps of bait density
on the ground. NERD also facilitates the planning of helicopter flight paths and allows for the instant identification of
areas with low or high bait density. During the recent and successful rodent eradication campaign on Banco Chinchorro in
Mexico, carried out during 2015, NERD results were used to enable dynamic decision-making in the field and to ensure
the efficient use of resources.
Keywords: aerial dispersal, bait density, invasive species, rodenticide
INTRODUCTION
Island ecosystems are vulnerable to the threat posed by
invasive species due to the combination of high levels of
endemism and isolation, coupled with smaller population
sizes (Loope & Mueller-Dombois, 1989; D’Antonio
& Dudley, 1995; Reaser, et al., 2007). Invasive rodent
species such as Rattus rattus are particularly harmful to
island ecosystems. Worldwide, invasive rodents are found
on more than 80% of the world’s islands and their high
potential for dispersal indicates that this number is on the
rise (Russell, et al., 2008; Harris, et al., 2012). The presence
of invasive rodents on islands can lead to rapid population
decreases of both flora and fauna and the extirpation
of endemic species (Towns, et al., 2006; Medina, et al.,
2011) as invasive rodent species begin to dominate
communities (Angel, et al., 2009; Towns, et al., 2013).
Island biodiversity is not only affected by the presence of
invasive rodents; in cases where rodent invasion is severe,
key island ecosystem functions and services are often lost
(e.g., Towns, et al., 2006). Island ecosystems are unable to
recover while rodents are present; as such, the first step to
restore ecological functioning and island biodiversity is the
eradication of invasive rodent species via the dispersal of
rodenticide (Towns & Broome, 2003; Harris, et al., 2012).
The aerial-based dispersal methods of rodenticide bait
via helicopter are preferable to ground-based methods in
many circumstances (Towns & Broome, 2003; Broome,
et al., 2014). Aerial bait dispersal strategies are designed
to cover large areas rapidly, reduce the complications
associated with complex topography, and target potential
refuge sites (Towns & Broome, 2003; Howald, et al.,
2007). The evaluation of the effectiveness of aerial
rodenticide dispersal is informed by bait density maps that
show the spatial variation of bait on the ground (Broome,
et al., 2014). Traditionally, bait density maps have been
created with in situ measurements or from GPS helicopter
trajectories although there are challenges associated with
both methods. To obtain in situ measurements, quadrat
bait density sampling is carried out on the ground and
requires a substantial investment of both time and human
resources. The effectiveness of this method depends on the
topography, accessibility, and climate of the island at the
time of sampling, in addition to existing time constraints
and available manpower. In contrast, the spatial estimation
of bait density from recorded GPS helicopter trajectories is
time intensive and can be imprecise as it is based on several
untested assumptions, the principal one being that the bait
density remains constant within the treated polygon.
We have developed a method for the numerical
estimation of rodenticide density (NERD) that improves
upon the aforementioned methods. NERD creates bait
density maps using GPS helicopter trajectories but is not
constrained by the assumptions of traditional GIS analysis.
NERD does not assume that bait density is constant within
the treated polygon nor is it time intensive. Results from
NERD are both automatic and instantaneous, allowing for
modifications to helicopter flight plans during an ongoing
eradication. During helicopter refuelling, GPS data from
the helicopter are downloaded into NERD and bait density
maps are returned in minutes.
The NERD algorithm combines two models. The first
model estimates the mass flow rate as a function of the bait
bucket aperture diameter and the second model describes
the bait density profile perpendicular to the flight path of
the helicopter. By combining the two models, bait density
on the ground is estimated as a function of the aperture
diameter of the bait bucket and the speed of the helicopter.
In this paper, we present the first field implementation of
NERD on the island of Banco Chinchorro, Mexico, a small
false atoll in which rodents were most likely introduced
during the 19th century (Samaniego, et al., 2017).
METHODS
Study site
Banco Chinchorro is comprised of four flat keys
that create a false atoll measuring 0.5–539 ha, located
in the Caribbean Sea approximately 35 km off the coast
of Quintana Roo, Mexico, and is classified as both a
Biosphere Reserve and Ramsar site (CONANP, 2000;
2006; Samaniego, et al., 2017). Banco Chinchorro presents
a wet tropical climate and is primarily covered with
mangrove vegetation, composed of Rhizophora mangle,
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 47–50. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
47
Island invasives: scaling up to meet the challenge. Ch 1A Rodents: Planning
Laguncularia racemosa, Avicennia germinans, and
Conocarpus erectus, and has tropical trees such as Thrinax
radiata, Bursera simaruba, and Tournefortia gnaphalodes
(Samaniego, et al., 2017). The island provides habitat
for a number of crab species, the American crocodile
(Crocodylus acutus) and the seabird Fregata magnificens
(Samaniego, et al., 2017). Prior to eradication efforts, the
invasive rodent (Rattus rattus) occurred at densities from
6.5–47.9 rats/ha on Cayo Centro to 25.3–102.5 rats/ha on
Cayo Norte Major (Samaniego, et al., 2017). The extensive
mangrove presence on Banco Chinchorro and the presence
of the C. acutus makes ground-based evaluation methods
of bait density both hazardous and ineffectual.
Relationship between density, mass flow rate, and
helicopter speed
The combination of the two models comprising NERD
is presented. Here, we show that the function σ(x,y) used to
represent superficial bait density (kg/m2) complies with the
following equation
where ṁ is the bait flow (kg/s), s is the speed of the
helicopter (m/s) and w is the swath width (m).
We set the origin of a Cartesian coordinate system on
the middle point of the bottom side of a rectangle with base
w and height δy. This way, the bottom side is found at y = 0,
the top side at y = δy, the left side at
, and the right
side at
. The rectangle represents one dispersion
cell.
After the helicopter completes a pass, in each point
(x,y) of the dispersion cell, a superficial bait density is
obtained σ(x,y). In instances where two or more dispersions
cells overlap, we simply add the density from each cell
to get the total density on the overlap. The definition
of the superficial bait density of mass m indicates that
. Rewriting the superficial density substituting
dA by dydx and integrating along the dispersion cell, it
follows that
.
(1)
Assuming superficial density is uniform with respect to the
helicopter’s flight path, represented in Fig. 1, equation (1)
becomes
.
(2)
The left-hand side of the equation represents the linear bait
density, which is related with the mass flow of bait from
the bucket and the speed of the helicopter. A helicopter
equipped with a dispersion bucket with a constant mass
flow rate,
(3)
flies from the point (0,0) to the point (0,δy) with a speed
of
.
(4)
Combining equations (3) and (4), the linear bait density
(5)
is obtained.
Finally, setting equations (2) and (5) equal to each other,
we obtain
.
48
(6)
Fig. 1 Hypothetical island with bait swaths. Each vertical
band represents one bait swath. Each shaded rectangle
represents one dispersion cell. Shade intensity
corresponds to bait density, with darker shades indicating
higher densities.
Equation (6) relates the bait density on the ground
with the mass flow rate and the speed of the helicopter. In
order to get an explicit form of σ, a model is fitted to crossdensity profiles, such as the ones shown in Fig. 2.
Simplified relationship between density, mass flow
rate, and helicopter speed
The required bait density for the successful eradication
of an invasive species on an island is determined by
evaluating the ecosystems present and the biology of the
target species. Once this density has been determined,
NERD can be used to estimate the aperture of the bait
bucket needed for the eradication operation in question
and to plan helicopter flight paths. During the planning
phase of an eradication campaign, prior to arriving on the
island, a simplified relationship between density, mass
flow rate, and helicopter speed is used where bait density
is assumed to be constant along and across the flight path
of the helicopter.
Assuming density is independent of x, i.e. σ does not
change perpendicular to the flight path, equation (6) can be
easily solved to obtain
.
(7)
To write equation (7) as a function of the aperture diameter
of the bait bucket, we express the mass flow rate of bait as
a function of the aperture diameter, ṁ(d). To do this, the
bait in the bucket was weighed and the time required to
empty the bucket was measured and repeated using several
aperture diameters (Fig. 3).
The resulting three-dimensional model,
(8)
is shown in Fig. 4.
An implementation of this model can be found at <http://
github.com/IslasGECI/nerd>.
Rojas-Mayoral, et al.: Improving efficiency by rodenticide density
Fig. 2 Bait density profile perpendicular to the flight path of
the helicopter during a test flight in Oxnard, CA in 2013.
Each black dot shows the bait density measured within
a quadrat.
RESULTS AND DISCUSSION
NERD was used to plan and carry out the 2015
eradication campaign on Cayo Centro of Banco Chinchorro.
Given the desired helicopter speed, NERD was used to
determine the aperture of the bait bucket and the flight
paths of the helicopter required to achieve the desired
bait density within the target polygon. The results of the
2015 rodent eradication campaign on Banco Chinchorro
are detailed by Samaniego et al. (2017). During the course
of the eradication campaign, NERD was operated by two
people and generated an updated bait density map multiple
times each day providing instantaneous visualisations of
the current state of bait application over the island, such
as the map shown in Fig. 5. These visualizations provided
feedback in real time, allowing for helicopter course
corrections and promoting the efficient use of rodenticide
bait.
Fig. 5 shows the final bait density map estimated with
NERD for the eradication campaign. From this map, it is
apparent that all terrestrial areas of Banco Chinchorro were
estimated to be covered with at least 60 kg/ha of rodenticide,
Fig. 3 Mass flow rate (kg/s) as a function of aperture
diameter d (mm). Each dot represents a calibration
event and the black curve is the quadratic model fitted
to the data.
Fig. 4 Surface bait density σ (kg/ha) as a function of aperture
diameter d (mm) and speed s (km/hr). The horizontal
axis shows the aperture diameter of the bait bucket and
the vertical axis shows the speed of the helicopter. The
resulting bait density on the ground is shown in white
superimposed numbers and in the second vertical
grayscale axis.
which was the target bait density for this campaign. The
colormap of Fig. 5 indicates bait density on the ground
(kg/ha), with warmer colours corresponding to lower bait
densities. The large red polygons that appear on the map
represent inland lagoons, which were not covered with
rodenticide bait excepting a few swaths that correspond
to the presence of sandbars within the lagoons. Around
these lagoons, manual bait placement was carried out by
a team of field operatives. The maps generated by NERD
were also used by this team to ensure even bait coverage
and avoid excess bait application. Overall, few areas in
Fig. 5 show bait densities near 100 kg/ha, indicating that
helicopter flight paths were rarely redundant. Furthermore,
any small isolated areas of low bait density were always
surrounded by areas with target bait densities of at least
60 kg/ha.
Fig. 5 Estimated bait density (kg/ha) resulting from the
aerial operation of the rodent eradication campaign in
Banco Chinchorro, Mexico, during 2015.The shade
bar on the right indicates predicted bait density on the
ground (kg/ha), with lighter shades indicating lower
densities. The large white polygons show the location of
inland lagoons.
49
Island invasives: scaling up to meet the challenge. Ch 1A Rodents: Planning
The information provided by NERD was indispensable
to the eradication campaign on Banco Chinchorro and
allowed for immediate decisions to be made regarding not
only the aerial dispersal of rodenticide bait, but also for
the manual placement of bait on the ground. Until now,
efforts to generate bait density maps have been inefficient
and results were often not available until after the end of an
eradication campaign. NERD provides information in real
time, enabling dynamic decision making in the field and
ensuring the efficient use of resources.
REFERENCES
Angel, A., Wanless, R.M. and Cooper, J. (2009). ‘Review of impacts of
the introduced house mouse on islands in the Southern Ocean: Are mice
equivalent to rats?’ Biological Invasions 11: 1743–1754.
Broome, K.G., Cox, A., Golding, C., Cromarty, P., Bell, P. and McClelland,
P. (2014). Rat Eradication Using Aerial Baiting: Current Agreed Best
Practice Used in New Zealand (Version 3.0). Wellington, New Zealand:
Department of Conservation.
CONANP. (2000). Programa de Manejo Reserva de la Biosfera Banco
Chinchorro. Mexico City, Mexico: INE-SEMARNAT.
CONANP. (2006). Programa de Conservación y Manejo Parque Nacional
Arrecife Alacranes. Mexico City, Mexico: Comisión Nacional de Áreas
Naturales Protegidas.
D’Antonio, C.M. and Dudley, T.L. (1995) ‘Biological Invasions as Agents
of Change on Islands Compared with Mainlands’. In: P.M. Vitousek,
L. Loope and H. Adsersen (eds.) Islands: Biological Diversity and
Ecosystem Function, pp. 103–121. Berlin, Germany: Springer Verlag.
Harris, D.B., Gregory, S.D., Bull, L.S., and Courchamp, F. (2012). ‘Island
prioritization for invasive rodent eradications with an emphasis on
reinvasion risk’. Biological Invasions 14(6): 1251–1263.
Howald, G., Donlan, C.J., Galvan, J.P., Russell, J.C., Parkes, J.,
Samaniego, A., Wang, Y., Veitch, D., Genovesi, P., Pascal, M., Saunders,
A. and Tershy, B. (2007) ‘Invasive rodent eradication on islands’.
Conservation Biology 21(5): 1258–1268.
Loope, L. and Mueller-Dombois D. (1989). ‘Characteristics of Invaded
Islands, with Special Reference to Hawaii’. In: J.A. Drake, H.A.
Mooney, F. di Castri, R.H. Groves, F.J. Kruger, M. Rejmanek and
M. Williamson (eds.). Scope 37 – Biological Invasions: A Global
Perspective, pp. 257–280. Chichester: John Wiley & Sons.
Medina, F.M., Bonnaud, E., Vidal, E., Tershy, B.R., Zavaleta, E.S.,
Donlan, C.J., Keitt, B.S., Le Corre, M., Horwarth, S.V. and Nogales,
M. (2011). ‘A global review of the impacts of invasive cats on island
endangered vertebrates’. Global Change Biology 17(11): 3503–3510.
Reaser, J. K., Meyerson, L. A., Cronk, Q., De Poorter, M., Eldrege, L.
G., Green, E. and Vaiutu, L. (2007). ‘Ecological and socioeconomic
impacts of invasive alien species in island ecosystems’. Environmental
Conservation 34(2): 98–111.
Russell, J.C., Towns, D.R. and Clout, M.N. (2008). Review of Rat Invasion
Biology: Implications for Island Biosecurity. Science for Conservation
286. Wellington, New Zealand: Department of Conservation.
Samaniego-Herrera, A., Aguirre-Muñoz, A., Bedolla-Guzmán, Y.,
Cárdenas-Tapia, A., Félix-Lizárraga, M., Méndez-Sánchez, F.,
Reina-Ponce, O., Rojas-Mayoral, E. and Torres-García, F. (2017).
‘Eradicating invasive rodents from wet and dry tropical islands in
Mexico’. Oryx 1–12: DOI:10.1017/S0030605316001150.
Towns, D.R., Atkinson, I.A.E. and Daugherty, C.H. (2006). ‘Have
the harmful effects of introduced rats on islands been exaggerated?’
Biological Invasions 8(4): 863–891.
Towns, D.R. and Broome, K.G. (2003). ‘From small Maria to massive
Campbell: Forty years of rat eradications from New Zealand islands’.
New Zealand Journal of Zoology 30(4): 377–398.
Towns, D.R., West, C.J. and Broome, K.G. (2013). ‘Purposes, outcomes
and challenges of eradicating invasive mammals from New Zealand
islands: An historical perspective’. Wildlife Research, 40: 94–107.
50
D. Rueda, V. Carrion, P.A. Castaño, F. Cunninghame, P. Fisher, E. Hagen, J.B. Ponder, C.A. Riekena, C. Sevilla, H. Shield, D. Will and K.J. Campbell
Rueda, D.; V. Carrion, P.A. Castaño, F. Cunninghame, P. Fisher, E. Hagen, J.B. Ponder, C.A. Riekena, C. Sevilla, H. Shield, D. Will and
K.J. Campbell. Preventing extinctions: planning and undertaking invasive rodent eradication from Pinzon Island, Galapagos
Preventing extinctions: planning and undertaking invasive rodent
eradication from Pinzon Island, Galapagos
D. Rueda1, V. Carrion2, P.A. Castaño2, F. Cunninghame3, P. Fisher4, E. Hagen5, J.B. Ponder6, C.A. Riekena7, C. Sevilla1,
H. Shield8, D. Will5 and K.J. Campbell2,9
Galapagos National Park Directorate, Charles Darwin Ave., Puerto Ayora, Galapagos Islands, Ecuador. 2Island
Conservation, Charles Darwin Ave., Puerto Ayora, Galapagos Islands, Ecuador.
<karl.campbell@islandconservation.org>. 3Charles Darwin Foundation, Charles Darwin Ave., Puerto Ayora,
Galapagos Islands, Ecuador. 4Landcare Research, P.O. Box 69040, Lincoln 7640, New Zealand. 5Island Conservation,
2100 Delaware Ave, Santa Cruz, CA 95060, USA. 6The Raptor Center, University of Minnesota, 1920 Fitch Avenue,
St. Paul, MN 55108, USA. 7Bell Laboratories, Inc., 3699 Kinsman Blvd., Madison, WI 53704, USA. 8P.O. Box
5184, Springlands 7201, Blenheim, Marlborough, New Zealand. 9School of Geography, Planning & Environmental
Management, The University of Queensland, St Lucia 4072, Australia.
1
Abstract Invasive black rats (Rattus rattus) were successfully eradicated during 2012 from Pinzon Island in the Galapagos
archipelago using the rodenticide brodifacoum. Potential exposure to brodifacoum in Pinzon tortoises (Chelonoidis
ephippium), Pinzon lava lizards (Microlophus duncanensis) and Galapagos hawks (Buteo galapagoensis) was mitigated
by captive holding of subpopulations. This was successful for all species during and shortly after baiting, however
mortality of Galapagos hawks occurred post-release, likely due to the persistence of residual brodifacoum in lava lizards.
Since 2013, Pinzon tortoise hatchlings are surviving in-situ for the first time in at least 120 years and the eradication of
black rats is expected to have significant benefits for at least 15 other island-endemic species.
Keywords: brodifacoum, endemic species, eradication, Rattus rattus, restoration
INTRODUCTION
Islands are centres of endemism and endangerment,
with about one-fifth of the world’s threatened amphibians,
one-quarter of the threatened mammals and more than
one-third of the threatened birds being endemic to islands
(Fonseca, et al., 2006). Invasive non-native species are
major extinction drivers, with predators like rodents being
particularly damaging (Bellard, et al., 2016; Doherty, et al.,
2016). Four rodent species (Rattus rattus, R. norvegicus,
R. exulans, Mus musculus) have been introduced to
islands holding 88% of all insular critically endangered
or endangered terrestrial vertebrates (TIB Partners, 2014).
Invasive rodents cause population declines and extinctions
of insular flora and fauna and interrupt ecosystem
processes with negative cascading effects (Fukami, et
al., 2006; Towns, et al., 2006; Jones, et al., 2008; Kurle,
et al., 2008). To recover endangered populations and
restore ecosystem processes, invasive rodents on islands
are increasingly targeted for eradication, with at least 650
eradication attempts of introduced Rattus spp. populations
to date (Russell & Holmes, 2015). Eradication of invasive
mammals from islands results in positive responses by
native species with few exceptions (Jones, et al., 2016).
Pinzon Island (1,815 ha), in the Galapagos archipelago,
is uninhabited and is entirely within the Galapagos National
Park. Pinzon endemics include three reptiles (Pinzon
Island tortoise (Chelonoidis ephippium), Pinzon lava
lizard (Microlophus duncanensis), Pinzon leaf-toed gecko
(Phyllodactylus duncanensis)), six land snails (Bulimulus
duncanus, B. eschariferus ventrosus, B. pinzonensis, B.
pinzonopsis, B. prepinguis, Bulimulus sp. undescribed),
and six insects in the orders Homoptera and Hemiptera.
Thirteen species considered threatened by the IUCN are
present, such as marine iguanas (Amblyrhynchus cristatus),
Galápagos hawk (Buteo galapagoensis), land snails and
the cactus Opuntia galapageia, along with several species
of unassessed conservation status (IUCN, 2015).
The island was most heavily used by whalers harvesting
tortoises in the early to mid-1800s and it is during this period
that black rats (R. rattus) were most likely introduced, with
specimens first collected in 1891 (Patton, et al., 1975).
Black rats are the only invasive mammals that successfully
populated the island. On visiting Pinzon Island in 1903,
Rolland Beck noted “We… captured altogether nearly
thirty live tortoises…. We were much chagrined, however,
at finding no very small specimens, but soon came to the
conclusion that the large rats, of recent introduction, and
now common everywhere on the island, eat the young
as soon as they are hatched” (Beck, 1903 p. 174). Heavy
predation by black rats on eggs and hatchlings saw a halt of
recruitment into the tortoise population for over a century,
leaving fewer than 65 old tortoises that had survived
human harvesting efforts (MacFarland, et al., 1974; Jensen,
et al., 2015). In response, a ‘head-starting’ programme
was initiated nearly 50 years ago. This entailed collecting
eggs or recently hatched individuals from nests on-island,
transporting them to the Galapagos National Park’s centre
on Santa Cruz Island where hatchlings were reared ex-situ
until 4–5 years old, at which time they were repatriated
back to Pinzon Island (Jensen, et al., 2015). Elsewhere in
the Galapagos Archipelago, invasive black rats have been
implicated in the extinction of native rodents, declines
and extirpations of sea- and land-bird populations and
other fauna (Cruz & Cruz, 1987; Steadman, et al., 1991;
Dowler, et al., 2000). By consuming seeds and seedlings
they impede vegetation regeneration and alter forest
dynamics, affecting entire ecosystems (Clark, 1981).
Impacts on invertebrates have not been quantified in the
Galapagos Archipelago but likely occur based on reports
from elsewhere (e.g. Towns, et al., 2006).
Conservationists attempted to eradicate black rats from
Pinzon Island in 1988 utilising rodenticide bait dumps
(coumatetralyl powder combined with rice in paper bags)
and hand broadcast of baits containing brodifacoum and
coumatetralyl (Cayot, et al., 1996; Harper & Carrion,
2011). The project was unsuccessful, although rodents were
not detected for nine months after the operation (Cayot, et
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 51–56. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
51
Island invasives: scaling up to meet the challenge. Ch 1A Rodents: Planning
al., 1996). This rodent suppression resulted in recruitment
of Pinzon tortoises, anecdotal reports of increases in
the abundance of juvenile marine iguanas, populations
of Pinzon lava lizards and Galapagos doves (Zenaida
galapagoensis), and decreases in populations of short-eared
owl (Asio flammeus galapagoensis) and Galapagos hawks
(Muñoz, 1990; Cayot, et al., 1994; Cayot, et al., 1996).
A cessation of predation of Pinzon tortoise hatchlings by
black rats was recorded, however an 80% predation rate by
native Galapagos hawks occurred for two years after the
eradication attempt (Morillo Manrique, 1992). Ambitious
for its time, this failed eradication attempt set back rodent
eradications in the archipelago for the next three decades,
with the exception of attempts on just a few small (<20 ha)
islands (Harper & Carrion, 2011).
Large-scale feral pig (Sus scrofa), goat (Capra hircus)
and donkey (Equus asinus) eradications were implemented
in the Galapagos Archipelago throughout the late 1990s
and 2000s (Cruz, et al., 2005; Carrion, et al., 2007;
Carrion, et al., 2011) renewing interest in large-scale
rodent eradications. In 2007, an international workshop
laid out a plan for developing capacity and confidence to
eventually eradicate rodents from inhabited Floreana Island
(17,253 ha) with complexity and scale being increased at
each step (CDF & GNPS, 2007). Later in 2007, North
Seymour Island (184 ha) was hand baited with wax blocks
containing brodifacoum, successfully eradicating black
rats (Harper, et al., 2011). In 2011, the first aerial broadcast
of brodifacoum baits in South America eradicated
rodents from Rabida and 11 other islands totalling 705
ha (Campbell, et al., 2013). Pinzon (1,815 ha) and Plaza
Sur (12 ha) islands were originally considered within the
group of islands to be targeted along with Rabida but their
operations were delayed to allow trials to be conducted for
increasing certainty of non-target risks to tortoises and for
pilot mitigation strategies for Galapagos hawk (Campbell,
et al., 2013). As part of the Rabida project, 20 Galapagos
hawks were kept in captivity and released once the risk of
mortality from rodenticide poisoning was considered past
(Campbell, et al., 2013).
Bait application
As with previous rodent eradications in the archipelago,
bait application was timed for the last three months of
the dry season (October–December), when rat breeding
ceases and their numbers are at a minimum, after a typical
six-month dry-spell (Clark, 1980). Bait type used was
‘Brodifacoum 25D Conservation’ (Bell Laboratories,
Madison WI). Baits were 2.5 g compacted crushed grain
pellets of 13 mm diameter, containing 25 μg (25 ppm)
of brodifacoum per kg of bait, blue dye and pyranine
biomarker, a non-toxic, odourless and tasteless dye that
fluoresces green under UV light. Bait was applied in two
aerial applications 23 days apart at an average rate of 6.72
kg/ha for the first application (15–17 November, 2012)
and 4.85 kg/ha for the second application (8–9 December,
2012; Fig. 1). Pre-eradication trials in 2010 had determined
that target application rates of 6 kg/ha followed by 3 kg/ha
ensured bait was available in all habitats for at least four
days. It had been planned to have bait applications 7–10
Here we describe the successful eradication of invasive
black rats from Pinzon Island and the measures taken to
mitigate negative impacts of rodenticide bait application
on non-target wildlife.
METHODS
Site description
Pinzon Island, located in the centre of the Galapagos
Archipelago, has a maximum elevation of 458 m and
approximately 18 km of rocky coastline with steep cliffs on
the southern and north-western coasts. Large lava blocks
dominate the slopes of Pinzon. There are two craters at the
centre of the island. The vegetation is xerophytic and there
are no permanent bodies of water. Two small islets, each
of approximately 0.4 ha in size, lie close inshore. Pinzon
has no terrestrial visitor sites and is more than 10 km from
any other island with invasive rodents, making unassisted
reinvasion highly unlikely.
Baseline genetic sampling of rodents from Pinzon
In 2011, black rats were trapped, euthanised and
samples taken (n=89) for future genetic analyses in case
rodents were detected after the eradication attempt. If
this occurred, as island populations of black rats can be
differentiated in the Galapagos (Willows-Munro, et al.,
2016), genetic samples from the pre- and post-eradication
attempt could be compared to help determine whether
reintroduction or eradication failure occurred (Abdelkrim,
et al., 2007).
52
Fig.1 Bait density (kg/ha) maps of Pinzon Island from (A)
first, and (B) second bait applications. Circles indicate
where baits in paper bags were applied (bola baiting
points).
Rueda, et al.: Rodent eradication Pinzon Island, Galapagos
days apart, however a pregnant rat was reported after the
first bait application, prompting a decision to extend the
duration between applications to maximise the probability
that all rats would be exposed to bait (Keitt, et al., 2015).
An experienced pilot flew the helicopter (Eurocopter
AS350-B2, France) guided by GPS, pre-programmed flight
lines and light-bar (Tracmap flight unit, New Zealand). The
helicopter was fitted with a custom agricultural style bait
spreader bucket (CSI Helicopters, New Zealand) that was
used to spread bait systematically over Pinzon Island (Fig.
1). Pre-programmed flight-lines 40 m apart provided a 100%
overlap for inland areas, as previous bucket calibration
indicated the bucket had an 80 m effective baiting swath
width. Interior flight-lines ran coast-to-coast, with lines
starting and ceasing 40 m inside of the coast to minimise the
amount of bait entering the marine environment. Interior
flight-lines were flown approximately north–south on the
first application and east–west on the second. Two flightlines were flown around the coast. The ‘outer’ coastal swath
was flown along the coastline with a deflector attached to
the bucket, providing 40 m unidirectional sowing towards
the inland, to minimise bait entering the ocean. The inner
coastal swath was carried out with the standard bucket, 60
m inland from the coast, thereby achieving a 50% overlap
with the outer coastal swath. Sections of cliff over 50ᵒ slope
on the southern side of the island were treated as a separate
block to achieve twice the bait application rate of the
interior, which is considered best practice (Broome, et al.,
2014). GPS tracks were inspected periodically throughout
each application. Any gaps identified in bait coverage were
then baited in subsequent flights the same day.
Hand-baiting was conducted around the on-island
camp, temporary hawk aviaries and one islet. The second
islet was baited by hand from the helicopter using paper
bags with 10 baits in each to achieve target application
rates. Any areas along the coast that may not have received
bait due to extreme steepness, overhangs, and deep cut
gullies were also hand baited with bait in paper bags. Bait
availability plots (25 m × 1 m; n=10) were used to monitor
bait persistence after each aerial application at points from
the coast to the highlands on the northern side of the island.
Plots contained the number of pellets that corresponded to
the bait application rate for each application. Each bait
pellet was marked with a pin flag, which was removed as
pellets were consumed. Plots were checked daily between
the first and second applications and for 13 days following
the second application.
Two boats acted as a floating base during helicopter
baiting operations. One boat, fitted with a helicopter
landing platform, also acted as the helicopter refuelling
station. The second boat was fitted with a wooden platform
from which bait was loaded into the bait spreader bucket as
the helicopter hovered to one side.
Non-target species
Brodifacoum is the most commonly used toxicant for
rodent eradications on islands and has the highest success
rate (DIISE, 2016). Although an effective rodenticide,
brodifacoum is highly toxic to mammals and birds, is
known to persist in tissue containing vitamin K epoxide
reductase (Eason, et al., 2002) and therefore presents risks
to non-target wildlife through primary and secondary
pathways of exposure (Broome, et al., 2015). Reptiles are
considered to be less susceptible to brodifacoum (Weir,
et al., 2015) but may also present a secondary exposure
pathway to their predators.
An a priori non-target risk assessment which included
Pinzon wildlife was conducted in 2010 (Campbell,
2010). A revised assessment (Fisher & Campbell, 2012)
incorporated a suite of new information from the 2011
rodent eradications, and captive feeding trials used to
assess risk of brodifacoum exposure in giant tortoises, lava
lizards, geckos and snakes (Fisher, 2011a; Fisher, 2011b).
Lava lizard samples were taken from Rabida Island before
and after bait application to assess the incidence and
persistence of residual brodifacoum in lava lizards but all
these samples perished when a freezer was unplugged.
Population-level impacts of brodifacoum applications
were assessed for lava lizards and land birds using a
before-after control-impact study design on Rabida,
Bartolome, Bainbridge #3 and Beagle Sur islands, with
Pinzon acting as a control. Based heavily upon the 2012
non-target risk assessment the Galapagos National Park
Directorate and other partners determined that mitigation
actions should be conducted for Pinzon tortoise, Galapagos
hawk, Pinzon lava lizard, lava gull (Larus fuliginosus) and
endemic land snails. Mitigation plans were developed
for each taxon (Cunninghame, 2012; Cunninghame, et
al., 2012; Oberg & Campbell, 2012; Parent & Campbell,
2012) except tortoises. Mitigation plans for lava gulls and
land snails were not implemented. Lava gulls were not
present on Pinzon Island during operations, and in searches
undertaken before bait application all snails found were
estivating so would not be exposed to bait.
Fifteen adult Pinzon tortoises were brought into
captivity two years prior to baiting operations, housed on
Santa Cruz Island and returned in good health two years
after the rodent eradication was complete. Forty Pinzon
lava lizards were taken into captivity prior to baiting and
were maintained in enclosures on Pinzon Island. Termite
larvae were provided as food every other day. Ten days after
the second application the potential for bait consumption
by lava lizards, as determined by bait degradation plots,
was determined to be minimal and all surviving individuals
were released near their capture sites.
Sixty hawks were taken into captivity on Pinzon Island,
most prior to baiting operations, held in purpose-built
aviaries and maintained on diets of goat meat, day-old
chicks and (prior to baiting) rats. All hawks were ringed
and genetic samples taken for future study. Four additional
hawks were captured, ringed, and treated with injectable
(intramuscular) vitamin K1, however due to limited aviary
space they were released. Three hawks were identified,
but never captured. Captive hawks were released 12–14
days after the second aerial application of bait. Telemetry
transmitters were fitted to 32 hawks before release.
Confirmation of eradication
Efficacy of rat detection methods was demonstrated
prior to the eradication. Corrugated plastic chew cards with
peanut butter (Oberg, et al., 2014), visual sightings, and
signs of activity (prints, faeces, gnawed seed pods) readily
indicated rodent presence across Pinzon Island. In January
2015 (25 months after the second bait application), these
same methods were used to confirm black rat eradication
with 1,140 chew cards deployed for at least 54 days, spaced
at 25 m intervals along a trail network covering the island.
RESULTS
Baiting operations successfully applied bait across the
island at the desired rates, as determined by helicopter GPS,
baiting rate and effective swath width being overlaid on
island maps (Fig. 1). Monitoring conducted more than two
years after bait applications did not detect invasive rodents
on Pinzon Island. None of the 1,140 chew cards deployed
had rodent sign, while nearly 100% of chew cards placed
pre-eradication did. Seed pods of Acacia spp. were intact
and showed no sign of rodent damage across the island.
53
Island invasives: scaling up to meet the challenge. Ch 1A Rodents: Planning
Based on this evidence we conclude that black rats were
eradicated from Pinzon Island.
Bait availability plots indicated that after the first
application (6.3 kg/ha) the average remaining density
across all plots was above 2 kg/ha until day three (2.07
± 1.75 kg/ha) and did not drop below 1 kg/ha until day
12 (0.9 ± 1.04 kg/ha). One plot had no bait available on
day four; tortoises were observed consuming the baits.
When the second application occurred, the average bait
density remaining was approximately 0.5 ± 1.14 kg/ha.
Bait availability plots for the second application (4.2 ±
1.19 kg/ha), indicated average availability remained above
2 kg/ha until day seven (2.7 ± 1.64 kg/ha), and less than 1
kg/ha remaining at day 12 (0.99 ± 1.67 kg/ha). Individual
plots went to zero within two days due to Pinzon tortoises
consuming bait.
Consumption of bait by Pinzon lava lizards and Pinzon
tortoises was observed at higher rates than anticipated
and evidenced by faeces containing blue dye, however
no mortality in these species was observed in the wild.
Additionally, mitigation efforts were successful at
maintaining a separate population of Pinzon lava lizards
and Pinzon tortoises as insurance in the case of any
unexpected mortality in wild populations. Two lava lizards
escaped captivity and five captive lizards died (survival
rate of 87%).
All captive Galapagos hawks survived captivity and
were released in healthy condition. Between 12 and 170
days after release, mortality of 22 tracked Galapagos
hawks was recorded (Rueda, et al., 2016). Necropsy of
four of these birds showed signs of anticoagulant toxicosis,
with 379 ppb brodifacoum measured in one hawk liver
(Rueda, et al., 2016). Monitoring of live-caught Pinzon
lava lizards also showed residual brodifacoum in liver, for
at least 850 days after bait application (Rueda, et al., 2016).
The fate of 28 released hawks remains unknown, but they
likely died. The remaining Pinzon Island Galapagos hawk
population (n=10) was recaptured, placed into captivity in
June 2013 and treated with Vitamin K1, while toxicological
monitoring of Pinzon lava lizards continued (Rueda, et
al., 2016). These captive Galapagos hawks from Pinzon
Island, representing 15% of the original population, were
released when risk was considered acceptable in July
and August 2016 with telemetry and GPS transmitters.
Within three months of release, Galapagos hawks from
Pinzon Island had nests with eggs. As of April 2018, nine
nesting attempts have resulted in five healthy chicks, two
nest failures, one unknown outcome and one pending (P.
Castaño, unpublished data 2018). These and related events
will be reported in greater detail elsewhere. Galapagos
hawks continue to be monitored on Pinzon Island, as does
toxicological monitoring in Pinzon lava lizards.
The eradication of black rats and actions taken to
mitigate non-target impacts on Pinzon Island cost an
estimated $1,501,000 (2013 US dollars). Cost breakdown
estimates include planning ($101,000), implementation
($909,000), non-target species management ($101,000)
and indirect costs ($390,000) (Holmes, et al., 2015).
DISCUSSION
Recovery of native and endemic species due to the
successful eradication of invasive black rats from Pinzon
is already evident, with ongoing monitoring expected
to reveal further biodiversity gains. Pinzon tortoise
hatchlings are now surviving in the wild for the first time
in over 120 years (Tapia Aguilera, et al., 2015). With
natural recruitment now occurring the Pinzon tortoise
head-starting may soon no longer be required (Jensen,
et al., 2015). Land-bird surveys in early 2018 found two
54
species (cactus finch Geospiza scandens, Galapagos rail
Laterallus spilonota) never before recorded from the island
(Fessl, et al., unpublished data 2018). Endemic land snails
also appear to be on the increase, indeed a new species
of land snail was discovered two years post-eradication
in permanent snail monitoring plots on the island and is
currently being described (C. Parent, unpublished data
2015). With a major threat now removed, threatened land
snails and other species may now be eligible for downlisting from the IUCN Red List. Similarly, on Rabida Island,
two years after invasive Norway rats (Rattus norvegicus)
were eradicated, two island endemic land snails that were
considered extinct for over 100 years were rediscovered
(Campbell, et al., 2013; C. Parent, unpublished data 2012).
Also, on Rabida Island, a leaf-toed gecko was found posteradication in late 2012 (Campbell, et al., 2013). The only
known geckos from Rabida were recorded from subfossils
estimated at more than 5,700 years old, which were
classified to genus only (Steadman, et al., 1991). Although
the specimen was identified at the time as the archipelago
endemic Phyllodactylus galapagensis (W. Tapia Aguilera
pers. comm. 2013), a recently proposed taxonomic split
divides P. galapagensis into four species by major islands,
including Pinzon (Torres-Carvajal, et al., 2014). Future
analyses including samples of geckos from Rabida Island
may also see a unique species identified for that island.
Eradication of black rats from Pinzon Island was
arguably a cost-effective conservation action at US$827
/ ha, resulting in the removal of a significant threat
for at least 15 Pinzon Island endemic species, several
archipelago endemic species and 12 IUCN threatened
species. The negative impact of this conservation action
on Pinzon’s population of Galapagos hawks is expected
to be short-term, with breeding already underway on the
island. However, without additional mitigation actions this
population may have been lost due to secondary poisoning,
potentially requiring a translocation to re-establish
Galapagos hawks on Pinzon Island. Longer-term impacts
will only be discovered in time.
The persistence of brodifacoum residues in Pinzon
lava lizards for at least 850 days was unexpected (Rueda,
et al., 2016) and, as it was unknown at the time, was not
considered within a priori risk assessments. Ingestion
of lizards carrying residual brodifacoum for prolonged
periods was likely a significant contributor to unpredicted
and unexpectedly high mortality of released Galapagos
hawks (Rueda, et al., 2016). The use of a prescribed
duration for captive holding was, in hindsight, an error.
Future mitigation efforts should use biological criteria
(e.g. bait availability, sentinel animals) relevant to the
pathways being managed to determine when captive
held or translocated non-target species be released after
brodifacoum bait has been used for rodent eradication.
Pinzon is currently the largest island in the Galapagos
to be freed of invasive rodents, and the fourth-largest island
globally to be eradicated of black rats (behind Macquarie,
Rangitoto and Australia Islands; DIISE, 2016). Continuing
on, as suggested in the original roadmap (CDF & GNPS,
2007), the next island in the Galapagos archipelago
being targeted for rodent eradication is Floreana Island,
nearly an order of magnitude larger than Pinzon, with
160 human inhabitants, pets, livestock, surface water and
a suite of wildlife species that are expected to require
mitigation actions (Island Conservation, 2013). Floreana
Island represents significant challenges but also major
opportunities for incorporating social well-being targets
in invasive species eradication projects, as well as
biodiversity targets to benefit 55 IUCN threatened species
and creating the conditions for the reintroduction of 13
species extirpated by invasive species.
Rueda, et al.: Rodent eradication Pinzon Island, Galapagos
Removing non-native invasive rodents from islands
is a proven approach to protecting endemic biodiversity
(Jones, et al., 2016), and anticoagulant rodenticides are
currently the most reliably effective method to achieve
this. Until alternative rodent-specific methods become
available (Campbell, et al., 2015), practitioners will have
to become increasingly skilled at mitigating risks to nontarget species related to rodent eradications to ensure the
conservation benefits of this powerful tool are maximised.
ACKNOWLEDGEMENTS
We thank Galapagos National Park Directorate and
Island Conservation staff for assisting in planning, logistics,
and the field work conducted to implement this project. L.
Cayot, J. Flanagan, J. Gibbs, C. Parent, P. Parker, T. de
Vries, and staff from project partner organisations provided
information relevant to developing or assisted in developing
mitigation plans for non-target species. G. Harper and D.
Brown assisted in operational planning. New Zealand
Department of Conservation’s Island Eradication Advisory
Group and Island Conservation’s Eradication Advisory
Team provided peer review of earlier drafts of plans.
Thanks to the editors and two anonymous reviewers for
constructive feedback on earlier drafts of this manuscript.
Financial support
The work described here was supported by Galapagos
National Park Directorate, Island Conservation, The
Leona M. and Harry B. Helmsley Charitable Trust (IC),
Bell Laboratories (IC), The David and Lucile Packard
Foundation (IC), The Raptor Center at the University of
Minnesota, Morris Animal Foundation (Grant D14ZO-061;
TRC), Galapagos Conservancy (IC) and the Charles
Darwin Foundation. The Galapagos Invasive Species Fund
(IC) supported development of this manuscript. This work
was carried out as part of management operations of the
Galapagos National Park Directorate.
REFERENCES
Abdelkrim, J., Pascal, M. and Samadi, S. (2007). ‘Establishing causes of
eradication failure based on genetics: Case study of ship rat eradication
in Ste. Anne Archipelago’. Conservation Biology 21: 719–730.
Beck, R. (1903). ‘In the Home of the Giant Tortoise’. Annual Report of
the New York Zoological Society 7:160–174.
Bellard, C., Cassey, P. and Blackburn, T.M. (2016). ‘Alien species as a
driver of recent extinctions’. Biology Letters 12: 20150623.
Carrion, V., Donlan, C.J., Campbell, K.J., Lavoie, C. and Cruz, F. (2011).
‘Archipelago-wide island restoration in the Galápagos Islands: Reducing
costs of invasive mammal eradication programs and reinvasion risk’.
PLOS ONE 6: e18835.
Cayot, L.J., Rassmann, K. and Trillmich, F. (1994). ‘Are marine
iguanas endangered on islands with introduced predators’. Noticias de
Galápagos 53: 13–15.
Cayot, L.J., Muñoz, E. and Murillo, G. (1996). ‘Conservation of the
tortoises of Pinzón Island’. Charles Darwin Research Station Annual
Report 1988–1989: 29–35.
CDF and GNPS (2007). Project Pinzon: Ecosystem Restoration of the
Galapagos Islands through Introduced Rodent Removal. April 2–12,
2007. Puerto Ayora, Galápagos: Fundación Charles Darwin / Servicio
Parque Nacional Galápagos.
Clark, D.A. (1981). ‘Foraging patterns of black rats across a desertmontane forest gradient in the Galapagos Islands’. Biotropica 13:
182–194.
Clark, D.B. (1980). ‘Population ecology of Rattus rattus across a desertmontane forest gradient in the Galapagos Islands’. Ecology 61: 1422–
1433.
Cruz, F., Donlan, C.J., Campbell, K. and Carrion, V. (2005). ‘Conservation
action in the Galápagos: Feral pig (Sus scrofa) eradication from Santiago
Island’. Biological Conservation 121: 473–478.
Cruz, J.B. and Cruz, F. (1987). ‘Conservation of the dark-rumped petrel
Pterodroma phaeopygia in the Galapagos Islands, Ecuador’. Biological
Conservation 42: 303–311.
Cunninghame, F. (2012). Lava Gull (Leucophaeus fuliginosus) Mitigation
During an Aerial Broadcast of Brodifacoum Cereal Baits on Plaza
Sur Island, Galapagos. Unpublished report. Puerto Ayora, Galapagos
Islands: Charles Darwin Foundation.
Cunninghame, F., Ponder, J. and Campbell, K. (2012). Galápagos Hawk
(Buteo galapagoensis) Mitigation During an Aerial Broadcast of
Brodifacoum Cereal Baits on Pinzón Island, Galápagos. Unpublished
report. St Paul, Minnesota: The Raptor Center.
DIISE. (2016). ‘The Database of Island Invasive Species Eradications,
Developed by Island Conservation, Coastal Conservation Action
Laboratory UCSC, IUCN SSC Invasive Species Specialist Group,
University of Auckland and Landcare Research New Zealand’. <http://
diise.islandconservation.org>. Accessed 16 February 2016.
Doherty, T.S., Glen, A.S., Nimmo, D.G., Ritchie, E.G. and Dickman, C.R.
(2016). ‘Invasive predators and global biodiversity loss’. Proceedings
of the National Academy of Sciences: 201602480.
Dowler, R.C., Carroll, D.S. and Edwards, C.W. (2000). ‘Rediscovery
of rodents (Genus Nesoryzomys) considered extinct in the Galapagos
Islands’. Oryx 34: 109–117.
Eason, C.T., Murphy, E.C., Wright, G.R.G. and Spurr, E.B. (2002).
‘Assessment of risks of brodifacoum to non-target birds and mammals
in New Zealand’. Ecotoxicology 11: 35–48.
Fisher, P. (2011a). Non-Target Risk to Galápagos Reptiles from Application
of Brodifacoum Bait for Rodent Eradication: Captive Feeding Trials.
Unpublished report. Lincoln, New Zealand: Landcare Research.
Broome, K., Cox, A., Golding, C., Cromarty, P., Bell, P. and McClelland,
P. (2014). Rat Eradication Using Aerial Baiting: Current Agreed Best
Practice Used in New Zealand (V3.0). New Zealand Department of
Conservation internal document DOCDM-839096. Wellington, New
Zealand: Department of Conservation.
Fisher, P. (2011b). Non-Target Risk to Galápagos Tortoises from
Application of Brodifacoum Bait for Rodent Eradication: Captive
Feeding Trials. Unpublished report. Lincoln, New Zealand: Landcare
Research.
Broome, K.G., Fairweather, A.A.C. and Fisher, P. (2015). Brodifacoum
Pesticide Information Review. Version 2015/1. Docdm-25436.
Hamilton, New Zealand: Department of Conservation.
Fisher, P. and Campbell, K. (2012). Non-Target Risk Assessment for
Rodenticide Application on Pinzón and Plaza Sur, Galápagos Islands.
Unpublished report. Lincoln, New Zealand: Landcare Research.
Campbell, K. (2010). Non-Target Risk Assessment for Rodenticide
Applications in the Galapagos Islands. Unpublished report. Puerto
Ayora, Galapagos Islands: Island Conservation and Charles Darwin
Foundation.
Fonseca, G.A.B.d., Mittermeier, R.A. and Mittermeier, C.G. (2006).
Conservation of Island Biodiversity: Importance, Challenges, and
Opportunities, Washington, DC: Center for Applied Biodiversity
Science at Conservation International.
Campbell, K.J., Carrion, V. and Sevilla, C. (2013). ‘Increasing the Scale of
Successful Invasive Rodent Eradications in the Galápagos Islands’. In:
L.J. Cayot, D. Cruz and R. Knab (eds.) Galapagos Report 2011–2012,
pp. 194–198. Puerto Ayora, Galapagos, Ecuador: GNPS, GCREG, CDF
and GC.
Fukami, T., Wardle, D.A., Bellingham, P.J., Mulder, C.P.H., Towns,
D.R., Yeates, G.W., Bonner, K.I., Durrett, M.S., Grant-Hoffman, M.N.
and Williamson, W.M. (2006). ‘Above- and below-ground impacts of
introduced predators in seabird-dominated island ecosystems’. Ecology
Letters 9: 1299–1307.
Campbell, K.J., Beek, J., Eason, C.T., Glen, A.S., Godwin, J., Gould, F.,
Holmes, N.D., Howald, G.R., Madden, F.M., Ponder, J.B., Threadgill,
D.W., Wegmann, A. and Baxter, G.S. (2015). ‘The next generation
of rodent eradications: Innovative technologies and tools to improve
species specificity and increase their feasibility on islands’. Biological
Conservation 185: 47–58.
Harper, G.A. and Carrion, V. (2011). ‘Introduced rodents in the galápagos:
colonisation, removal and the future’. In: C.R. Veitch, M.N. Clout and
D.R. Towns (eds.) Island invasives: eradication and management, pp.
63–66. Occasional paper SSC no. 42. Gland, Switzerland: IUCN and
Auckland, New Zealand: CBB.
Carrion, V., Donlan, C.J., Campbell, K., Lavoie, C. and Cruz, F.
(2007). ‘Feral donkey (Equus asinus) eradications in the Galápagos’.
Biodiversity and Conservation 16: 437–445.
55
Island invasives: scaling up to meet the challenge. Ch 1A Rodents: Planning
Harper, G.A., Zabala, J. and Carrion, V. (2011). ‘Monitoring of a
population of Galápagos land iguanas (Conolophus subcristatus) during
a rat eradication using brodifacoum’. In: C.R. Veitch, M.N. Clout and
D.R. Towns (eds.) Island invasives: eradication and management, pp.
309–312. Occasional paper SSC no. 42. Gland, Switzerland: IUCN and
Auckland, New Zealand: CBB.
Holmes, N.D., Campbell, K.J., Keitt, B.S., Griffiths, R., Beek, J., Donlan,
C.J. and Broome, K.G. (2015). ‘Reporting costs for invasive vertebrate
eradications’. Biological Invasions 17: 2913–2925.
Island Conservation (2013). Floreana Island Ecological Restoration:
Rodent and Cat Eradication Feasibility Analysis V.6.1. (English
and Spanish). Unpublished report. Santa Cruz, California: Island
Conservation.
IUCN (2015). ‘The IUCN Red List of Threatened Species. Version 2015–
4’. <http://www.iucnredlist.org>. Accessed 19 November 2015.
Jensen, E.L., Tapia, W., Caccone, A. and Russello, M.A. (2015). ‘Genetics
of a head-start program to guide conservation of an endangered
Galápagos tortoise (Chelonoidis ephippium)’. Conservation Genetics
16: 823–832.
Jones, H.P., Tershy, B.R., Zavaleta, E.S., Croll, D.A., Keitt, B.S.,
Finklestein, M.E. and Howald, G.R. (2008). ‘Severity of the effects of
invasive rats on seabirds: A global review’. Conservation Biology 22:
16–26.
Jones, H.P., Holmes, N.D., Butchart, S.H.M., Tershy, B.R., Kappes,
P.J., Corkery, I., Aguirre-Muñoz, A., Armstrong, D.P., Bonnaud, E.,
Burbidge, A.A., Campbell, K., Courchamp, F., Cowan, P., Cuthbert,
R.J., Ebbert, S., Genovesi, P., Howald, G.R., Keitt, B.S., Kress, S.W.,
Miskelly, C.M., Oppel, S., Poncet, S., Rauzon, M.J., Rocamora, G.,
Russell, J.C., Samaniego-Herrera, A., Seddon, P.J., Spatz, D.R., Towns,
D.R. and Croll, D.A. (2016). ‘Invasive mammal eradication on islands
results in substantial conservation gains’. Proceedings of the National
Academy of Sciences 113: 4033–4038.
Keitt, B., Griffiths, R., Boudjelas, S., Broome, K., Cranwell, S., Millett,
J., Pitt, W. and Samaniego-Herrera, A. (2015). ‘Best practice guidelines
for rat eradication on tropical islands’. Biological Conservation 185:
17–26.
Kurle, C.M., Croll, D.A. and Tershy, B.R. (2008). ‘Introduced rats
indirectly change marine rocky intertidal communities from algae- to
invertebrate-dominated’. Proceedings of the National Academy of
Sciences 105: 3800–3804.
MacFarland, C.G., Villa, J. and Toro, B. (1974). ‘The Galápagos giant
tortoises (Geochelone elephantopus) part I: Status of the surviving
populations’. Biological Conservation 6: 118–133.
Morillo Manrique, G.E. (1992). ‘Estado Poblacional y Ecología
Reproductiva de la Tortuga Gigante de la Isla Pinzón (Geochelone
elephantopus ephippium), Galápagos’. Undergraduate degree in
biological sciences thesis. Quito, Ecuador: Pontificia Universidad
Católica del Ecuador.
Muñoz, E. (1990). ‘Seguimiento de la Campaña de Erradicación de Ratas
en la Isla Pinzón’. Carta Informativa (Estacion Cientifica Charles
Darwin, Servicio Parque Nacional Galápagos) 30: 2.
Oberg, E. and Campbell, K. (2012). Pinzon Lava Lizard Mitigation Plan:
Protocols for Capture and Captive Holding. Unpublished report. Puerto
Ayora, Galapagos Islands: Island Conservation.
Oberg, E., Alifano, A., Pott, M. and Wegmann, A. (2014). ‘Detecting
Rodents in the Presence of Land Crabs: Indicator Blocks Outperform
Standard Rodent Detection Devices at Palmyra Atoll’. In: R.M. Timm
and J.M. O’Brien (eds.) Proceedings of the 26th Vertebrate Pest Control
Conference, pp. 370–374. Davis, Califorina: University of California,
Davis.
Parent, C. and Campbell, K. (2012). Land Snail Mitigation Plan for
Pinzón Island During the Application of Brodifacoum Bait for Rodent
Eradication. Unpublished report. Berkley, California: University of
California, Berkley and Island Conservation.
Patton, J.L., Yang, S.Y. and Myers, P. (1975). ‘Genetic and morphologic
divergence among introduced rat populations (Rattus rattus) of the
Galapagos Archipelago, Ecuador’. Systematic Biology 24: 296–310.
Rueda, D., Campbell, K.J., Fisher, P., Cunninghame, F. and Ponder, J.B.
(2016). ‘Biologically significant residual persistence of brodifacoum
in reptiles following invasive rodent eradication, Galapagos Islands,
Ecuador’. Conservation Evidence 13: 38.
Russell, J.C. and Holmes, N.D. (2015). ‘Tropical island conservation: Rat
eradication for species recovery’. Biological Conservation 185: 1–7.
Steadman, D., Stafford, T., Donahue, D. and Jull, A. (1991). ‘Chronology
of Holocene vertebrate extinction in the Galápagos Islands’. Quaternary
Research 36: 126–133.
56
Tapia Aguilera, W., Málaga, J. and Gibbs, J.P. (2015). ‘Giant tortoises
hatch on Galapagos Island’. Nature 517: 271.
TIB Partners. (2014). ‘The Threatened Island Biodiversity Database
(TIB): Developed by Island Conservation, University of California
Santa Cruz Coastal Conservation Action Lab, Birdlife International and
IUCN Invasive Species Specialist Group. Version 2014.1’. <www.tib.
islandconservation.org>. Accessed 23 December 2016.
Torres-Carvajal, O., Torres‐Carvajal, O., Barnes, C.W., Pozo‐Andrade,
M.J. and Tapia, W. (2014). ‘Older than the islands: Origin and
diversification of Galápagos leaf‐toed geckos (Phyllodactylidae:
Phyllodactylus) by multiple colonizations’. Journal of Biogeography
41: 1883–1894.
Towns, D.R., Atkinson, I.A.E. and Daugherty, C.H. (2006). ‘Have
the harmful effects of introduced rats on islands been exaggerated?’
Biological Invasions 8: 863–891.
Weir, S.M., Yu, S., Talent, L.G., Maul, J.D., Anderson, T.A. and Salice,
C.J. (2015). ‘Improving reptile ecological risk assessment: Oral and
dermal toxicity of pesticides to a common lizard species (Sceloporus
occidentalis)’. Environmental Toxicology and Chemistry 34: 1778–
1786.
Willows-Munro, S., Dowler, R.C., Jarcho, M.R., Phillips, R.B., Snell,
H.L., Wilbert, T.R. and Edwards, C.W. (2016). ‘Cryptic diversity in
black rats Rattus rattus of the Galápagos Islands, Ecuador’. Ecology
and Evolution 6: 3721–3733.
A. Schiavini, J. Escobar, E. Curto and P. Jusim
Schiavini, A.; J. Escobar, E. Curto and P. Jusim. First results from a pilot programme for the eradication of beavers for environmental restoration in Tierra Del Fuego
First results from a pilot programme for the eradication of beavers for
environmental restoration in Tierra Del Fuego
A. Schiavini123, J. Escobar1, E. Curto4 and P. Jusim135
Centro Austral de Investigaciones Científicas, CONICET, Houssay 200, 9410 Ushuaia, Argentina. <aschiavini@
wcs.org>. 2 Wildlife Conservation Society, Argentinian Representation. 3Universidad Nacional de Tierra del Fuego,
Ushuaia, Argentina. 4Dirección General de Áreas Protegidas y Biodiversidad, San Martín 1401, Ushuaia, Argentina.
5
Universidad Nacional de Buenos Aires, Facultad de Ciencias Exactas y Naturales, Departamento de Ecología,
Genética y Evolución, Buenos Aires, Argentina.
1
Abstract A pilot project for the eradication of beavers (Castor canadensis) in Tierra del Fuego started as part of a binational agreement, signed between Argentina and Chile, to restore the affected environments. The project covers nine
pilot areas of different landscapes and land tenures in the Argentinian part of Isla Grande de Tierra del Fuego. We report on
the results from operations in the first of the pilot areas. From October 2016 to January 2017, ten trappers (named restorers
for advocacy purposes) used body-grip traps, snares and an air rifle, in a first phase, which included 2,237 trapping
nights and 1,168 trap-sets. Shooting efforts were not monitored. Traps were set for 1,401 trapping-nights and caught 175
beavers at a success rate of 12.5% (captures per trap night). Snares were set for 936 snare-nights and caught 22 beavers
at a success rate of 2.3%. Seven beavers were shot. Most beavers (65%) were removed during the first week of trapping
in the different watercourse sections. Stopping trapping for a week or more did not increase efficiency. From March
to May 2017 restorers removed 24 survivors and/or reinvaders, including 10 from two previously untrapped colonies.
Capture efficiency for this removal period was low for body-gripping traps but not for snares. The sex ratio of catches was
47% females to 53% males. The age structure of catches was 15% kits, 29% yearlings, 51% adults, with 4% not aged.
An estimated total of 41 colonies was trapped, giving an average of 5.6 animals per colony. After nominal eradication
was declared by restorers, 154 camera trapping nights were deployed to assess eradication success. Nine cameras (of 26
cameras used) detected beavers. Therefore, eradication was not achieved using the methods and efforts in the first part of
the pilot study. This highlights the need for more effort or the application of different techniques or trapping strategies.
For example, daily checking of traps may cause the animals to be cautious so, the next step in the programme will involve
exploring alternative trapping methods to reduce disturbance.
Keywords: Argentina, Castor canadensis, eradication programme, management, pilot study, trapping
INTRODUCTION
North American beavers (Castor canadensis) are semiaquatic and territorial rodents. They live in family groups
generally composed of two breeding adults, two yearlings
and two kits; the yearlings are forced to leave the natal
colony by the age of two (Lizarralde & Escobar, 1997;
McTaggart & Nelson, 2003). The family group controls a
group of adjacent dams, defending its territory from other
beavers. Each family group can build one or more lodges
(although they may also den in the river banks) and share
a single food cache.
In 1946, 20 beavers were introduced from Canada to
Tierra del Fuego, South America (Pietrek & Fasola, 2014),
with the aim of developing a fur industry. Beavers found
extensive suitable habitats, high availability of food, lack
of predators and unoccupied territory (Lizarralde, 2004).
These features allowed beavers to spread quickly throughout
Tierra del Fuego (Skewes, et al., 2006; Anderson, et al.,
2009). Several impacts on the environment of Tierra del
Fuego were reported and it was suggested that beavers
caused the largest landscape-level alteration to the region
since the Holocene (Anderson, et al., 2009). The most
obvious impacts are the reduction of the riparian vegetation
due to their activities, which includes the building of at least
70,000 dams in Argentinian Tierra del Fuego (Eljall, et al.,
2016), affecting at least 31,000 ha of forests, grasslands
and peat bogs (Henn, et al., 2016), as well as the fen areas
(Westbrook, et al., 2017). The beech forests of Tierra del
Fuego are not adapted to the impact of beavers, so their
impacts are long lasting (Anderson, et al., 2009). Their
dams also limit the dispersal of native fish and the water
in their dams changes the benthic communities, modifying
the macroinvertebrate assemblages by engineering changes
to the fluvial and riparian environment (Anderson, et al.,
2006). Beavers also modify the dynamics of the streams
by altering sedimentation (Vazquez, 2002; Martin, et al.,
2015). Last, but no less important, beavers impact the
economy by flooding roads and culverts, and affecting
ranching activity, reducing pastures by flooding as well as
affecting fences.
Attempts to control beavers by commercial hunting
during the 1990s and 2000s failed. Beavers were detected
in continental South America in the 1990s (Skewes, et al.,
2006; Wallem, et al., 2007; Schiavini, et al., 2008; Anderson,
et al., 2009), although recent dendrochronological evidence
takes their arrival date to 1968 (Graells, et al., 2015). The
presence of beavers in the continent raised alarm about the
possibility of their dispersal through the greater American
continent. In view of these issues, Argentina and Chile
started, in 2005, to discuss a change in strategy.
Eradication was deemed as feasible (Parkes, et al., 2008),
and adopted as a strategy by Argentina and Chile in 2008,
after signing a bi-national agreement for the restoration of
the southern ecosystems affected by the beaver (Malmierca,
et al., 2011). At present, both countries are performing pilot
projects, funded by the Global Environment Facility (GEF)
and national counterparts. The pilot project in Argentina
is under the umbrella of the major project “Strengthening
the Governance for the Protection of Biodiversity through
Formulation and Implementation of the National Strategy
for Invasive Exotic Species” GEF Project ID 4768. The
project runs from 2015 to 2019, covering nine pilot areas
of Tierra del Fuego.
The objectives of the project (Schiavini, et al., 2016) are
essentially to answer questions raised during the feasibility
study: building capacity, learning about technical and
organisational challenges of the process, showing the
environmental benefits of beaver removal, and deciding
the next steps between the two countries.
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 57–63. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
57
Island invasives: scaling up to meet the challenge. Ch 1A Rodents: Planning
Several research priorities and questions in relation to
the eradication of beavers are expected to be answered by
the pilot project:
● How much effort is needed to eradicate beavers and
to declare eradication on a small scale?
● What factors affect effectivity of trapping? The tools
used? The sequence of deployment? Learning by
beavers to avoid traps?
● What is the effort demanded for active surveillance
to avoid reinvasion?
● How to develop passive surveillance from society?
● Is the bureaucracy able to accommodate the
dynamics of eradication projects?
● Are any beavers found, after nominal eradication is
declared, likely to be survivors or reinvaders?
● Does the environment recover in a short time frame
after beaver removal?
The nine Argentinian pilot areas cover an area of 1,017
km2, with a range of 14–238 km2 (Fig. 1). In this paper, we
report the results of operations achieved in the first pilot
area, Esmeralda-Lasifashaj, and discuss the challenges
revealed for the larger major project.
MATERIALS AND METHODS
The Esmeralda-Lasifashaj area (54 km2) belongs
to the ecological region of the forest range (Collado,
2007). The landscape represents a U-shaped valley with
the valley bottom covered with Sphagnum peat bogs and
poorly drained mires (Figs 2 and 3). Slopes are covered
with southern beech forests (Nothofagus spp.) with the
vegetation line reaching about 700 m altitude. The main
valley is surrounded by eight lateral valleys. The area is
open to reinvasion as it has no geographical boundaries
that limit beaver dispersal, mainly from the west and
east. However, it was proposed as a pilot area for several
reasons: it is located only 20 km from Ushuaia city, is
Fig. 2 An aerial view of a series of beaver dams in the
bottom of the main valley of Esmeralda-Lasifashaj pilot
area.
Fig. 3 An aerial view of a series of beaver dams in the
Esmeralda-Lasifashaj pilot area, in an area of poor
drainage at the contact between peat bogs and forest.
Note the riparian forest impacted by cutting.
used by the public for recreation and tourism, and the area
holds a permanent cross-country ski trail, which is affected
by beavers. For these reasons, the area was selected as a
way of showing the environmental, social and economic
benefits of removing beavers.
Fig. 1 The Argentinian sector of Isla Grande de Tierra
del Fuego. Numbers refer to each pilot area. 1: Arroyo
Gamma. 2: Arroyo Asturiana. 3: Rio Malengüena. 4:
Río Mimica. 5: Arroyo Indio. 6: Esmeralda-Lasifashaj. 7:
Arroyo Grande, 8: Rio Pipo, 9: south of Tierra del Fuego
National Park. The black circle shows the location of the
pilot area Esmeralda-Lasifashaj.
58
The dams and lodges built by beavers are so
conspicuous that they can be identified in satellite images.
During the planning process, beaver dams and lodges were
mapped using Google Earth and integrated with the dams
identified by Eljall, et al. (2016). Then, 363 locations of
beaver activity were loaded into the GPS units used during
the operation (Garmin eTrex 20x), to be used as a general
guide for moving through the terrain to the areas impacted
by beavers.
Schiavini, et al.: Beavers in Tierra Del Fuego
The skills of the personnel involved in hunting should
include not only good trapping skills, but also the ability
to spend several days in the field in the harsh weather of
Tierra del Fuego and deliver good trapping data, essential
for assessing trapping efforts and eradication success.
Good, traditional trappers work with a focus on yield,
while personnel needed for eradication need to “look for
the last animal”. With this change in focus, 10 people
were selected and trained from a group of 39 people
interviewed. The training was performed by our own
personnel, staff from the National Parks Administration
and from the volunteer fire brigade. Training included the
use of trapping tools, data recording and first aid in the
field. The final selection included a combination of people
with previous trapping skills and people with good outdoor
abilities and a willingness to learn. Hunters are publicly
called “restorers” as a way of helping to advocate for
the final objective of the project, i.e. building the correct
conditions for environmental restoration by means of
beaver eradication.
The trapping equipment and tools were purchased
with advice from the Animal and Plant Health Inspection
Service of the USA, who also provided a handbook for
best-practice management. Two main tools are being
tested, body-gripping traps and non-powered cable devices
(snares), complemented with a PCP air rifle. The group
was commanded by a chief of operations and assisted by a
logistics officer.
The spatial and temporal progression of trapping differs
from traditional trapping operations, where hunters deploy
their tools progressively through the landscape, usually
in a regular or grid mode. Given that the trapping target
is located along watercourses or sectors of poor drainage
such as edges of peatlands, trapping effort follows these
landscape features. For planning purposes, the pilot area
was divided into sectors that brought together groups of
sections of channel or activity detected during planning.
Watercourse sections were trapped inside sectors until
"nominal" eradication was achieved, when trappers moved
to another watercourse section. After nominal eradication
of a sector, operations progressed to another sector.
At the watercourse section scale, trapping was made
according to decisions made by each restorer. A “trapset” is a trap (either a body-gripping trap or a snare) set
at a particular location and for a number of consecutive
trapping nights. Traps are usually set along watercourses
and near dams with beaver activity denoted by the girdling
of trees, fresh beaver trails, freshly gnawed branches in
front of the dams, castor mounds, and /or accumulation
of submerged tree branches with leaves. Traps are also set
either in trails or slides made by beavers or in purposemade openings at the front of the dam. The limits of beaver
colonies are not always evident. However, during fall and
winter, family groups gather at one lodge, so colonies are
more easily distinguishable. During spring and summer,
young animals disperse from their natal colonies, so the
movement of animals leads to colony boundaries being
confused. Also, traps can be set in the same place for more
than one night. After a number of trapping nights, hunters
noticed a reduction in their trapping efficiency, and at
some point, they decided that a "nominal" eradication was
achieved in this watercourse section and moved to another
section. As a result, data recording is quite different from
some other hunting and trapping operations, where hunters
either traverse a landscape searching for their prey, or traps
are set up more permanently at sites or along transects or
grids.
The records of trapping and yields attempted to reflect
the operation in great detail. An account of each trap set and
its subsequent outcome (set, capture, activation without
capture, not activated, removed) was recorded every day,
taking into account the use of both the body-gripping
traps and snares, with each one requiring daily checks
for humanitarian reasons. Each trap had a unique number
for identification. For data recording, an application
was built into Cybertracker software (Steventon, 2017),
allowing us to build a database with a record of each trap
(set, revision and retirement, with or without capture), as
well as ancillary data (e.g. location of placement, use of
attractant). The application is available upon request, or at
<http://cybertrackerwiki.org/index.php?title=Community_
applications>. For data recording, we used an outdoor
rugged tablet (Boolean A71, Boreal Technologies Inc). The
database can be transferred to Spreadsheets or to any GIS
system, as Cybertracker software can export shapefiles.
Restorers also carried a GPS unit for tracking their activity.
Operations ran from October 24, 2016 to January 31,
2017 in the first phase. From March 2 to May 15, 2017
(Fig. 4), the area was checked again to remove survivors/
invaders. Restorers worked mostly daily, during blocks of
five days or four trapping nights, commuting each day from
Ushuaia to the pilot area that is traversed by a National
Route highway. When restorers worked on the lateral
valleys, they camped for between three and five days. A
Robinson R44 helicopter was used to search for dams in
specific areas (Johnston & Windels, 2015) and to transport
personnel and equipment to lateral valleys. Two colonies
were left untrapped until the survivors-reinvaders removal
phase, as they were used by tourist operators during the
summer. Tour operators agreed as this would be the last
time they would be using these colonies for their tours.
Trapped animals were aged in the field, based on
external measurements, as kits, yearlings or adults, and
were sexed by detection of the baculum. Samples were
stored for accurate age determination, the breeding status
of females and for future assessment of the accuracy of
genetic tools to distinguish survivors from new invaders in
areas free of beavers.
For verification of eradication, an independent team
visited a sample of the watercourse sections, as restorers
declared the “nominal” eradication, between December 12,
2016, and May 24, 2017. Twenty-six camera traps were
set in front of artificial castor mounds with beaver lure at
a 1–2m distance from the camera and no more than 1m
from the water body. Each camera was placed at a height
of between 20 and 40 cm from the ground to capture full
images of beavers, and operated, on average, six days, with
a range of 3–10 days. Cameras were located both in the
main valley and in all the lateral valleys.
Fig. 4 Gantt chart including the first and second eradication
step (the arrows mark the first and last capture) and the
period of camera-trap vigilance (the arrows mark the first
and the last detection).
59
Island invasives: scaling up to meet the challenge. Ch 1A Rodents: Planning
As operations took place during spring and summer,
territorial limits were difficult to assess. The total number
of colonies was estimated based on the spatial distribution
of catches following Johnston & Windels (2015). The
Esmeralda-Lasifashaj area was divided into 18 sectors
for data analysis. All statistical analysis was performed in
Infostat (Rienzo, et al., 2016).
A monitoring plan measuring the environmental
benefits of removal of the beavers is being developed by
independent groups. The monitoring includes assessment
of the of trees that will not be subject to beaver cutting after
beaver removal, water quality, macroinvertebrate diversity,
metabolism of the watercourse and fish diversity.
RESULTS
Mop-up phase
From October 2016 to January 2017, restorers walked
2,930 km over the area (Fig. 5). For logistic purposes,
a helicopter was flown for nine hours. Trapping nights
were derived from trapping records by summing up trap
revisions and retirements. An additional 5% was added to
the effort for the offset for traps that were not checked daily
(based on an analysis of a subset of data).
Body-gripping traps were deployed in 715 trap sets,
yielding 1,401 trapping nights. Snare traps were deployed in
453 trap sets, yielding 936 trapping nights. This represents
a total of 2,337 trapping nights with 1,168 sets. Each trap
operated on average 1.97 nights with a range of one to four
nights. Rifle effort was not monitored, as it was employed
in an opportunistic fashion. A total of 197 beavers were
removed by trapping; 175 with body-gripping traps and
22 with snares, together with seven individuals that were
shot (Fig. 6). The trapping efficiency was 12.5% for bodygripping traps and 2.3% for snares, giving an average
efficiency of 9% for trapping.
The capture efficiency for each day of the working
blocks was assessed. For example: during the first day
of the working block the main activity was setting traps;
during the second day of the working block there were
443 reviews or removals and 46 catches, which gives an
Fig. 5 Tracks recorded by restorers in the pilot area
Esmeralda-Lasifashaj. Some tracks were not recorded
due to failure of the GPS units.
60
efficiency of 10.4%; on the sixth or seventh day, very little
field work was performed. This analysis was then limited
to reviews and retirement of traps from Tuesday to Friday.
Using a test of more than two proportions (Zar, 2010), the
null hypothesis of the difference of proportions revealed
no differences in catch efficiency over the different days of
the week (χ2 statistic, p=0.152, df=3). Therefore, restorers
did not reduce trapping efficiency through cumulative
disturbance by working consecutive days in a watercourse
segment, since the efficiency was similar between the
days of the working block. Another explanation might be
that even though beavers are more "relaxed" or “naïve”
to trapping early in the week (i.e. Tuesdays), restorers
gradually perform better in a particular area during the
week, compensating for the increasing caution of beavers
with improved trapping sets.
The effect of disturbance from hunting over the weeks
was also assessed, checking if leaving a section of the
watercourse without trapping for a week after trapping
for one or two weeks increases the trapping efficiency by
reducing the awareness of traps by the beavers. The scarce
data available for this analysis revealed no positive effect
by leaving a watercourse section without traps. The first
week of trapping in the watercourse´s section yielded 65%
of the beavers, giving an average capture efficiency higher
than the efficiency of the rest of the trapping days (10.3%
and 9% respectively; p <0.0001, difference of proportions
of Infostat). The capture efficiency did not differ between
the main valleys and the lateral valleys, comparing the 10
channel sections of the main valley with the six channel
sections of the lateral valleys (p=0.88).
A total of 151 traps (289 trapping nights) were set
with attractant (beaver hormone, food lure): 142 traps
(263 trapping nights) set with attractant, six traps (13 trap
nights) with attractant added after the first review and
three traps (13 trap nights) with attractant added after the
second review. These 151 traps produced 13 catches (289
trap nights), giving an efficiency of 4.5%. If only beaver
lure was considered, there were 10 catches in 89 traps (163
trapping nights), giving an efficiency of 6.1%.
The sex ratio of catches did not differ from 1:1 (p=0.26,
45% females vs 52% males, 3% unsexed). Also, the
Fig. 6 Catches recorded for the pilot area EsmeraldaLasifashaj
Schiavini, et al.: Beavers in Tierra Del Fuego
proportion of females did not differ between body-gripping
and snares (difference of proportions = 0.003; p≈1). The
age assessment made by restorers revealed an age structure
of 15% kits, 29% yearlings and 51% adults (4% not aged),
with similar proportions of age classes between bodygripping and snare traps (p=0.21). During the first days
of trapping along each watercourse section, 83% of the
sections yielded females while 50% of them yielded males
(marginally significant difference, p=0.043).
Survivors/reinvaders removal phase
During this phase, restorers walked 380 km deploying
735 trap nights (529 body-gripping and 206 snares). This
represented 23% and 31% of the previous walking and
trapping effort respectively. Twenty-four animals were
removed (22 with body-gripping traps and two with snares,
Fig. 6). From them, 10 animals came from the two colonies
left untrapped during the mop-up phase, and therefore 14
animals should be considered survivors-reinvaders. The
main valley provided most of the captures (83%), although
most of the trapping effort was focused there (83%).
Capture efficiency was 3.97% for body-gripping traps
and 0.97% for snares. Trapping efficiency, compared
with the first phase, was lower for body-gripping traps
(p<0.0001), but not for snares (p=0.20).
One of the two colonies originally left untrapped
yielded six males (one adult, four juveniles and one kit),
one female and one animal of unidentified sex. The second
colony yielded two males (one juvenile and one kit) and
two females (one adult and one kit).
The survivors/reinvaders captured consisted of 10 males
(six adults, three juveniles and one kit), six females (four
adults and two juveniles) and two animals of unidentified
sex. Five sites provided only males in this phase (including
a site with only three males). Attractant was used in only
seven of the sets, therefore the outcome was not analysed
due to the low sample size.
Population assessment
Analysis of the spatial distribution of catches concluded
that 41 colonies were trapped (plus a few recolonised
sites). The average number of beavers per colony was 5.6,
although this may exclude offspring, presumably dead
inside dens (see Discussion). The survivors/reinvaders
came from what we identified as 11 different colonies. As
beavers were dispersing during the time of operations, it
is difficult to compare the age/sex of the beavers caught
during the mop up with those captured during the survivor/
reinvader phase.
Non-target catches
Trap specificity was 90%. Non-target catches were
recorded only during the first phase. One culpeo fox
(Lycalopex culpaeus), and one upland goose (Chloephaga
picta) were released alive. Native species killed included
two spectacled ducks (Speculanas specularis), three
unidentified ducks and two upland geese (Chloephaga
picta). Exotic species captured included 10 muskrats
(Ondathra zibethicus) and one mink (Neovison vison)
which were killed and one grey fox (Lycalopex griseus)
which was released alive.
Eradication verification phase
The 26 cameras yielded a total of 154 camera trapping
nights. Nine cameras detected beavers after a period
between zero to five days (average two days), and 17
cameras did not detect animals after a period of between
three and 10 days (average six days). In addition, two
persons walked 155 km to check for signs of presence/
absence at the same time that the cameras were set. The
last beaver detection was confirmed on 24 May, 2017, nine
days after the last capture. Later in the year, from August
to October, surveys for survivors/reinvaders were planned
to continue.
DISCUSSION
This is the first eradication attempt for beavers from
one area in a short time frame. The finding of survivors/
reinvaders has two explanations, not mutually exclusive.
First, operations may not have reached the last individuals.
Second, the lack of physical barriers may ease the
movements of dispersing beavers from neighbouring
colonies. There had been two previous attempts at
beaver removal (Schiavini, et al., 2016). The first attempt
took place in the Tierra del Fuego National Park, where
a sustained control plan aimed to reduce the size of the
beaver colonies was followed by their complete removal
from 2,000 ha in 2011. The second attempt took place in the
provincial protected area of Reserva Provincial Corazón de
la Isla in 2014, where beavers were removed from 4,900 ha
in two months, although the project was discontinued for
financial reasons and this area has been included as one of
the pilot areas to be treated in the near future.
The estimated efficiency of body-gripping traps (12%)
was lower than the 22% reported by Lizarralde, et al.
(1996) for Tierra del Fuego. However, it must be noted
that the first estimate derives from tests for trapping aimed
at performance-oriented catches per number of captures.
In contrast, the complete removal of animals from one
area explains the lower trapping efficiency reported here.
Results from the next pilot areas will allow us to have a
broader view of the calculation.
The original trapping set and reviewing approach
required daily checking of traps. The presence of people
walking every day over the dams and dens, and in the
vicinity of colonies, can make beavers more “cautious”,
affecting the likelihood of removing the last animals.
The potential of beavers “learning” from disturbance and
becoming wary (sensu Morrison, et al., 2007) is a problem
for efficient eradication operations. Initial data analysis
did not reveal the cumulative effect of the presence of
the restorers in the capture efficiency. Neither did it find
beneficial effects of not setting traps for a number of days.
Because part of this pilot area was subject to different
intensities of trapping over the years, animals from there
may already have been cautious to human disturbance.
However, capture efficiency did not differ between areas
with more historical trapping effort (the main valleys)
and areas less accessible to trapping (the lateral valleys),
suggesting a lack of “memory” from previous trapping
disturbance in the area.
The next trials will give us a chance to answer the
questions raised above, and explore alternative trapping
effort schemes – for example, the exclusive use of bodygripping traps. This lethal tool would allow us to leave
traps unattended for several days, reducing the likelihood
of disturbance. However, the size and weight of bodygripping traps limit the number of traps a person can
transport and manage during a day, and the trade-off is that
trapping effort would be overestimated by this approach.
Nevertheless, the benefits of eradication would overcome
the uncertainty associated with estimating the eradication
effort.
The unexpectedly small number of kits present in
the catch may be because they were too young to leave
the dens. The trapping effort coincided with much of the
breeding season. Also, the lodges were not destroyed as part
61
Island invasives: scaling up to meet the challenge. Ch 1A Rodents: Planning
of the management process because we wanted to avoid
the escape of animals from their colonies. Consequently,
the most likely scenario is that kits remained in the den and
starved after the mother was captured. This poses a potential
constraint on the timing of future eradication attempts if
animal welfare issues are considered. Although the sex
ratio of the capture was even overall, females outnumbered
males by 1.66:1 (p = 0.043) early in the trapping of each
watercourse section, when 83% of females were caught.
These numbers support the idea of greater mobility of
females outside the lodges due to their maternal duties.
Trapping efficiency was lower during the survivor/
reinvader removal phase than during the first phase. This
is to be expected due to fewer remaining animals, and/
or because they may have “learnt” to be more cautious.
However, it is expected that reinvaders would not be as
cautious as survivors. More data are needed to explore
this issue. During the mop-up phase we could not identify
family colonies accurately from the spatial distribution
of catches, and consequently we could not discriminate
survivors from reinvaders based on their sex and/or age. It
is expected that genetic analyses would assist in identifying
survivors from reinvaders.
Of the 28 individuals captured during the survivor/
reinvader phase, 18 came from colonies previously trapped;
10 males, six females and two of undetermined sex. In
five sites only males were captured, and three males were
captured at one site. Most of the females were captured
at the same site next to males. The sex ratio of captures
for this phase did not differ significantly from 1:1 (p=0.3),
although male catch seemed to be larger. This could be a
reflection of greater male dispersion from neighbouring
areas, following source–sink dynamics.
Analysis of the spatial distribution of catches indicated
that 41 colonies were trapped, plus a few recolonised sites.
These values are in agreement with previously known
colony densities for the area. Lizarralde (1993) reported
4.72 colony sites/km, defining a colony site as “a pond,
or series of ponds used by a colony of beavers throughout
the year or years”, different than the usual definition of a
colony, that refers to a family group living in a series of
ponds and sharing a common food cache. Lizarralde &
Escobar (pers. comm. 2000) reported, for 1998 and 1999,
densities of 0.91 and 0.45 active colonies/km for the Olivia
River and of 0.67 and 0.52 colonies/km for the Lasifashaj
River, respectively. Schiavini, et al. (2016), reported
densities of 0.42 and 0.37 colonies/km for the Olivia and
Lasifashaj rivers in March 2010.
The estimated number of beavers per colony (5.6
individuals/colony) may underrepresent kits for the
reasons explained above. On the other hand, since trapping
occurred during a period of high juvenile mobility, the total
catch is likely to overestimate the number of individuals
per colony, since it would include animals from colonies
neighbouring the pilot area.
Eradication was not achieved during operations in
this first pilot area since beavers were detected by trapcameras during the verification phase and the removal and
revision work continued after the month of May. The main
reasons are likely to be that the area is open to reinvasion
and that trapping took place during a time of high juvenile
dispersal. In view of these preliminary results, a large-scale
eradication programme in the Isla Grande de Tierra del
Fuego (48,000 km2), must consider the spatial progression
of the operations, adjusted to the possibility of reinvasion
of the area under management and to the biological cycle of
beaver dispersal. Large-scale operations should be carried
out either in larger areas, covering areas with physical
barriers for reinvasion, and/or restorers should cover the
62
landscape in a more structured way. It is expected that the
experience gained in the rest of the trial will allow us to
adjust the strategy.
ACKNOWLEDGEMENTS
This is part of the project: Strengthening of Governance
for the Protection of Biodiversity Through the Formulation
and Implementation of the National Strategy on Invasive
Alien Species (NSIAS), funded by the Global Environment
Facility (GEF), National and Provincial counterparts and
executed by FAO (GCP/ARG/023/GFF). The work was
also funded by Wildlife Conservation Society, the Consejo
Nacional de Investigaciones Científicas y Técnicas and
by a grant from the Agencia Nacional de Promoción de la
Ciencia y Tecnología from Argentina (PICT 2934-2012).
REFERENCES
Anderson, C.B., Griffith, C.R., Rosemond, A.D., Rozzi, R. and Dollenz,
O. (2006). ‘The effects of invasive North American beavers on riparian
plant communities in Cape Horn, Chile: Do exotic beavers engineer
differently in sub-Antarctic ecosystems?’. Biological Conservation 128:
467–474.
Anderson, C.B., Pastur, G., Lencinas, M.V., Wallem, P.K., Moorman,
M.C. and Rosemond, A.D. (2009). ‘Do introduced North American
beavers Castor canadensis engineer differently in southern South
America? An overview with implications for restoration’. Mammal
Review 39: 33–52.
Collado, L. (2007). ‘La vegetación de Tierra del Fuego: de la estepa a la
selva’. In: J. Godoy Manríquez (ed.) Enciclopedia Patagonia Total, pp.
772–775. Buenos Aires: Barcelbaires Ediciones.
Eljall, A., Dieguez H., Menvielle M.F. and Hodara, K. (2016). ‘Distribución
y abundancia del castor en Tierra del Fuego, Argentina: Percepción
remota de la invasión de un ingeniero de los ecosistemas’ (Distribution
and abundance of the beaver in Tierra del Fuego, Argentina: Remote
perception of the invasion of an ecosystem engineer). Paper delivered
at the “VI Reunión Binacional de Ecología”. September 18–22 2016.
Iguazú, Argentina.
Graells, G., Corcoran, D. and Aravena, J.C. (2015). ‘Invasion of North
American beaver (Castor canadensis) in the province of Magallanes,
Southern Chile: Comparison between dating sites through interviews
with the local community and dendrochronology’. Revista Chilena de
Historia Natural 88(3): 1–9.
Henn, J.J., Anderson, C.B., and Pastur, G.M. (2016). ‘Landscape-level
impact and habitat factors associated with invasive beaver distribution
in Tierra del Fuego’. Biological Invasions 18: 1679–1688.
Johnston, C.A. and Windels, S.K. (2015). ‘Using beaver works to
estimate colony activity in boreal landscapes’. The Journal of Wildlife
Management 79: 1072–1080.
Lizarralde, M.S. (1993). ‘Current status of the introduced beaver (Castor
canadensis) population in Tierra del Fuego, Argentina’. Ambio 22:
351–358.
Lizarralde, M.S., Deferrari, G., Escobar, J., Álvarez, S. and Camilión,
C. (1996). ‘El trampeo de animales mamíferos: Un ensayo para el
trampeo de castor y rata almizclera con trampas Conibear 330-2 y
110-2’ (Trapping of mammals: An essay for the trapping of beaver and
muskrat with Conibear traps 330-2 and 110-2). Publicación Técnica de
la Dirección de Recursos Naturales de la Provincia de Tierra del Fuego
4: 1–12.
Lizarralde, M.S. and Escobar, J. (1997). ‘Avances Sobre la Ecología de los
Mamíferos Silvestres Introducidos en la Provincia de Tierra del Fuego’
(Advances in ecology of Introduced Wild Mammals in the Province of
Tierra del Fuego). La Fauna Silvestre en Tierra del Fuego, No. 27.
Lizarralde, M.S. (2004). ‘Invader species in Argentina: a review about
the beaver (Castor canadensis) population situation on Tierra del Fuego
ecosystem’. Interciencia 29: 352–358.
Malmierca, L., Menvielle, M.F., Ramadori, D., Saavedra, B., Saunders,
A., Soto Volkart N. and Schiavini A. (2011). ‘Eradication of beaver
(Castor canadensis), an ecosystem engineer and threat to Southern
Patagonia’. In: C.R. Veitch, M.N. Clout and D.R Towns (eds.) Island
invasives: eradication and management, pp. 87–90. Occasional paper
SSC no. 42. Gland, Switzerland: IUCN and Auckland, New Zealand:
CBB.
Martin, S.L., Jasinski, B.L., Kendall, A.D., Dahl, T.A. and Hyndman,
D.W. (2015). ‘Quantifying beaver dam dynamics and sediment retention
using aerial imagery, habitat characteristics, and economic drivers’.
Landscape Ecology 30: 1129–1144.
Schiavini, et al.: Beavers in Tierra Del Fuego
McTaggart, S.T. and Nelson, T.A. (2003). ‘Composition and demographics
of beaver (Castor canadensis) colonies in central Illinois’. The American
Midland Naturalist 150: 139–150.
Morrison, S.A., Macdonald, N., Walker, K., Lozier, L. and Shaw, M.R.
(2007). ‘Facing the dilemma at eradication’s end: uncertainty of absence
and the Lazarus effect’. Frontiers in Ecology and the Environment 5:
271–276.
Parkes, J.P., Paulson, J., Donlan, C.J. and Campbell, K. (2008). Control
of North American Beavers in Tierra del Fuego: Feasibility of
Eradication and Alternative Management Options. Landcare Research,
LC0708/084, Lincoln, NZ.
Pietrek, A.G. and Fasola, L. (2014). ‘Origin and history of the beaver
introduction in South America’. Mastozoología Neotropical 21: 355–
359.
Rienzo, J.A.D., Casanoves, F., Balzarini, M.G., Gonzalez, L., Tablada,
M. and Robledo, C. W. (2016). InfoStat versión 2016. <http://www.
infostat.com.ar>.
Schiavini, A., Cerezani, R., Silva, M.R., Loekemeyer, N., Malmierca,
L., Escobar, J., Deferrari G. and Lizarralde, M. (2008). ‘Gestión y
técnicas de control de castores en Argentina. Lecciones aprendidas’
(Management and control techniques of beavers in Argentina. Lessons
learned). In: B. Saavedra and C.A. Silva (eds.) Proceedings of Taller
Internacional para el Control de Castores en la Patagonia. Digital
edition, pp. 73–92. Chile: Wildlife Conservation Society. < http://www.
karukinkanatural.cl>.
Schiavini, A., Carranza, M.L., Deferrari, G., Escobar, J., Malmierca, L.
and Pietrek, A.G. (2016). ‘Erradicación de especies invasoras: Ciencia,
actitud y entendimiento. El castor en Tierra del Fuego’ (Eradication of
invasive species: Science, attitude and understanding). Mastozoología
Neotropical 23: 279–288.
Skewes, O., Gonzalez, F., Olave, R., Ávila, A., Vargas, V., Paulsen, P. and
König, H.E. (2006). ‘Abundance and distribution of American beaver,
Castor canadensis (Kuhl 1820), in Tierra del Fuego and Navarino
islands, Chile’. European Journal of Wildlife Research 52: 292–296.
Steventon, J. (2017). Cybertracker v. 3.443. Cybertracker Software.
<http://www.cybertracker.org>.
Wallem, P.K., Jones, C.G., Marquet, P.A. and Jaksic, F.M. (2007).
‘Identificación de los mecanismos subyacentes a la invasión de Castor
canadensis (Rodentia) en el archipiélago de Tierra del Fuego, Chile’
(Identification of the mechanisms underlying the invasion of Castor
canadensis (Rodentia) in the archipelago of Tierra del Fuego, Chile).
Revista Chilena de Historia Natural 80: 309–325.
Westbrook, C.J., Cooper, D.J. and Anderson, C.B. (2017). ‘Alteration
of hydrogeomorphic processes by invasive beavers in southern South
America’. Science of the Total Environment 574: 183–190.
Zar, J. H. (2010). Biostatistical Analysis. New Jersey, USA: Pearson
Education.
63
M. Serr, N. Heard and J. Godwin
Serr, M.; N. Heard and J. Godwin. Towards a genetic approach to invasive rodent eradications:
assessing reproductive competitiveness between wild and laboratory mice
Towards a genetic approach to invasive rodent eradications: assessing
reproductive competitiveness between wild and laboratory mice
M. Serr1,2, N. Heard1 and J. Godwin1,2
Department of Biological Sciences, N.C. State University, Raleigh, NC, USA. <meserr@ncsu.edu>. 2Genetic
Engineering and Society Center, N.C. State University, Raleigh, NC, USA.
1
Abstract House mice are significant invasive pests, particularly on islands without native mammalian predators. As
part of a multi-institutional project aimed at suppressing invasive mouse populations on islands, we aim to create heavily
male-biased sex ratios with the goal of causing the populations to crash. Effective implementation of this approach will
depend on engineered F1 wild-lab males being effective secondary invaders that can mate successfully. As a first step in
assessing this possibility, we are characterising genetic and behavioural differences between Mus musculus strains in terms
of mating and fecundity using wild house mice derived from an invasive population on the Farallon Islands (MmF), a
laboratory strain C57BL/6/129 (tw2), and F1 wild-lab offspring. Mice with the ‘t allele’ (tw2) have a naturally occurring gene
drive system. To assess fertility in F1 wild-lab crosses, tw2 males were paired with wild-derived females from the Farallon
Islands (MmF). Results of these matings indicate litter sizes are comparable but that weaned pup and adult wild-lab mice
are heavier in mass. Next, we initiated tests of male competitiveness using larger (3 m2) enclosures with enrichment. We
introduced both an MmF and a tw2-bearing male to two MmF females to assess mating outcomes. Preliminary results of
these experiments show none of the offspring carried the t-allele. However, performing the same experiment with F1 wildlab males instead of a full lab background resulted in 70% of offspring carrying the tw2 allele. This indicates that F1 wildlab males may be able to successfully compete and secondarily invade. It will be important in subsequent experiments
to determine what characteristics contribute to secondary invasion success. More generally, a better understanding of
characteristics contributing to overall success in increasingly complex and naturalistic environments will be critical in
determining the potential of a gene drive-based eradication approach for invasive mice on islands.
Keywords: competition, gene drive, invasive rodents, reproductive fitness, secondary invasion
INTRODUCTION
Invasive rodents are a key biodiversity threat for
the majority of the world’s islands and eradication
campaigns are often employed to prevent loss of island
endemics (Howald, et al., 2007; Campbell, et al., 2015).
These eradications employ rodenticides and have been
successful in eliminating invasive rodents from over 400
islands (DIISE, 2017). Rodenticides, however, have a
higher failure rate with mice (Mus musculus), as opposed
to rats (Rattus spp.) (MacKay, et al., 2007) and their use
on inhabited islands presents severe logistical challenges.
Additionally, rodenticides are not species-specific and
present animal welfare concerns (Campbell, et al., 2015).
These challenges have created a compelling need for
alternative approaches to rodent eradication.
One potentially promising approach to eliminating
invasive mice from islands would be to bias offspring
sex ratios by genetically engineering mice that produce
only one sex of offspring. Pairing this approach with a
genetic drive mechanism to spread this trait in an invasive
mouse population would be critical. Key first steps are to
understand the processes of reproductive competitiveness
and the capability of an introduced mouse to introgress into
established island populations, a process we are terming
‘secondary invasion’. The phenomenon of secondary
invasions and multiple introductions has been documented
in invasive brown anole (Anolis sagrei) populations with
evidence that secondary invasions may be frequent and
can add genetic variation to existing invasive populations
(Kolbe, et al., 2004). This secondary invader phenomenon
in house mice, however, is less well understood and genetic
evidence suggests variation in how this occurs across
islands. Some studies suggest that secondary invaders
may be frequent (Berry, et al., 1991; Bonhomme & Searle,
2012) while others suggest instead only single primary
invasions (Hardouin, et al., 2010; Gabriel, et al., 2015).
For rodent eradications these secondary invaders would be
carrying the gene drive and spread of this construct through
the population would be necessary for this approach to be
effective.
The development of the CRISPR/cas9 genome editing
technology has recently revolutionised genetic engineering
capabilities (Barrangou & Doudna, 2016). This has
increased interest in genetic pest management approaches
first conceptualised by Burt (2003) and built upon by other
authors more recently (Sinkins & Gould, 2006; Esvelt,
et al., 2014). Many of these approaches centre on gene
drives, systems in which a genetic construct producing a
desired phenotype (e.g., sex ratio manipulation, sterility) is
preferentially inherited by offspring. These are considered
‘selfish’ genetic elements because the majority of offspring
will inherit the genetic construct and it therefore could
spread quickly through a population (Lyttle, 1991). In mice,
a naturally occurring gene drive is found on chromosome
17 and is termed the t-allele (Silver & Buck, 1993). The
t-allele bearing sperm impact the motility of non-t bearing
sperm and this leads to an inheritance rate of greater than
90% for the t-allele (Bauer, et al., 2005; Baker, 2008).
Homozygosity of the t-allele (t/t) is typically lethal, but this
is not true of the variant form termed the tw2 allele, although
homozygosity does cause sterility (Levene & Dunn, 1961).
A gene drive-based approach to eradication could use
either a naturally occurring drive or a synthetic drive based
on CRISPR/Cas9 and functional drives with this technique
have now been demonstrated in mosquitoes, flies, and
yeast (Harris, et al., 2012; DiCarlo, et al., 2015; Gantz &
Bier, 2015); see also early contributions by Craig, et al.
(1960) and Hamilton (1967). Theoretically, by biasing
offspring sex ratios heavily towards males, reproduction
could be impaired and populations reduced. One way
this could be done would be to use the Sry gene. The Sry
gene is the key male determining factor in mammals and
is sufficient to start the cascade of events leading to male
development (Hacker, et al., 1995). Placing the Sry into
an autosome induces development that is phenotypically
male in mice that are genotypically XX (Koopman, et
al., 1991). Inserting Sry into a naturally-occurring gene
drive such as the t-allele or a synthetic drive based on
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
64
up to meet the challenge, pp. 64–70. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Serr, et al.: Reproductive competitiveness wild vs laboratory mice
Fig.1 Depiction of the Sry gene inserted into the tw2 gene
drive accompanied by a depiction of how the population
would bias to be all male.
CRISPR/Cas9 should create the potential for reduction of
an invasive mouse population by reducing and ultimately
potentially eliminating production of fertile females (Fig.
1; Backus & Gross, 2016; Piaggio, et al., 2017; Prowse,
et al., 2017). A synthetic gene drive using CRISPR/Cas9
could theoretically be employed in a similar way to ensure
all offspring inherit a feminising gene.
Regardless of the genetic mechanism employed, the
reproductive competitiveness and relative fitness of gene
drive carriers are likely to be important in determining the
success of any genetic approach to reducing invasive mouse
populations. Assessing reproductive competitiveness is the
focus of this study. Since mice introduced with a gene drive
mechanism would essentially be secondary invaders into
an established invasive mouse population, it is important
to better understand processes affecting introgression into
established demes. Mice are social animals and dominant
males will often hold and defend a territory (i.e. deme) that
provides reproductive access to reproductive females while
subordinate males do not (Bonhomme & Searle, 2012).
How incoming mice are able to successfully integrate into
island demes is not clear. If a gene drive approach was
used, then the incoming males would need to compete
with the resident island males for females. Competition
and aggression tend to occur between male mice when
there are limited territories (Gray & Hurst, 1998). Mouse
populations living non-commensally on islands can instead
exhibit an ‘island syndrome’ where they show important
differences with commensal populations. These can
include increases in body mass and, importantly in the
context of this study, lower levels of aggression (Adler
& Levins, 1994; Gray & Hurst, 1998; Cuthbert, et al.,
2016). In the 1980s, a study was conducted by capturing
house mice on the Orkney island of Eday (commensal)
and releasing them onto the Isle of May, which was
uninhabited by humans but had an established population
of non-commensal wild house mice (Berry, et al., 1991).
This study followed the spread of genetic markers unique
to Eday and found that these alleles moved quickly through
the Isle of May population (Berry, et al., 1991; Jones, et
al., 1995). Differences in aggression may relate to whether
the mice are living commensally or not, with evidence
indicating that commensalism and perhaps increased
density favours more aggressive individuals (Berry, et al.,
1991; Gray & Hurst, 1998). Overall, the limited studies to
date have strongly suggested that island mice may not be
as competitive as their mainland/commensal counterparts
(Mackintosh, 1981; Berry, et al., 1991; Gray & Hurst,
1998).
Secondary invader success may also depend on female
mate choice (Jones, et al., 1995). In terms of female mate
choice, there is evidence that females prefer the scent of
foreign males and are more likely to mate with unrelated
males (Roberts & Gosling, 2003; Frynta, et al., 2010).
Importantly, however, there is also evidence of female
choice favouring non-t haplotype carrier males or males
carrying a different t-haplotype variant (Lenington et al.,
1994; Manser, et al., 2015; Sutter & Lindholm, 2016).
The relative fitness of gene drive carriers will be a critical
determinant of effectiveness for this approach. Fitness
costs have been documented with other forms of the t-allele
(Carroll, et al., 2004; Lindholm, et al., 2013), but have
not been examined for the tw2 variant to our knowledge.
Information about the t-allele presence on islands and
modelling of population dynamics would help us further
understand the transmission of the Sry/tw2gene drive in
island mouse populations (Backus & Gross, 2016).
Central questions
A critical aspect of exploring gene drive eradication
techniques for island rodents is that the gene drive
originates in a mouse strain with a standard laboratory
background that is amenable to manipulation. Laboratory
mice, however, have been inbred and housed in nonhierarchical social conditions for generations (Morse,
2007; Fawcett, 2012) and they have also undergone both
deliberate and inadvertent selection under these captive
conditions (Fawcett, 2012). It is encouraging to note,
however, that wild-type behaviour can be restored quickly
by backcrossing with wild-derived mice to create wild-lab
crosses (Chalfin, et al., 2014). The central goals of this
study are to one i) confirm that a gene drive mechanism
can be bred into a wild background and ii) assess whether
key reproductive measures such as litter size, pup weight,
and adult weight are impacted in F1 and F2 wild-lab mice.
We also present preliminary findings regarding the success
of laboratory and F1 wild-lab males in competitive mating
situations.
MATERIALS AND METHODS
Strains of mice
These studies employed several different strains of
mice. A primary laboratory strain is C57BL/6J referred to
as (B6) mice. B6 mice are the most common strain of lab
mice and are easily manipulated genetically (Silver, 1995).
Compared to other laboratory strains B6 mice are considered
more defensive and aggressive in response to perceived
threats (Blanchard, et al., 2009). A second strain was
donated from the Threadgill lab at Texas A&M University.
These mice are of a mixed C57BL/6J and a 129S1/SvlmJ
(B6;129) background (hereafter referred to as “lab” strain)
and carry the tw2 variant of the t-allele. The tw2 variant stems
from a wild background but was brought into laboratory
stocks in 1946 (Dunn & Morgan, 1953). These mice are not
transgenic (no Sry inserted) and so heterozygotes produced
are either male or female. The tw2 allele is inherited by 95%
of offspring in matings with a tw2/+ sire (Kanavy & Serr,
2017). To maintain tw2 mice, B6 females are mated to males
heterozygous for the tw2 allele (t/+). The wild-derived mice
(MmF) we use are derived from wild progenitors captured
on Southeast Farallon Island, which is part of the Farallon
National Wildlife Refuge, located about 30 miles off the
coast of California near San Francisco (Farallon, 2013).
Invasive mice are the only terrestrial mammals on the
island currently (Schoenherr, et al., 1999; Farallon, 2013).
These mice show annual cyclic population variation with
peak densities in late summer and early fall. MmF mice
do not carry the t allele (Threadgill, pers. comm. 2013).
65
Island invasives: scaling up to meet the challenge. Ch 1A Rodents: Planning
Some of the highest mouse densities ever recorded in noncommensal habitats are seen on Southeast Farallon Island
at over 1300/ha (490/acre) (Farallon, 2013; Newser, 2013).
Their diet consists primarily of invertebrates (Jones, et
al., 2006). The Farallons mice pose direct threats to an
endemic invertebrate and indirect threats to native seabirds.
The USFWS plans for a future mouse eradication with
rodenticide (Farallon, 2013). We established a colony of
wild-derived Farallons mice (MmF) at NCSU in 2013 and
they are now 8th generation derived from the wild. These
Farallon mice serve as the ‘island mouse’ model being used
to form demes for testing the ability of secondary invaders
to establish and mate successfully.
All experiments were conducted under an approved
Institutional Animal Care and Use Committee protocol
at North Carolina State University between 2015–2017.
Mice were maintained in a temperature-controlled
greenhouse with natural lighting and conditions suitable
for reproduction year round. Animals were fed ad libitum
with 5058 LabDiet® and daily health and welfare checks
were performed. To test if mating between wild-derived
MmF females and laboratory males occurred pairs of lab
males with wild-derived MmF females were created and
housed in 29 cm wide × 40 cm long × 19 cm high standard
laboratory cages. Each cage contained aspen bedding,
natural cotton, a 15 cm PVC tube and black oil sunflower
seeds for enrichment. Mice were housed in this manner
with weekly cage changes. To minimise disturbance,
mice were transferred over to a clean cage using a 15 cm
PVC pipe whenever possible. Pups were weaned at the
mouse standard of 21 days +/- 3 days (Silver, 1995) and
the litter size, sex and weight of the pups in grams were
recorded. In addition, an ear punch or tail snip was taken
for genotyping. Pups were then weighed as adults and their
weight in grams was collected for nulliparous individuals
between the ages of 70–140 days.
Tests of male competition were conducted in seminatural enclosures. The size of these ‘arenas’ is 3 m2,
closely approximating the size of those used by Slade, et al.
(2014). To allow for formation of hierarchies and nesting,
we added enrichment and complexity in the form of sand,
bricks, plastic blocks (‘Legos’) supporting multilevel
clear Plexiglass structures, galvanized wire mesh (1.25 ×
1.25 cm mesh size), cardboard boxes and cardboard egg
cartons, and PVC pipes for environmental complexity. For
trials, all mice were placed into the arena at the same time.
Males were either weight matched to within 1 g (~5% of
body weight) or age matched within 8–10 weeks. All mice
used in the arenas were nulliparous and sexually mature.
Coloured ear tags as well as Clairol ‘Just For Men’ Black
Hair dye® was used to identify males. Trials included
combinations of MmF and tw2 males as well as MmF and
F1 wild-lab males. At the start of each trial, both males and
two non-related MmF females were placed into the arena
and filmed for one hour. During this hour, we counted
the number of bouts, chases and attempts to copulate, or
time in proximity with females, as a means of assessing
dominance. Animal welfare checks and monitoring for
pups were performed daily. Any pups born in the enclosures
were weaned at the standard of 21 days and a tissue sample
was collected for genotyping.
To confirm the presence of the tw2 haplotype, we used
a modified protocol where we amplified a portion the
Hba-ps4 (alpha-globin pseudogene-4) locus (Schimenti
& Hammer, 1990). The procedure uses a ‘dirty’ DNA
extraction developed by one of our collaborators at Texas
A&M University (Kanavy, pers. comm. 2016). Tissue is
collected and either a 2–3 mm tail snip or a 2 mm ear punch
is used. The ‘dirty’ DNA extraction buffer contains (50 μl
5 M NaOH, 4 μl 0.5 M EDTA, and 10 ml sterile water).
66
100μl of extraction buffer is then added to the tissue sample
and incubated at 95°C for 20 minutes. After vortexing and
cooling 5 μl of 1 M HEPES is added. The sample is then
centrifuged at 6,000 g for five minutes and 40 μl of DNA
is extracted from the top. DNA electrophoresis of PCR
products shows a distinct band at 198 bp for wildtype mice
(+/+) while tw2 homozygotes (t/t) display a band at 214 bp
and heterozygotes (t/+) show the presence of both bands.
Statistical analyses were conducted using JMP® Pro
12.2.0 (SAS) where 1-way ANOVAS were used for adult
weights and litter sizes. A mixed model ANOVA with the
fixed effect of litter size was used to separate litter size
from pup weight to compare pup weights. Next, post-hoc
analyses including orthogonal contrasts and Tukey’s HSD
tests were used to identify group differences. Litter sizes
and weights are presented as mean ± SEM.
RESULTS
Adult weights were taken for males and females.
Sample sizes for males were as follows: B6 (33), tw2 (24),
MmF (53), F1 (21), and F2 (22). For females sample
sizes were: B6 (19), tw2 (25), MmF (44) and F1, (23). The
average day of age that adult male weights were measured
at was the following: B6=80.43± 21.95; tw2 =90.43±27.65;
MmF 92.63±34.90; F1 93.03±19.46; and F2 89.48±28.27.
Similarly, for females the average day of age that the adult
weight was taken was: B6 91.24±28.99; tw2 88.66±24.09;
MmF 89.20±36.14; and F1 82.25±38.15. Adult weights
varied by strain and sex, F8,257=28.35, p<0.0001. In addition,
tw2carrying males (tw2, F1, F2) were larger than MmF males,
F=58.00, p<0.0001. Similarly, tw2 carrying females (tw2 and
F1) were larger than MmF females, F=7.75, p=0.0058 (Fig.
2). Due to space restrictions for husbandry, not enough F2
adult females had been reared to allow calculation of a
meaningful average for this group.
While litter size varied across strains F5,141=4.59,
p<0.0007, MmF, F1 and F2 wild-lab mice had litter sizes
that were comparable (Fig. 3). Sample sizes for litter size
were as follows: B6 (27); tw2 (20); MmF (45); MmF/B6
(19); F1 (21); and F2 (20). There were no differences
detected in the sex ratios for pups born, nor in the time of
gestation (data not shown).
Weaning weight was measured with a mixed model
ANOVA with litter size being a fixed effect. The samples
are as follows: B6 (18); tw2 (14); MmF (44); MmF/B6 (20);
Fig. 2 Adult weight by strain and sex. 1-way ANOVA,
F8,257=28.35, p=0.0001. Tukey’s HSD reveals significant
differences in weights indicated by letters. Sample sizes
are indicated in parentheses.
Serr, et al.: Reproductive competitiveness wild vs laboratory mice
Dominance was again based on initiation of chasing or
fighting with the MmF male and by time spent pursuing
or mating with females. When subordination did occur, the
subordinate males appeared to place themselves so as not to
be visible to the dominant individual. Behavioural results
are ongoing and were beyond the scope of this manuscript.
DISCUSSION
Fig. 3 Litter size by strain 1-way ANOVA, F5,141=4.59,
p=0.0007 indicates significant differences in litter
size across strains. Tukey’s HSD reveals significant
differences in weights indicated by letters. Sample sizes
are indicated in parentheses.
F1 (13); and F2 (20). Pup weaning weight was significantly
different across strains (F133,383 =13.922, p=0.0001) and the
highest weaning weights were found in F1 wild-lab F2 and
F1s respectively (Fig. 4). Highest mean weights at weaning
were 10.46± 0.40 g (F1) and 9.82± 0.33 g (F2).
In the arenas, preliminary trials of male competition
between tw2 males (laboratory strain) and MmF males
revealed no tw2 transmission based on genotyping (three
trials with 35 pups total). The tw2 male initially appeared
behaviourally dominant. He pursued females and chased the
MmF male away, but on subsequent days was subordinate
and tended to stay on top of the feeder out of view of the
MmF male. Preliminary trials with MmF males and F1
wild-lab males (eight trials, 47 pups) revealed strongly
contrasting results and a 70% transmission rate of the tw2
allele. Here, five of the eight litters did carry the tw2 with
31 of 33 pups from these litters confirmed. The F1 wild-lab
males appeared to be behaviourally dominant throughout
the trial in the same five trials where tw2 pups were produced.
Fig. 4 Weaning weight with fixed effect of litter size
Mixed Model, strain F5,119=4.98, p=0.0004., litter size
F1,117=12.46, p=0.0006. Tukey’s HSD reveal significant
differences in pup weights across strains, which is
indicated by letters. Sample sizes are indicated in
parentheses.
Relative fitness of gene drive carriers is likely to be
critical in determining the success of this approach (Burt,
2003; Manser, et al., 2015; Backus & Gross, 2016). Carriers
of gene drive constructs would need to be successful in
reproduction and reproductive competition if a genetic
approach to invasive rodent eradication is to be effective.
This work establishes some key initial conditions for
this success. First, lab mice and wild mice can breed and
produce viable litters. Second, while litters of the common
lab background tw2 mice were smaller than those of wildderived mice under the more naturalistic conditions used
in this study, the F1 wild-lab litters were of comparable
size to those having two wild-derived parents. Preliminary
results also suggest F1 wild-lab males may have strong
potential for reproductive success, a likely prerequisite for
initial introgression of gene drive constructs into an island
population.
This work established that wild-derived Farallon
females will mate with laboratory males in standard cages
and at similar frequencies to those seen in matings with
wild-derived males (M. Serr, unpublished data). This was
an initial but critical step in assessing reproductive output
across strains and in F1 wild-lab mice. Furthermore, results
indicate that both F1 wild-lab and F2 wild-lab backcrossed
mice have litter sizes that are not different statistically
than those of Farallon mice. This is important in terms of
fitness and exploring the effectiveness of using the Sry/
tw2 haplotype technique. It is also important to note that
the reverse holds true, as wild-derived MmF males will
mate with B6 and tw2 females in standard laboratory cages
although sample sizes are not adequate for statistically
meaningful comparisons. Results for pup weights indicate
F1 and F2 wild-lab pups have the greatest weight at weaning
and that this trend continues for adult males. Body size
affects male competitiveness in mice (Cunningham, et al.,
2013; Ruff, et al., 2017) with evidence suggesting that in
semi-natural enclosures male mice of intermediate weight
have the highest fitness (Ruff, et al., 2017). Matching mice
based on body size for our experiments helps rule out this
confounding factor, but for a potential gene drive release it
could be beneficial for the drive-bearing mice released to
weigh more than their wild counterparts.
Preliminary results from experiments in our larger
arenas examining competition suggested a surprising
pattern. Arena trials between MmF and tw2 males suggest
the wild-derived MmF males are dominant to pure
laboratory strain males, preventing transmission of the tw2
allele. Interestingly, however, weight-matched F1 wild-lab
males carrying the tw2 allele appear more competitive and
behaviourally dominant to MmF males. Consistent with
this observation, we find a 70% transmission rate of the tw2
allele in arena trials analysed thus far. In addition, of the
three trials where the F1 wild-lab male was not dominant
MmF litter sizes were small with two of the three litters
only having two pups each. This suggests that F1 wild-lab
males are strong competitors and that females will mate
with F1 wild-lab males even when both male types are
present. It will be important to conduct further arena trials
to assess this competitiveness with greater sample sizes
and also assess the competitiveness of F2 wild-lab males.
Other reproductive comparisons we are conducting include
measuring testes weights. Testes weight is correlated to
67
Island invasives: scaling up to meet the challenge. Ch 1A Rodents: Planning
total sperm count in mice (Le Roy, et al., 2001). Testes
weight can also predict dominance and mating success,
as mice with higher testicular weight are more likely to
initiate mating with females and attack behaviour towards
conspecific males (McKinney & Desjardin, 1973). Finally,
nesting behaviour and the temperature of nests will be
important to examine across wild-derived, laboratory
and F1 wild-lab mice as anecdotal observations suggest
poor nest construction by laboratory mice. This could be
important too because in cooler environments studies have
indicated that nest building behaviour, thermoregulation,
and fitness are correlated (Bult & Lynch, 1997).
Our results suggest that F1 wild-lab males could be
efficient secondary invaders. This would be generally
consistent with other studies from island populations
(Jones, et al., 1995; Bonhomme & Searle, 2012). However,
the situation may be different for females. Introduction of
mice from a commensal population on the Isle of Eday to
the Isle of May did not lead to the spread of mitochondrial
DNA markers, which are maternally inherited. These
results were in contrast to those for a Y-chromosome
marker and suggested females were unable to secondarily
establish while males did (Jones, et al., 1995). Studies from
other islands have corroborated these results in suggesting
no integrations of new maternal haplotypes from laterarriving females (Searle, et al., 2009; Gabriel, et al.,
2010; Jones & Searle, 2015). This apparent male-female
asymmetry in secondary establishment ability, however,
has not been experimentally tested. One approach to
addressing this apparent asymmetry is having records
of detailed behaviour in more naturalistic arena settings.
We have designed and implemented a Radio Frequency
Identification (RFID) system for tracking mouse
movements. RFID tracking allows collection of detailed
behavioural records and works well with wild house mice
(Weissbrod, et al., 2013; Auclair, et al., 2014). Behavioural
measures include time spent at nest boxes, running wheels
and food. With this information we can assess the number
of visits, the timing of visits, the number of interactions
and time in social contact with one another (König, et al.,
2015; Lopes, et al., 2016).
A second approach is to test the ability of different
strains to establish dominance in a standard test termed
resident-intruder paradigm. A previous study used this
approach to compare competitive behaviour in house
mice from the Isle of Eday and the mainland, finding the
island mice were significantly less aggressive (Gray &
Hurst, 1998). Expanding trials to increasingly complex
naturalistic experimental arenas should give insight into
the relative abilities of male and female mainland mice
to secondarily invade and therefore genetically introgress
into an island population.
Other factors that could influence the potential success
of an eradication effort include mate-choice and tolerance
of island conditions. Mate-choice factors known for mice
include odorant cues such as urinary proteins and ultrasonic
vocalisations (Hurst & Beynon, 2004; Blanchard, et al.,
2009; Musolf, et al., 2010). Island conditions and climate,
in particular, could be important influences on the success
of introduced mice (Berry, 1992). The island syndrome
for rodents predicts increased body mass and decreased
aggression (Adler & Levins, 1994; Gray & Hurst, 1998;
Cuthbert, et al., 2016). In addition, the island syndrome in
rodents is often associated with high population densities,
increased reproductive output, and increased survival rates
on islands (Adler & Levins, 1994). Mice are able to adapt
to new conditions and islands (Anderson, 1978; Bronson &
Pryor, 1983) and this adaptation could be critical for fitness,
although any construct would presumably be introgressed
into an island genetic background relatively quickly as it
68
spread. The population genetic structure of the mice already
present on an island would be critical for a synthetic gene
drive, but other factors including the rate of inbreeding,
ratio of reproductive males to females, and age structure
of the mouse population(s) might also prove important.
These are also likely to impact spread of either a synthetic
or natural drive like the t-haplotype considered here. In
regions with seasonality and temperature variations, mouse
populations often undergo a ‘boom and bust’ cycle, as seen
in the Farallon Islands, where the populations can erupt
only to die off with changes in temperature. The timing
of release of secondary invaders will likely be important
in these situations (Singleton, et al., 2005; Farallon, 2013;
Backus & Gross, 2016). Both natural and sexual selection
could influence the number of drive carrier mice that would
be required for eradication success. A study by Backus &
Gross (2016) modelling the Sry/tw2 gene drive found that
the relative fitness of the mice carrying the gene drive
determined whether multiple releases would be required.
Similarly, Prowse, et al. (2017) modelled synthetic gene
drives and found that a sex reversing drive would require
multiple releases to achieve eradication success.
The concept of reducing invasive mouse populations
through release of genetically-modified mice is still in the
early stages of development. Many key issues will need
to be addressed to determine whether this is a feasible
approach. We have shown that an island-derived wild
strain will mate with tw2-carrying laboratory males and
produce comparable litter sizes to those of wild–wild
matings. Promisingly, we also see that pup-weaning
weights are larger for F1 and F2 wild-lab mice and that
F1 wild-lab males may be stronger competitors in seminatural enclosures. A key future step will be to scale up
trials in arena size and environmental complexity. Larger
enclosures could be used with greater numbers of mice to
test whether a gene drive can spread under controlled and
biosecure, but naturalistic conditions. Finally, beyond the
technical issues discussed above, social license for any
environmental releases would be crucial (NASEM, 2016).
As gene drives are a new technology still in development,
input from the relevant publics and regulatory authorities
will be very important moving forward and this input is
also likely to lead to additional interesting and important
questions that developers will need to address.
REFERENCES
Adler, G.H. and Levins, R. (1994). ‘The island syndrome in rodent
populations’. The Quarterly Review of Biology 69(4): 473–490.
Anderson, P. K. (1978). ‘The serendipitous mouse’. Natural History 87:
38–43.
Auclair, Y., König, B., Ferrari, M., Perony, N. and Lindholm, A.K. (2014).
‘Nest attendance of lactating females in a wild house mouse population:
Benefits associated with communal nesting’. Animal Behaviour 92:
143–149.
Backus, G.A. and Gross, K. (2016). ‘Genetic engineering to eradicate
invasive mice on islands: Modeling the efficiency and ecological
impacts’. Ecosphere 7(12): e01589.
Baker, A.E. (2008). ‘Mendelian inheritance of t haplotypes in house
mouse (Mus musculus domesticus) field populations’. Genetics research
90: 331–339.
Barrangou, R. and Doudna, J.A. (2016). ‘Applications of CRISPR
technologies in research and beyond’. Nature Biotechnology 34(9):
933–941.
Bauer, H., Willert, J., Koschorz, B. and Herman, B. (2005). ‘The
t complex–encoded GTPase-activating Protein Tagap1 acts as a
transmission ratio distorter in mice’. Nature Genetics 37(9): 969–973.
Berry, R.J., Triggs, G.S., King, P., Nash, H.R. and Noble, L.R. (1991).
‘Hybridization and gene flow in house mice introduced into an existing
population on an island’. Journal of Zoology 225(4): 615–632.
Serr, et al.: Reproductive competitiveness wild vs laboratory mice
Berry, R.J. (1992). ‘The significance of island biotas’. Biological Journal
of the Linnean Society 46: 3–12.
Blanchard, C.D., Crawley, J.N., Arakawa, H. and Blanchard, R. (2009).
‘Social Behaviors in Wild and Laboratory Mice with a Special Emphasis
on the C57BL/6J inbred strain’. In: W.E. Crusio, F, Sluyter, R.T. Gerlai,
and S, Pietropaolo (eds.) Behavioral Genetics of the Mouse: Volume I,
pp. 205–217.Cambridge, MA: Cambridge University Press.
Bonhomme, F. and Searle, J. (2012). ‘House Mouse Phylogeography’.
In: M. Macholan (ed.) Evolution of the House Mouse, pp. 278–296.
Cambridge, MA: Cambridge University Press.
Bronson, F.H. and Pryor, S. (1983) ‘Ambient temperature and reproductive
success in rodents living at different latitudes’. Biology of Reproduction
29: 72–80.
Bult, A. and Lynch, C.B. (1997). ‘Nesting and fitness: Lifetime
reproductive success in house mice bidirectionally selected for
thermoregulatory nest-building behavior’. Behavior Genetics 27(3):
231–240.
Burt, A. (2003). ‘Site-specific selfish genes as tools for the control
and genetic engineering of natural populations’. Proceedings of the
Royal Society of London: Biological Sciences, 270(1518): 10.1098/
rspb.2002.2319.
Campbell, J.K., Beek, J., Eason, C.T., Glen, A.S., Godwin, J., Gould,
F., Holmes, N.D., Howald, G.R. and Madden, F.M. (2015). ‘The next
generation of rodent eradications: Innovative technologies and tools to
improve species specificity and increase their feasibility on islands’.
Biological Conservation 185: 47–58.
Carroll, L.S., Meagher, S., Morrison, L., Penn, D.J. and Potts, W.K.
(2004). ‘Fitness effects of a selfish gene (the Mus T complex) are
revealed in an ecological context’. Evolution 58: 1318–1328.
Chalfin, L., Dayan, M., Levy, D.R., Austad, S., Miller, R., Iraqi, F.,
Dulac, C. and Kimchi, T. (2014). ‘Mapping ecologically relevant social
behaviours by gene knockout in wild mice’. Nature Communications
4569: 10.1038/ncomms5569.
Craig, G., Hickey, W. and Van de Hey, R. (1960). ‘An inherited maleproducing factor in Aedes aegypti’. Science 132: 1887–1889.
Cunningham, C.B., Wanless, R.M., Angel, A., Burle, M.H., Hilton, G.M.,
Louw, H Visser, P., Wilson, J.W. and Ryan, P.G. (2013). ‘Competitive
ability in male house mice (Mus musculus): Genetic influences’.
Behavior genetics 43(20): 151–60.
Cuthbert, R.J., Wanless, R.M., Angel, A., Burle, M.H., Hilton, G.M.,
Louw, H Visser,P., Wilson, J.W, and Ryan, P.G. (2016). ‘Drivers of
predatory behavior and extreme size in house mice Mus musculus on
Gough Island’. Journal of Mammalogy 97(2): 533–544.
DiCarlo, J.E., Chavez, A., Dietz, S.L., Esvelt, K.M. and Church, G.M.
(2015). ‘Safeguarding CRISPR-Cas9 gene drives in yeast’. Nature
Biotechnology 33(12): 1250–1255.
DIISE (2017). ‘Database of Island Invasive Species Eradications,
developed by Island Conservation, Coastal Conservation Action
Laboratory UCSC, IUCN SSC Invasive Species Specialist Group,
University of Auckland and Landcare Research New Zealand’. <http://
diise.islandconservation.org>. Accessed 13 June 2017.
Dunn, L.C., and Morgan, W.C. (1953). ‘Alleles at a mutable locus found
in wild house mice (Mus musculus) Genetics 39: 391–402.
Esvelt, K.M., Smidler, A.L., Catteruccia, F. and Church, G.M. (2014).
‘Concerning RNA-guided gene drives for the alteration of wild
populations’. Elife 3: e03401.
Farallon (2013). Farallon National Wildlife Refuge South Farallon
Islands Invasive House Mouse Eradication Project: Revised Draft
Environmental Impact Statement. San Francisco Bay National Wildlife
Refuge Complex, Fremont, California: United States Fish and Wildlife
Service.
<https://www.fws.gov/uploadedFiles/Farallon_House_
Mouse_Eradication_Revised_DEIS_compressed.pdf>. Accessed 27
June 2017.
Fawcett, A. (2012). Animal Research Review Pane Guideline 22:
Guidelines for the Housing of Mice in Scientific Institutions. Australia:
Animal Welfare Unit, NSW Department of Primary Industries.
<http://www.animalethics.org.au/__data/assets/pdf_file/0004/249898/
Guideline 22-mousehousing.pdf/>. Accessed 27 June 2017.
Frynta, D., Volfová, R., Fraňková-Nováková, M. and Stejskal, V. (2010).
‘Oestrous females investigate the unfamiliar male more than the
familiar male in both commensal and non-commensal populations of
house mice’. Behavioural Processes 83(1): 54–60.
Gabriel, S.I., Jóhannesdóttir, F., Jones, E.P. and Searle, J.B. (2010).
‘Colonization, mouse-style’. BioMed Central Biology 8(131): 2–4.
Gabriel, S.I., Mathias, M.L. and Searle, J.B. (2015). ‘Of mice and the
“Age of Discovery”: The complex history of colonization of the
Azorean archipelago by the house mouse (Mus musculus) as revealed
by mitochondrial DNA variation’. Journal of Evolutionary Biology
28(1): 130–145.
Gantz, V.M. and Bier, E. (2015). ‘The mutagenic chain reaction: A
method for converting heterozygous to homozygous mutations’.
Science 348(6233): 442–444.
Gray, S. and Hurst, J. (1998). ‘Competitive behaviour in an island
population of house mice, Mus domesticus’. Animal Behaviour 56(5):
1291–1299.
Hacker, A., Capel, B., Goodfellow, P. and Lovell-Badge, R. (1995).
‘Expression of Sry, the mouse sex determining gene’. Development
121(6): 1603–1614.
Hamilton, W.D. (1967). ‘Extraordinary sex ratios’. Science 156(3774):
477–488.
Hardouin, E.A., Chapuis, J.L., Stevens, M.I., Bettine van Vuuren,
J., Quillfeldt, P. Scavetta, R.J., Teschke, M. and Tautz, D. (2010).
‘House mouse colonization patterns on the sub-Antarctic Kerguelen
Archipelago suggest singular primary invasions and resilience against
reinvasion’. BioMed Central Evolutionary Biology 10(1): 325.
Harris, A.F., McKemey, A.R., Nimmo, D., Curtis, Z., Black, I., Morgan,
S.A., Oviedo, M.N., Lacroix, R., Naish, N., Morrison, N.I., Collado,
A., Stevenson, J., Scaife, S. Dafa’alla, T., Fu, G., Phillips, C., Miles,
A., Raduan, N., Kelly, N., Beech, C., Donnelly, C.A., Petrie, W.D.
and Alphey, L. (2012). ‘Successful suppression of a field mosquito
population by sustained release of engineered male mosquitoes’. Nature
Biotechnology 30: 828–830.
Howald, G., Donlan, C., Galván, J., Russell, J., Parkes, J., Samaniego, A.,
Wang, Y., Veitch, D., Genovesi, P., Pascal, M., Saunders, A. and Tershy,
B. (2007). ‘Invasive rodent eradication on islands’. Conservation
Biology 21: 1258–1268.
Hurst, J.L. and Beynon, R.J. (2004). ‘Scent wars: The chemobiology of
competitive signaling in mice’. BioEssays, 26(12): 1288–1298.
Jones, C.S., Noble, L.R., Jones, J.S., Tegelström, H., Triggs, G.S. and
Berry, R.J. (1995). ‘Differential male genetic success determines gene
flow in an experimentally manipulated mouse population’. Proceedings
Biological Sciences260(1359): 251–256.
Jones, M.A., Golightly, M.A., Howald, G.R.A, Saminego, A. (2006).
‘Annual Variation in the Diet of House Mice (Mus musculus) on
Southeastern Farallon Island’. Unpublished Report. Arcata, California:
Department of Wildlife, Humboldt State University.
Jones, E.P. and Searle, J.B. (2015). ‘Differing Y chromosome versus
mitochondria DNA ancestry, phylogeography, and introgression in
the house mouse’. Biological Journal of the Linnaean Society115(2):
348–361.
Kanavy, D. and Serr, M. (2017). ‘Sry gene drive for rodent control: Reply
to Gemmell and Tompkins’. Trends in Ecology and Evolution 32(5):
315–316.
Kolbe, J.J., Richard, E.G., Schettino, L.R., Lara, A.C., Larson, A. and
Losos, J.B. (2004). ‘Genetic variation increases during biological
invasion by a Cuban lizard’. Nature 431(7005): 177–181.
König, B., Lindholm, A.K., Lopes, P.C., Dobay, A., Steinert, S. and
Buschmann, F.J.U. (2015). ‘A system for automatic recording of
social behavior in a free-living wild house mouse population’. Animal
Biotelemetry 3(1): 39.
Koopman, P., Gubbay, J., Vivian, N., Goodfellow, P. and Lovell-Badge, R.
(1991). ‘Male development of chromosomally female mice transgenic
for Sry’. Nature 351(6322): 117–21.
Lenington, S., Coopersmith, C., and Erhart, M. (1994). ‘Female
preference and variability among t-haplotypes in wild house mice’. The
American Naturalist, 143(5): 766–784.
Le Roy, I., Tordjman, S., Migliore-Samour, D., Degrelle, H. and
Roubertoux, P. (2001). ‘Genetic architecture of testis and seminal
vesicle weights in mice’. Genetics 158(1): 333–340.
Levene, H. and Dunn, L.C. (1961). ‘Population dynamics of a variant
t-allele in a confined population of wild house mice’. Evolution 15(4):
385–393.
Lindholm, A.K., Musolf, K., Weidt, A. and König, B. (2013). ‘Mate
choice for genetic compatibility in the house mouse’. Ecology and
Evolution 3(5): 1231–1247.
Lopes, P.C., Block, P. and König, B. (2016). ‘Infection-induced
behavioural changes reduce connectivity and the potential for disease
spread in wild mice contact networks’. Scientific Reports 6: 31790.
69
Island invasives: scaling up to meet the challenge. Ch 1A Rodents: Planning
Lyttle, T.W. (1991). ‘Segregation distorters’. Annual Review Genetics 25:
51–57.
MacKay, J.W.B., Russell, J.C. and Murphy, E.C. (2007). ‘Eradicating
House Mice from Islands: Successes, Failures, and the Way Forward’.
In: G.W. Witmer, W.C. Pitt and K.A. Fagerstone (eds.) Managing
Vertebrate Invasive Species: Proceedings of an International
Symposium, pp. 293–304. Fort Collins, Colorado, USA: USDA/APHIS
Wildlife Services, National Wildlife Research Center.
Mackintosh, J.H. (1981). ‘Behaviour of the house mouse’. Symposia of
the Zoological Society of London 47: 337–365.
Manser, A., König, B. and Lindholm, A.K. (2015). ‘Female house mice
avoid fertilization by t haplotype incompatible males in a mate choice
experiment’. Journal of Evolutionary Biology 28(1): 54–64.
McKinney, T.D. and Desjardin, C. (1973). ‘Postnatal development of
the testis, fighting behavior, and fertility in house mice’. Biological
Reproduction 9: 279–294.
Morse, H.I. (2007). ‘The Mouse in Biomedical Research, Volume 1’.
In: J. Fox, S. Barthold, M. Davisson, C. Newcomer, F. Quimby and A.
Smith (eds.). The Mouse in Biomedical Research: History, Wild Mice,
and Genetics, pp. 1–11. Burlington: Elsevier Inc.
Musolf, K., Hoffmann, F. and Penn, D.J. (2010). ‘Ultrasonic courtship
vocalizations in wild house mice, Mus musculus musculus’. Animal
Behaviour 79(3): 757–764.
National Academies of Sciences, Engineering, and Medicine. (2016).
Gene Drives on the Horizon: Advancing Science, Navigating
Uncertainty, and Aligning Research with Public Values. Washington,
DC: The National Academies Press.
Newser, K.S. (2013). ‘World’s most rodent-packed island has just 2
options’. USA Today. (9 August 2013). <www.usatoday.com/story/
news/nation/2013/08/19/newserrodent-island/2670717/>. Accessed 8
June 2017.
Piaggio, A.J., Segelbacher, G., Seddon, P.J., Alphey, L., Bennett, E.L.,
Carlson, R.H., Friedman, R.M., Kanavy, D., Phelan, R., Redford, K.H.,
Rosales, M., Slobodian, L. and Wheeler, K. (2017). ‘Is it time for
synthetic biodiversity conservation?’ Trends in Ecology and Evolution
32(2): 97–107.
Prowse, T.A.A., Cassey, P., Ross, J.V., Pfitzner, C., Wittmann, T. and
Thomas, P. (2017) ‘Dodging silver bullets: Good CRISPR gene-drive
design is critical for eradicating exotic vertebrates’. Proceedings of the
Royal Society B: Biological Sciences 284: 20170799.
Roberts, S.C. and Gosling, L.M. (2003). ‘Genetic similarity and quality
interact in mate choice decisions by female mice’. Nature Genetics
35(1): 103–106.
Ruff, J.S., Cornwall, D.H., Morrison, L.C., Cauceglia, J.W., Nelson,
A.C., Gaukler, S.M., Meagher, S., Carroll, L.S. and Potts, W.L. (2017).
‘Sexual selection constrains the body mass of male but not female
mice’. Ecology and Evolution 7(4): 1271–1275.
Schimenti, J. and Hammer, M. (1990). ‘Rapid identification of mouse t
haplotypes by PCR polymorphism(PCRP)’. Mouse Genome 87: 108.
Schoenherr, A., Feldmeth, R. and Emerson, M. (1999). Natural History of
The Islands of California. Los Angeles: University of California Press.
Searle, J.B., Jones, C.S., Gündüz, İ., Scascitelli, M., Jones, E.P., Herman,
J.S., Rambau, R.V., Noble, L.R., Berry, R.J., Giménez, M.D. and
Jóhannesdóttir, F. (2009). ‘Of mice and (Viking?) men: Phylogeography
of British and Irish house mice’. Proceedings. Biological Sciences
276(1655): 201–207.
Silver, L. and Buck, C. (1993). ‘The mouse t complex distorter-3 (Tcd3) locus and transmission ratio distortion’. Genetics Research 62(2):
133–137.
Silver, L. (1995). Mouse Genetics: Concepts and Applications. New
York: Oxford University Press.
Singleton, G.R., Brown, P.R., Pech, R.P., Jacob, J., Mutze, G. and Krebs,
C.J. (2005). ‘One hundred years of eruptions of house mice in Australia
– a natural biological curio’. Biological Journal of the Linnean Society
84(3): 617–627.
Sinkins, S.P. and Gould, F. (2006). ‘Gene drive systems for insect disease
vectors’. Nature Review Genetics 7: 427–435.
Slade, B., Parrott, M.L., Paproth, A., Magrath, M.J.L., Gillespie, G.R.
and Jessop, T.S. (2014). ‘Assortative mating among animals of captive
and wild origin following experimental conservation releases’. Biology
Letters 10: 20140656.
70
Sutter, A., and Lindholm, A.K. (2016). ‘Meiotic drive changes sperm
precedence patterns in house mice: potential for male alternative mating
tactics?’. BioMed Central Evolutionary Biology 16(1): 133.
Weissbrod, A., Shapiro, A., Vasserman, G., Edry, L., Dayan, M., Yitzhaky,
A., Hertzberg, L., Feinerman, O. and Kimchi, T. (2013). ‘Automated
long-term tracking and social behavioural phenotyping of animal
colonies within a semi-natural environment.’ Nature Communications
4: 2018.
D. Will, G. Howald, N. Holmes, R. Griffiths and C. Gill
Will, D.; G. Howald, N. Holmes, R. Griffiths and C. Gill. Considerations and consequences
when conducting aerial broadcast applications during rodent eradications
Considerations and consequences when conducting aerial broadcast
applications during rodent eradications
D. Will1, G. Howald1, N. Holmes1, R. Griffiths1 and C. Gill2
Island Conservation, 2100 Delaware Ave Suite 1 Santa Cruz, CA 95060, USA. <david.will@islandconservation.org>.
2
Coastal Conservation, 775 Abbington Lane, Tappen, British Columbia V0E 2X3, Canada.
1
Abstract Aerial broadcast application is currently one of the most common methods for conducting rodent eradications
on islands, particularly islands greater than 100 ha or with complex and difficult topography where access by ground
teams is difficult. Overall, aerial broadcast applications have a high success rate, but can be burdened by logistical,
regulatory, and environmental challenges. This is particularly true for islands where complex shorelines, sheer terrain, and
the interface with the marine environment pose additional risks and concerns. Using data collected during ten eradication
projects we investigate the influence that operational realities have on broadcast applications. We tested the association
between the amount of bait used and island size, topography, and the desire to reduce bait application into the marine
environment and then compared planned bait application to actual bait application quantities. Based on our results, islands
of decreasing size and increasing coastal complexity tended to use more bait than anticipated and experienced greater
variability in localised bait densities. During operations, we recommend analysing flight data to identify treated areas
with localised bait densities that fall below the target application rate. We recommend that areas with low localised bait
densities may result in biologically significant gaps that should receive an additional application of bait based on project
risk variables such as target home range size, non-target bait competitors, and alternative foods. We also recommend a
common language for discussing aerial broadcast applications and where future work can be done to improve operational
decision making.
Keywords: bait density, gaps, geographic information systems (GIS), island invasives, operational monitoring
INTRODUCTION
History of aerial broadcast applications
One of the primary principles for rodent eradication
is ensuring sufficient bait is distributed to every potential
rodent home range, so that every rodent is exposed to bait
for long enough to cause mortality (Bomford & O’Brien,
1995; Howald, et al., 2007). The aerial application of
rodenticide is one of the most common and effective ways
for eradicating rodents from islands (Holmes, et al., 2015).
Aerial broadcast techniques were first developed in the
1980s and methodology and principles were developed over
several decades as lessons learnt were applied to projects of
increasing size and complexity (Towns & Broome, 2003).
The first aerial applications relied on the use of modified
“monsoon” fire-fighting buckets slung beneath a helicopter
and flown by eye or guided by ground personnel. These
early projects were often successful in removing rodents,
despite difficulty in controlling application rates and the
need to use hand spreading to fill gaps (Garden, et. al.,
2019). The advent of specialised mechanical spreading
buckets to control bait application rates and distribution,
and global positioning systems (GPS) to guide pilots along
straight flight paths and record bait spread, revolutionised
aerial application techniques (Garden, et. al., 2019). These
changes allowed rodent bait to be delivered with far greater
precision over much larger areas, resulting in the successful
removal of rodents from islands larger than 10,000 ha
(Campbell 11,300 ha; Macquarie 12,800 ha; and South
Georgia 108,700 ha) (Broome, 2009; Russell & Broome,
2016; Martin & Richardson, 2017).
Aerial application principles
It is impossible to predict where all rodent home ranges
are and, because rodents are highly tolerant of a wide range
of habitat types, the whole island must be assumed to
support rodents, and the entire island is ultimately treated.
Bait application rates are set to ensure that bait is readily
available in all potential rodent home ranges and target bait
application rates are often informed by bait availability
trials (Pott, et al., 2015) or rates used on similar islands
that were previously successful (Broome, et al., 2014).
These rates are conservatively selected to ensure enough
bait for all the rodents on the islands while accounting for
loss and uptake by non-target competitors, like land crabs,
that reduce the amount of bait rodents are exposed to (Pott,
et al., 2015).
In general, one bait application rate is targeted across
an entire island because stratification increases complexity
and the risk of gaps in bait coverage (i.e. areas where some
rodents may not be exposed to bait), increasing the risk
of eradication failure (Keitt, et al., 2015). Subsequently,
projects are generally designed to use parallel flight lines
with 50% overlap between lines and additional parallel
flights along the coast to reduce the risk of gaps. Projects
may apply additional bait on steep cliffs because they have
a larger surface area (3D) than planar area (2D), resulting
in un-even bait distribution from bait falling downslope
(Broome, et al., 2014).
Challenges in aerial application
There are technical limitations of helicopters and
mechanical bait spreaders in applying bait over an entire
island. Operational realities, like wind, flight speed and
turning capabilities of the helicopters, steep terrain, and
unevenness of bait pellet distribution from the mechanical
spreader can impact bait placement on the ground, leading
to potential gaps in coverage. To ensure sufficient coverage
the pilot must reapply bait over potential gap areas, resulting
in locally increased bait densities where this additional
application partially overlaps with previous flight lines.
Additional complications arise when areas need to be
excluded from aerial application, such as human habitation,
inland water features or the marine environment. These
operational constraints tend to increase the total amount
of bait needed because additional overlapping flight lines
are required to ensure no gaps in coverage exist along edge
boundaries.
When trying to eradicate a rodent population, planning
tends to focus on targeting the worst-case scenario,
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 71–78. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
71
Island invasives: scaling up to meet the challenge. Ch 1A Rodents: Planning
ensuring that there are no gaps, meaning that bait overlaps
with the smallest known home range. However, it is not
well understood if applying less bait could constitute
biologically significant gaps where reduced bait availability
within a rodent’s home range decreases the likelihood of a
rodent being exposed to a lethal dose. The potential risks
posed by biologically significant gaps may be particularly
relevant on tropical islands, which tend to have more
non-target bait competitors and alternative food sources
(Holmes, et al., 2015), or when targeting multiple rodent
species.
These challenges have generally led to an “overengineering” approach to project design under the
perception that more bait increases the likelihood of
eradication success (Cromarty, et al., 2002); however,
higher bait use has trade-offs, such as increasing risk to nontarget species (Parkes, et al., 2011). We sought to improve
existing knowledge of what constitutes an ‘optimal’ bait
application rate, and what is a biologically relevant gap
in baiting. We examined ten projects to 1) understand
factors influencing the difference in bait use between what
was planned and what happened on the ground, and 2)
characterise localised bait application rates amongst these
ten projects to further understand what may constitute a
gap. Specifically, we asked:
What are the differences in total bait used between
three baiting scenarios and what physical and operational
factors are associated with these differences?
How does localised bait application rate vary and
how do areas estimated to be below the target application
compare to rodent home range size?
METHODS
Aerial application terminology
The target application rate is the desired rate of bait
deployment, in mass per unit area (e.g. kg/ha), to be applied
across the island. The target application rate is usually
based on bait availability trials and is set to maintain bait
availability for a certain period. The average application
rate is the total amount of bait distributed over an island
divided by the area of the island, in bait mass per unit area,
and is generally used for comparing eradication projects.
In general, bait is applied via a modified fertiliser
bucket underslung from a helicopter that distributes bait
either 360 degrees (full swath) or 180 degrees (half swath
or directional) from the bucket. Each bucket throws bait
pellets a certain distance as a function of bait product
size and weight and the speed of the distribution spinner.
The swath width is the effective distance that baits are
consistently sown, which is conservatively set during
calibration trials and less than the maximum distance the
bucket can throw bait.
The flow rate is the rate, in mass per unit time (i.e.
kg/sec), at which bait is distributed by the bucket. This
may be controlled in a variety of ways, depending on the
mechanics of a bucket, but is often controlled manually
with aperture discs that vary in size to restrict how much
bait can enter the spinner.
A bucket’s sow rate is the rate, in mass per unit area
(e.g. kg/ha), that bait is distributed from the bucket and is
a function of the helicopter’s flight speed and the bucket’s
flow rate. In general, a faster flight speed will decrease the
sow rate while a larger aperture disc will increase the sow
rate.
Using a GPS unit, bait is generally spread in parallel
flight lines employing planned overlap between flight lines
to reduce the possibility of gaps in bait coverage. When
72
using overlap the sow rate must be reduced to achieve the
desired target application rate (i.e. using a planned 50%
overlap buckets would require a sow rate of 5 kg/ha if
the target application rate was 10 kg/ha). In areas where
multiple flight line swaths overlap localised bait densities
achieved on the ground, in mass per unit area (e.g. kg/ha),
may be higher than the target application rate, and where
planned overlap does not occur bait densities may be lower
– resulting in undertreated areas. The GPS unit assists
helicopter pilots during bait application by indicating
deviance from the desired flight line and displaying the
current flight speed.
Supplemental bait is additional bait needed to fill
unplanned gaps, undertreated areas, or areas that require
additional treatment like steep cliffs or preferred habitat.
Contingency bait is bait held in reserve to replace spoiled
bait and is generally intended to be left unused at the end
of an operation.
Data from aerial broadcast eradication projects
Between 2008 and 2016 aerial baiting data were
collected and analysed across ten different rodent
eradication projects representing a variety of different
island habitats, sizes, strategies, outcomes, and regulatory
environments (Table 1). We used these data for our
analyses.
For each operation, an aerial baiting plan was developed
to estimate the total amount of bait required to complete
the operation. High resolution satellite imagery (<1 metre
per pixel) was acquired and used to estimate the island
area by digitising along the mean high-water mark at a
scale of 1:2,500. Treatment area estimates were generated
by calculating the area from hypothetical parallel flight
lines over the island with 50% overlap, using an estimated
effective swath width, and a single directional coastal
boundary swath, at half the estimated effective swath
width, along the coastline. For the nine projects with the
most conservative regulatory guidelines that restricted bait
entry into the marine environment, the start and end of the
parallel flight lines were brought in from the coast by half
of the estimated effective swath width, and an additional
coastal overlap buffer was estimated that overlapped with
the ends of the interior flight lines and the coastal swath.
On several operations, areas were identified for
supplemental treatment (e.g. steep cliffs) or exclusion
from aerial treatment (e.g. inland bodies of water, human
habitation) and treatment areas were calculated based on
the operational parameters. Steep cliff areas were estimated
by acquiring Digital Elevation Models (DEM) with a
resolution of 30 metres per pixel or better. Slope estimates
were calculated based on the DEM and used to identify
areas for additional treatment. Exclusion zones were
treated like the coastal edge, with flight line ends starting
and stopping at least half the effective swath width from
the exclusion boundary and a half swath flown around the
exclusion boundary to minimise gaps.
To estimate the total amount of bait required per
application treatment, area estimates were multiplied by
the sow rates required to achieve the target application rate
on the ground.
Aerial bait tracking
During each operation, a tracking worksheet was
completed that recorded detailed information about
each bucket load including: helicopter departure time,
helicopter arrival time, bucket type, disc size, bait placed
in the bucket, bait returned in the bucket, and cumulative
area treated as recorded by GPS (TracMap Ltd., Otago,
New Zealand iOS 1.7.2). For each bucket load the amount
2012
2016
2010
2011
2015
2015
2015
2010
2012
2013
2012
2015
2008
EC
EC
EC
FP
US
US
EC
US
FP
FP
FP
EC
US
CA
EC
NZ
US
Project
ActeonGambier
Desecheo
Desecheo
Rabida
Palmyra
ActeonGambier
ActeonGambier
ActeonGambier
Rabida
Wake
Murchison
& Faraday
Pinzon
Antipodes
Hawadax
Year
Tropical
Tropical
Tropical
Tropical
Tropical
Tropical
Tropical
Tropical
Tropical
Tropical
Tropical
Habitat
Coral
Coral
Volcanic
Volcanic
Volcanic
Coral
Coral
Volcanic
Volcanic
Volcanic
Volcanic
Pinzon
Antipodes
Hawadax
Tropical
Volcanic
Temperate Volcanic
Temperate Volcanic
Rabida
Tropical
Volcanic
Wake+
Tropical
Coral
Murchison and Faraday Temperate Volcanic
Temoe
Tenarunga
Desecheo*
Desecheo
Bartolome
Palmyra
Vahanga
Gambier (Manui,
Makaroa, Kamaka*)
Plaza Norte
Plaza Sur
Bainbridges, Sombrero
Chino and Beagles
Block
*Project failed to remove invasive rats.
+Project successfully removed one of two species of rats.
2015
2010
2012
2010
Country
Rabida
Pinzon
Rabida
Island type
5
10
35
Max. elev. (m)
430
366
340
340
5
190
5
10
220
220
80
10
10
140
Table 1 Operational data analysed to evaluate factors influencing total bait used.
Size (ha)
1,789.6
2,129.5
2,900.0
499.0
637.0
806.0
431.0
425.0
117.1
117.1
129.3
234.9
380.0
88.0
8.4
14.8
72.2
Coastline (km)
18.3
33.7
43.8
11.1
39.6
40.9
36.9
23.5
7.9
7.9
7.2
64.0
25.0
8.1
2.1
2.7
9.2
No. of flight lines
408
661
1096
117
776
607
431
214
85
159
68
664
288
103
9
25
116
Supplemental
treatment
Cliff
Cliff
Coast
Coast
None
None
None
None
None
Cliff
Coast
None
None
None
None
None
Coast
Coastal overlap
buffer
TRUE
TRUE
TRUE
TRUE
TRUE
TRUE
FALSE
FALSE
TRUE
TRUE
TRUE
TRUE
FALSE
FALSE
FALSE
FALSE
FALSE
Target rate (kg/ha)
6.0
16.0
6.0
6.0
15.2
16.0
24.0
24.0
20.0
34.0
6.0
80.0
24.0
24.0
6.0
6.0
6.0
∆uniform.planned
9.2
7.1
5.3
19.1
15.6
11.8
28.5
16.5
3.2
36.7
47.9
12.2
21.1
34.1
0.0
31.1
41.7
∆planned.actual
2.5
20.8
1.2
6.5
5.6
23.4
13.7
13.5
-6.0
-1.7
2.0
-6.5
6.1
34.1
82.5
7.2
70.6
∆uniform.actual
12.0
29.3
6.5
26.9
22.1
38.0
46.1
32.2
-3.0
34.5
50.8
5.0
28.5
79.8
82.5
40.5
141.8
Will, et al.: Considerations of aerial broadcast for rodents
73
Island invasives: scaling up to meet the challenge. Ch 1A Rodents: Planning
of bait used and area treated were calculated and used to
estimate the sow rate achieved. The sow rate information
was relayed to project management and the pilot to inform
decisions about adjusting disc size or flight speed to ensure
a consistent sow rate.
Flight line data were downloaded from the GPS unit
and treatment polygons (spatial representations of where
bait was spread) were estimated by buffering the flight
lines based on the effective swath width calculated during
operational bucket calibration. Using the helicopter times
from the tracking worksheet and the times recorded in the
flight line GPS data, the recorded sow rates were assigned
to treatment polygons (now spatial representations of
where bait was spread and at what rate it was applied).
GIS-derived bait density estimates were calculated by
dissolving overlapping treatment polygons into new nonoverlapping polygons and summing the sow rates of the
overlapping parts. Bait density estimates and flight line
maps were reviewed to identify gaps or undertreated areas.
the planned amount of bait and the actual amount of bait
used (∆planned.actual) (Fig. 1). The variable ‘∆uniform.
planned’ represents the change in bait required between a
uniform application and what was planned to account for
physical island characteristics and strategy decisions such
as reducing bait into the marine environment. The variable
‘∆planned.actual’ represents the difference between
what was planned and what happened on the day due to
operational realities, such as unexpected deviations in sow
rates and flight path.
We used Spearman’s rank correlation to explore
relationships between variables we thought may influence
planning (∆uniform.planned) and how the reality of the day
affects the plan (∆planned.actual). To minimize the chance
of Type I error resulting from multiple pairwise tests, we
chose to test four variables (elevation, size, coastline, and
flight lines) for correlation with the two bait use scenarios
and penalized the p-value by a factor of 8 (P<0.0006).
The remaining explanatory variables were expressed as
boxplots and compared with exploratory statistics.
Factors associated with difference in planned and
actual bait amounts used
To evaluate what factors were associated with the
total bait applied during an aerial operation, aerial baiting
data from the ten projects, comprising 17 different island
blocks, were collated (Table 1). In some cases, an island
block comprised of multiple treatment units (i.e. motu
or small islets) that were treated collectively. There were
three projects where multiple island blocks were treated
as independent units. Ten exploratory factors thought to
be associated with differences in aerial bait applications
were collected for each application (Table 2). Only the
first application for each island block was analysed as
they were the most comparable because the amount of bait
applied during the second application could be influenced
by the amount of bait used during the first application, bait
availability monitoring data, or the use of supplemental
bait.
These ten factors were compared against two response
variables, referred to as bait use scenarios: 1) the percent
change between the bait amount in a hypothetical
uniform scenario, where bait is evenly distributed across
an island, and the planned amount of bait to be used
(∆uniform.planned); and 2) the percent change between
Fig. 1 Examples of bait use scenarios used and normalised
percent change, delta, in bait use between scenarios.
Uniform represents bait needed in an even distribution of
bait across island area, planned represents bait needed
based on predicted flights paths and overlap, and actual
represents bait used.
Table 2 Explanatory physical and operational characteristics evaluated.
Physical characteristic
Country
Habitat
Island type
Max. elevation
Size
Coastline
Operational characteristic
Target rate
Number of flight lines
Supplemental treatment
Coastal overlap buffer
74
Definition
Country operation was implemented in
Tropical or temperate
Volcanic or coral atoll
Maximum elevation in meters as a proxy for steep terrain
Size of area to be treated (km2)
Length of coastline to be treated (km)
Definition
Minimum application rate expected to be achieved on the
ground, in some cases the coast and interior had different
expected rates. The lowest expected rate was selected
The total number of flight lines flown
Cliff, coast, or none to represent areas that received
additional treatment above the target application rate
True or false if the coastal overlap buffer strategy was
employed to reduce bait into the marine environment
Will, et al.: Considerations of aerial broadcast for rodents
Variability in bait densities achieved
To evaluate the distribution of bait densities (kg/ha)
we used GIS-derived bait density estimates from the 17
island blocks. For each island block, the area and estimated
bait density of each polygon representing the bait density
achieved on the ground from overlapping swaths was
exported. Polygon areas representing areas smaller than
100 square meters (0.01 ha) were excluded as they were
smaller than what is commonly considered a significant
gap. For each island block, a bait density distribution was
calculated to represent the total amount of island area
treated at each bait density rate (e.g. 10 ha at 5 kg/ha)
by summing the areas of treatment polygons at each bait
density rate. To normalise bait density distributions across
island blocks, values were represented as a percentage of
the target application rate (e.g. 50% = half, 100% = target
rate, 200% = twice target) and areas as a percentage of the
total island area treated.
RESULTS
Factors associated with differences in planned and
actual bait amounts used
The 10 projects analysed most often occurred in tropical
regions (7 projects, 14 of 17 island blocks) and ranged in
size from 8–2,900 ha, and 5–430 m in elevation. Target
application rates ranged from 6 to 80 kg/ha, supplemental
baiting used in seven island blocks, the coastal buffer
overlap strategy used in 10 island blocks, and the number
of flight lines flown spanned 9–1096.
On average, 20% more bait than the uniform scenario
(∆uniform.planned) was planned for, and 16% more bait
was used than planned (∆planned.actual). The variables
∆uniform.planned and ∆planned.actual showed no
associations with the four factors investigated (elevation,
size, coastline length and the number of flight lines)
(Table 3). Median results of the 17 island blocks were 380
hectares, 214 flight lines, 80 m in elevation, and an 18 km
coastline. Although no statistical correlation was evident
among the island blocks and these factors, those blocks
below the median showed a mean ∆uniform.planned
that was two to three times greater than blocks above the
median, suggesting that compared to larger islands in our
sample, planning on smaller islands typically identified
proportionally more bait than a uniform distribution. The
same trend is evident for ∆planned.actual with mean values
for islands blocks below the median being one and a half
times greater than above the median, showing that among
our sample, smaller islands used proportionally more bait
than planned for, compared to larger islands. Of the 14
tropical island blocks, five were on coral atolls, and these
generally had a higher number of flight lines (M = 474.6,
SD=215.1), compared to volcanic islands (M=121.1,
SD=110.5).
Three island blocks conducted in the United States
(Desecheo 2012, 2016, and Palmyra) had a negative
∆planned.actual, putting less bait on the ground than
planned. The 10 blocks using the coastal buffer overlap
strategy to reduce bait into the marine environment showed,
on average, lower ∆uniform.planned and ∆planned.actual
compared to blocks that did not use this strategy.
Analysis of bait density estimates
On average, 5.1% (SD=3.8) of total island area received
less than 50% of the target application rate and 0.8%
(SD=1.6) of total island area received more than 400%
of target (Fig. 2). The GIS derived bait density estimate
polygons representing these areas had an average size of
0.12 ha (SD=0.2) and 0.03 ha (SD=0.04), respectively.
Bait density estimates from each island block are shown
in Fig. 3. Bait density estimates of less than 75% of the
target application rate were visually compared against
grids representing conservative minimum (0.01 ha) and
average (0.1 ha) rodent habitats on tropical islands based
on available literature (Fig. 4).
DISCUSSION
Factors associated with differences between planned
and actual bait amounts used
From a statistical perspective, the sample size we used
is considered small (n=17), and less than ideal because it
was opportunistically collected (and not experimentally
collated). From a conservation practitioners perspective,
the opportunity to compare 17 different island
blocks consistently is rare, and a positive example of
collaboratively working to answer questions relevant
across the island restoration field. A key result from our
investigation is that projects planned to use 20% more bait
than the hypothetical uniform application and used 16%
more bait than planned, suggesting that simply estimating
bait quantities by multiplying island area by target
application rate is insufficient to judge how much bait
will be needed. On average, the percent change between
the planned amount of bait and actual bait used was less
than the percent change between the hypothetical uniform
Table 3 Spearman’s correlation and p-value of factors
thought to influence bait use. Factors were considered
associated with changes in bait use if Rho > 0.3 and
p-value<0.006. Negative numbers represent a negative
association (i.e. as one factor increases the other
decreases) and positive numbers a positive association
(i.e. as one factor increases so does the other).
Scenario
∆uniform.
planned
∆planned.actual
Factor
Max. elevation
Size
Coastline
Flight lines
Max. elevation
Size
Coastline
Flight lines
Rho p-value
-0.193
0.458
-0.389
0.123
-0.288
0.262
-0.311
0.224
-0.252
0.328
-0.212
0.414
-0.185
0.477
-0.272
0.291
Fig. 2 Box plot of bait densities across projects represented
as % of total island area treated vs % of target application
rate.
75
Island invasives: scaling up to meet the challenge. Ch 1A Rodents: Planning
Fig. 3 Estimated bait density distributions per island block as % of total island area treated vs % of target application rate.
Projects are grouped into multi (i.e. multiple treatment areas), single (i.e. single continuous treatment area > 100 ha),
and small (i.e. < 100 ha).
amount of bait and actual bait used, suggesting that the
aerial bait plans were more accurate at forecasting bait use
but still underestimated actual bait required.
In general, smaller islands and islands with shorter
coastlines, less elevation, or fewer flight lines planned to use,
and actually used, a higher percentage of bait than projects
on larger islands or those with more topography or flight
lines. This suggests that small islands use proportionally
more bait and that projects with fewer flight lines are more
complex. While coastline length and maximum elevation
were likely not differentiated enough from island size to
detect a significant difference, the four-fold increase in
the number of flights flown on tropical coral atolls, which
have two coastal edges (lagoon and ocean), compared
to volcanic tropical blocks suggests coastal complexity
needs to be factored into planning. While the number of
flight lines is also related to size, projects with fewer flight
lines also have less room for error and could experience
greater variability in bucket sow rates. Small islands may
be able to improve bait applications, and reduce unplanned
bait use, by employing strategies to increase the number
of flight lines flown such as flying the parallel flight lines
twice per application at half the target rate.
Fig. 4 GIS derived bait density estimates showing shaded
areas less than 75% of the target application rate against
potential (A) minimum (0.01 ha) and (B) average (0.1 ha)
home range sizes from literature review.
76
Perhaps the most interesting result was that projects
implemented in the United States were the only projects,
on average, to use less bait than planned. The United
States has a complex regulatory environment, and aerial
broadcasts are required to stay below permitted application
rates. When implementing an eradication, projects in the
United States had to balance the desire to achieve the
desired target application rate with not exceeding the
permitted application rate. Striking this balance resulted
in projects using less bait than planned, particularly when
the desired target rate was close to the permitted rate.
This suggests that regulators should be involved early in
the planning process so that regulatory approval can be
sought to maximise project success. A single permitted
application rate, such as the one designated on the bait
product registration in the United States, is not necessarily
appropriate for every project and, when appropriate,
Will, et al.: Considerations of aerial broadcast for rodents
projects should develop site-specific operational strategies
using the best available science. Regulatory bodies should
review these strategies and recommended application rates
on a case by case basis.
Bait application variability and consequences
It is noteworthy that, on average, 5% of the total island
block area had bait density estimates less than half of the
target application and 0.4% had bait density estimates
greater than four times the target application rate (Fig.
2). This suggests a relatively high degree of precision
in balancing the risk of failure (i.e. low localised bait
densities) with unintended environmental impacts (i.e.
high localised bait densities). Comparing the distributions
of bait densities, larger (> 100 ha) single unit island blocks
(i.e. those treated as a single contiguous unit: Antipodes,
Desecheo, Hawadax, Pinzon, and Rabida) generally tended
to have less bait density variability, with more than 60% of
total island area near the target application rate, compared
to smaller islands (< 100 ha) or island blocks consisting
of multiple treatment units (Galapagos Islets, Gambier,
Palmyra, Plaza Norte, Temoe, Tenarunga, Vahanga, and
Wake) with less than 50% of total island area near the target
application rate (Fig. 3). This is logical given that large or
single unit island blocks have longer flight lines with which
to “settle” into consistent sow rates and a smaller coast
to size ratio resulting in fewer overlapping flights. Island
blocks with multiple treatment units, particularly tropical
coral atolls (Palmyra, Temoe, Tenarunga, and Vahanga),
tended to have a higher percentage of total island area
with localised bait densities more than twice the target
application. These tropical coral atolls have more coastline
for their size than other similarly sized islands, and thus
the consequences of the flight line overlap necessary to
minimise the chance of gaps near the coastline (i.e. higher
localised bait densities) are more pronounced. This result
underscores the trade-offs of ensuring complete coverage
along complex coastlines.
Examinations of the two failed projects (Desecheo in
2012 and Wake) suggested low bait densities as one of the
potential reasons contributing to failure (Derek Brown,
pers. comm.). The bait density distribution of the failed
2012 Desecheo project shows a larger proportion of the
island experienced localised bait densities less than half
the target application rate during the first application
(7.7%), compared to similar islands. Desecheo had a high
abundance of non-target bait competitors (up to 833 crabs/
ha) and bait availability plots in one habitat showed bait
availability reaching zero within two to three nights (Will,
et al., 2019). It seems likely that areas with localised bait
densities less than half the target application rate would
have experienced even less bait availability. On Wake, the
bait density distribution shows a smaller proportion of the
island achieved less than half the target application rate
(1.4%) compared to similar islands, but bait density maps
also show fewer flight lines extending up to the coastal
edge and the presence of bait gaps on the beaches between
the mean high-water mark and predominant vegetation.
These observations may be instructive in improving the
quality of future bait applications, suggesting that future
applications consider applying additional bait (i.e. reapply)
in areas with bait densities identified to be less than the
target application rate and consider minimising the amount
of untreated coastal edge on tropical coral atolls. These
are areas where bait availability may be much less than
expected and may not be immediately obvious when
inspecting flight line maps. It is impossible to know if
these improvements would have resulted in successful
eradication attempts on Desecheo in 2012 and on Wake,
but they would have removed questions about the quality
of bait coverage as a possible contributor to eradication
failure.
What is a significant biological gap?
Comparing actual bait densities achieved to the
hypothetically smallest potential home range size can
be instructive in informing risk tolerance for future
operations. Rodent home ranges are highly variable, but
amongst R. rattus have been recorded ranging from 0.012
to > 10 ha (Shiels, et al., 2016; Harper & Bunbury, 2015).
It is in the smaller home ranges, particularly for breeding
female rodents, where localised deficiencies in bait density
present the highest risk of a rodent not being exposed to a
lethal dose of bait (i.e. undertreated areas). We considered
any areas that achieved less than 75% of the target
application rate to be undertreated, which were generally
the result of flight line deviation and were small (<0.1 ha)
and irregularly-shaped (hundreds of meters long and <20
m wide). Despite their size and shape, these undertreated
areas were still large enough to encompass most, if not all,
of an assumed 0.01 ha potential minimum home range, but
a minority of an assumed 0.1 ha average home range (Fig.
4). This suggests that, at the extreme, localised deficiencies
in bait density could make bait less available than expected
in entire potential rodent habitats where rodents have small
home ranges.
Whether localised bait density deficiencies (i.e.
undertreated areas) constitute biologically significant gaps
is largely a consequence of toxicology, rodent biology
and island ecology, and is project dependent. Ultimately,
projects should anticipate that localised deficiencies in
bait density are almost inevitable and determine what
risk they pose to project success based on site specific
conditions. In the presence of alternative foods and nontarget bait competitors, or on islands targeting species
with small home range sizes or multiple rodent species,
areas that receive less than the target application rate
could result in insufficient bait availability and constitute
biologically significant gaps that pose a risk to project
success. Where biologically significant gaps are a concern,
project managers can either choose to increase the target
application rate to increase the localised bait density of
undertreated areas or set area size and application rate
thresholds (i.e. 0.1 ha or larger with a bait density less than
half the target application rate) to reapply bait.
Improving aerial application data analysis
Although GIS-derived bait density estimates provide
a useful metric for identifying gaps or undertreated areas,
they do have limitations and assumptions. A key limitation
is they are not a direct measure of bait on the ground,
and where possible on-ground measures of bait density,
particularly with adequate sample size, can improve
these data. Further, GIS-derived bait density estimates
assume a) that flight speed is constant along the length of
a flight line, b) bait pellet distribution across a swath is
even, and c) wind has no impact on bait spread. A novel
model called the Numerical Estimation of Rodenticide
Dispersal (NERD) models these assumptions to generate a
probability density function describing bait density and was
successfully implemented on several projects in Mexico
(Rojas-Mayoral, pers. comm.; Samaniego-Herrera, et al.,
2017). These sorts of novel models are highly appropriate
on high risk islands targeting species where smaller rodent
home ranges may be anticipated (e.g. tropical islands
where breeding may be expected). However, regardless of
the analysis method used, managers are advised to trust
in the broader rodent eradication principles and exercise
caution to avoid overanalysing baiting data.
77
Island invasives: scaling up to meet the challenge. Ch 1A Rodents: Planning
RECOMMENDATIONS
ACKNOWLEDGEMENTS
In summary, we propose the following recommendations
to improve the planning and implementation of aerial
broadcast applications for eradications.
We would like to acknowledge implementation partners
BirdLife International, Coastal Conservation, Island
Conservation, New Zealand Department of Conservation,
Parque Nacional Galápagos, Parks Canada Agency and
the Council of the Haida Nation, SOP Manu, The Nature
Conservancy, U.S. Air Force, U.S. Fish and Wildlife
Service and many funders for their dedication to planning
and conducting rodent eradication projects, without which
this work would not be possible.
Use high-resolution satellite imagery to estimate
island size. Accurate estimates of operational area will
improve estimates of the amount of bait needed and reduce
the risk of having insufficient bait or the cost penalties of
transporting and disposing of too much bait.
Create predicted flight plans to inform planning
and estimate bait requirements. Multiplying island area
by target rate is not an accurate estimate of bait needed.
Including flight line overlap between parallel swaths and
at the coastal boundary will improve accuracy of bait
total estimates, reducing the chance of having too little
bait. Additionally, predicted flight plans are useful in
communicating the desired strategy.
Projects should plan for small islands to use
more bait than anticipated and islands with complex
coastlines to experience greater variability in bait
densities. Coral atolls with lagoons have two coastal
edges, which increases complexity, and should plan to use
more bait and experience more areas of high localised bait
densities. Small, complex projects should plan on ordering
additional bait to treat gaps and compensate for areas of
unplanned overlap.
Managers of projects on small islands should
consider modifying operational strategies to reduce
using additional bait. Increasing the number of flight
lines by flying the island twice per application (with
sowing rates adjusted to achieve the target rate), reducing
the amount of bait in the bucket per load to reduce the
percentage of island covered per flight, or conducting
additional calibration runs to ensure consistency should be
considered.
Projects should seek site-specific regulatory approval
that maximises project success. A single permitted
application rate is not sufficient to maximise success for
all projects. Where appropriate, application rates should
be tailored to site-specific conditions and be informed by
the best available science. Additionally, to ensure clarity,
projects should seek site-specific approval to implement
predicted flight plans that describe the application rates
and strategy needed to maximise project success. This
is particularly relevant for projects implemented in the
United States.
Use bait density estimates to identify areas treated
below the target application rate. Tracking sowing rates
achieved per load and assigning them to flight line data
improves the understanding of bait coverage and allows
managers to identify undertreated areas. Novel or high-risk
projects should also consider using more fine scale bait
density modelling approaches like NERD (Rojas-Mayoral,
pers. comm.).
Projects should set gap size tolerances and
application rate thresholds to match project risk
variables. Clarify in advance of the project what constitutes
a biologically relevant baiting gap based on what is known
about the target species, island habitat, topography, and
presence of non-target bait competitors. It is highly likely
that a broadcast application will result in less than expected
bait availability in the smallest potential rodent home
ranges. For rodents with small home ranges, or tropical
islands with high densities of non-target bait competitors,
alternative food sources, or multiple rodent species a
smaller gap size or higher application rate threshold may
be warranted.
78
REFERENCES
Bomford, M. and O’Brien, O. (1995). ‘Eradication or control for
vertebrate pests?’ Wildlife Society Bulletin 23(2): 249–255.
Broome, K.G. (2009). ‘Beyond Kapiti – A decade of invasive rodent
eradications from New Zealand islands’. Biodiversity 10: 14–24.
Broome, K.G., Golding, C., Brown, K., Corson, P. and Bell, P. (2014).
Rat Eradication Using Aerial Baiting: Current Agreed Best Practice
Used in New Zealand (Version 3.1). Internal document DOC-29396.
Wellington, New Zealand: New Zealand Department of Conservation.
Cromarty, P., Broome, K., Cox, A., Empson, R., Hutchinson, W. and
McFadden, I. (2002). ‘Eradication Planning for Invasive Alien Animal
Species on Islands – the Approach Developed by the New Zealand
Department of Conservation’. In: C.R. Veitch and M.N. Clout (eds.)
Turning the tide: the eradication of invasive species, pp. 85–91.
Occasional Paper SSC no. 28. IUCN SSC Invasive Species Specialist
Group. IUCN, Gland, Switzerland and Cambridge, UK.
Garden, P., McClelland, P. and Broome, K. (2019). ‘The history of aerial
application of rodenticide in New Zealand.’ In: C.R. Veitch, M.N. Clout,
A.R. Martin, J.C. Russell and C.J. West (eds.) Island invasives: scaling
up to meet the challenge, pp. 114–119. Occasional Paper SSC no. 62.
Gland, Switzerland: IUCN.
Harper, G.A. and Bunbury, N. (2015). ‘Invasive rats on tropical islands:
Their population biology and impacts on native species’. Global
Ecology and Conservation 3: 607–627.
Holmes, N.D., Griffiths, R., Pott, M., Alifano, A., Will, D., Wegmann, A.S.
and Russell, J.C. (2015). ‘Factors associated with rodent eradication
failure’. Biological Conservation 185: 8–16.
Howald, G., Donlan, C.J., Galván, J.P., Russell, J.C., Parkes, J.,
Samaniego, A., Wang, Y., Veitch, D., Genovesi, P., Pascal, M., Saunders,
A. and Tershy, B. (2007). ‘Invasive rodent eradication on islands’.
Conservation Biology 21(5): 1258–1268.
Keitt, B., Griffiths, R., Boudjelas, S., Broome, K., Cranwell, S., Millett,
J., Pitt, W. and Samaniego-Herrera, A. (2015). ‘Best practice guidelines
for rat eradication on tropical islands’. Biological Conservation 185:
17–26.
Martin, A.R. and Richardson, M.G. (2017). ‘Rodent eradication scaled
up: Clearing rats and mice from South Georgia’. Oryx:10.1017/
S003060531700028X.
Parkes, J., Fisher, P. and Forrester, G. (2011). ‘Diagnosing the cause of
failure to eradicate introduced rodents on islands: Brodifacoum versus
diphacinone and method of bait delivery’. Conservation Evidence 8:
100–106.
Pott, M., Wegmann, A.S., Griffiths, R., Samaniego-Herrera, A., Cuthbert,
R.J., Brooke, M.d.L., Pitt, W.C., Berentsen, A.R., Holmes, N.D.,
Howald, G.R., Ramos-Rendón, K. and Russell, J.C. (2015). ‘Improving
the odds: Assessing bait availability before rodent eradications to aid in
selecting bait application rates’. Biological Conservation 185: 27–35.
Russell, J.C. and Broome, K.G. (2016). ‘Fifty years of rodent eradications
in New Zealand: Another decade of advances’. New Zealand Journal of
Ecology 40(2): 197–204.
Samaniego-Herrera, A., Aguirre-Muñoz, A., Bedolla-Guzmán, Y.,
Cárdenas-Tapia, A., Félix-Lizárraga, M., Méndez-Sánchez, F., ReinaPonce, O., Rojas-Mayoral, E. and Torres-García, F. (2017). ‘Eradicating
invasive rodents from wet and dry tropical islands in Mexico’. Oryx:
10.1017/s0030605316001150.
Shiels, A.B., Pitt, W.C., Sugihara, R.T. and Witmer, G.W. (2014). ‘Biology
and impacts of Pacific island invasive species. 11. Rattus rattus, the
black rat (Rodentia: Muridae)’. Pacific Science 68(2): 145–184.
Towns, D.R. and Broome, K.G. (2003). ‘From small Maria to massive
Campbell: Forty years of rat eradications from New Zealand islands’.
New Zealand Journal of Zoology 30(4): 377–398.
Will, D., Swinnerton, K.J., Silander, S., Keitt, B., Griffiths, R., Howald, G.,
Figuerola-Hernandez, C. and Herrera Giraldo, J.L. (2019). ‘Applying
lessons learned from tropical rodent eradications: a second attempt to
remove invasive rats from Desecheo National Wildlife Refuge, Puerto
Rico’. In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J.
West (eds.) Island invasives: scaling up to meet the challenge, pp. 154–
161. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
E.A. Bell
Bell, E.A. It’s not all up in the air: the development and use of ground-based rat eradication techniques in the UK
It’s not all up in the air: the development and use of ground-based rat
eradication techniques in the UK
E.A. Bell
Wildlife Management International Ltd, P.O. Box 607, Blenheim 7201, New Zealand, <biz@wmil.co.nz>.
Abstract Eradication techniques using ground-based devices were developed in New Zealand in the early 1970s to target
invasive rodents. Since then, different bait station designs, monitoring tools and rodenticide baits have been developed,
and changes in field techniques have improved and streamlined these operations. The use of these techniques has been
taken around the world to eradicate rodents from islands. Eradication technology has moved rapidly from ground-based
bait station operations to aerial application of rodenticides. However, regulations, presence of and attitudes of islandcommunities and presence of a variety of non-target species precludes the aerial application of rodenticides on islands in
many countries. As such, ground-based operations are the only option available to many agencies for the eradication of
invasive rodents from islands. It is important to recognise that the use of ground-based operations should be a valid option
during the assessment phase of any eradication proposal even in countries that can legally apply bait from the air; in many
instances the use of ground-based techniques can be as economic and rapid. The use of ground-based operations can also
facilitate opportunities for in-depth monitoring of both target and non-target species. Using examples of the techniques
and developments used in five ground-based rat eradication operations from the UK demonstrates how these methods
can be used safely and successfully around the world, even on islands in the order of hundreds of hectares and those with
communities.
Keywords: bait station, ground-based, inhabited, island, rodenticide
INTRODUCTION
Albeit unanticipated, the eradications of rats from
Rouzic Island, France in 1951 and Maria Island, New
Zealand in 1960 were the first successful rat eradication
operations on islands anywhere in the world (Towns &
Broome, 2003; Lorvelec & Pascal, 2005; Howald, et al.,
2007). These unintentional eradications spurred efforts in
New Zealand to develop and perfect eradication techniques
(Cromarty, et al., 2002; Thomas & Taylor, 2002; Towns
& Broome, 2003). Between 1965 and 1986, New Zealand
wildlife managers, ecologists and scientists used a range of
experimentally designed operations to determine the best
methods to consistently, successfully eradicate rats from
islands (Cromarty, et al., 2002; Towns & Broome, 2003).
Seabirds and other native species on islands are particularly
vulnerable to invasive mammal species, particularly
rats. The eradication of invasive mammals is considered
the first step in island restoration and the subsequent
recovery of native species and biodiversity. Since these
early ground-based operations, rats (Rattus rattus, R.
norvegicus, R. exulans) have been successfully eradicated
from over 400 islands ranging in size from 1 to 12,850
ha, around the world, using the full gamut of methods and
technology (Moors & Atkinson, 1984; Atkinson, 1985;
Towns & Broome, 2003; Howald, et al., 2007; Jones, et al.,
2008; Parks & Wildlife Service, 2008; Parks & Wildlife
Service, 2014, DIISE, 2015). Of these rodent eradications,
the largest ground-based rat eradication operation, was on
Langara Island in British Columbia at 3,100 ha, and the
largest ground-based rat eradication in the United Kingdom
(UK) was on the Isle of Canna at 1,300 ha (Taylor, et al.,
2000; Bell, et al., 2011; DIISE, 2015).
Techniques and technology developed in those early
eradications have since moved on from ground-based
hand-broadcast and bait station operations to aeriallyapplied rodenticide operations and these have now been
used across the globe. Advances in, and alterations to,
techniques and tools have streamlined ground-based
operations. Lessons learnt from each eradication have
improved the next operation. However, in several
countries, including the United Kingdom (but excluding
the United Kingdom Overseas Territories), methods to
eradicate rats are restricted to ground-based methods.
The presence of critical non-target species, sensitive
habitats, island communities and legislative requirements
have restricted methods and tools for island eradications
in these countries. This paper describes the history and
development of ground-based rat eradications using bait
stations in the United Kingdom using five eradication
operations as examples and covers lessons learnt and how
local communities have been involved.
INVASIVE RATTUS SPECIES ON UK ISLANDS
Both black (Rattus rattus) and brown (R. norvegicus)
rats are present in the UK (Nowak, 1999; Long, 2003).
Black rats were presumed to have been introduced by the
Romans (c. 110 AD) and the brown rat via shipping between
1720 and 1728 (Thomas, 1985; Corbet & Southern, 1977;
Yaldwen, 1999; McCann, 2005; Parslow, 2007). Brown
rats were first recorded in the Isles of Scilly in 1728 after
several shipwrecks occurred that year (Thomas, 1985;
Parslow, 2007). Although the brown rat displaced the black
rat throughout most of the UK, black rats can still be found
in a small number of locations, particularly port cities such
as London, Edinburgh and Falmouth (Matheson, 1962;
Bentley, 1959; Twigg, 1992; Long, 2003). The brown rat
is still present on 56% of UK islands over 100 ha (Long,
2003).
Rats are known to have very detrimental effects on
seabird populations through predation and competition for
food and habitat, causing local and global extinction of birds
on islands throughout the world (Moors & Atkinson, 1984;
Atkinson, 1985; Courchamp, et al., 2003; Towns, et al.,
2006; Jones, et al., 2008; Bell, et al., 2016). The eradication
of introduced predators from islands has become one of
the most important tools in avian conservation in recent
times and, with an initial investment, significant long-term
restoration benefits such as increased productivity and
population increases of seabirds and other native species
as well as the establishment of new seabird species can be
achieved. The eradication of rats from seabird islands is
recognised as a prerequisite for the restoration of seabird
populations (Atkinson, 1985; Moors, et al.,1992).
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 79–87. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
79
Island invasives: scaling up to meet the challenge. Ch 1B Rodents: Review
Seabird populations on many UK islands have been
recorded in decline and in at least four cases rats have
been identified as one of the contributing factors for these
declines (Campbell, 1892; Brooke, 1990; Mitchell, et al.,
2004; Brooke, et al., 2007; Swann, et al., 2007; Dawson,
et al., 2015; Hayhow, et al., 2017). Many species such as
puffin (Fratercula arctica) which is listed as threatened
due to their declining population status (IUCN, 2017),
Manx shearwater (Puffinus puffinus) and the European
storm petrel (Hydrobates pelagicus) may have limited
distribution due to the impacts of, and predation by, rats
(Heaney, et al., 2002; Mavor, et al. 2008). Currently, the
majority of the UK puffin and all European storm petrel
populations nest on rat-free islands (Mavor, et al. 2008,
Ratcliffe, et al., 2009). The protection and enhancement
of UK seabird breeding habitat has been recognised
as an important conservation priority, including under
international conservation agreements (Brooke, et al.,
2007; Ratcliffe, et al., 2009; Dawson, et al., 2015; Thomas,
et al., 2017a).
confirmed until years after the operation. Unfortunately,
there have been recent reports of rats on Inchgarvie and
rats reinvaded Handa in 2012 (Thomas, et al., 2017a).
Rat eradications have occurred on over a dozen islands
around the UK with brown rats being the most common
target species (Bell, et al., 2011; Thomas, et al., 2017a; Bell,
et al., 2019a; Pearson, et al., 2019). Black rats have been
targeted on Lundy Island and the Shiant Isles (Lock, 2006;
Appleton, et al., 2006; Thomas, et al., 2017a; Main, et al.,
2019). Many of the eradications have occurred on islands
with permanent staff or the presence of small communities
(Bell, et al., 2011; Bell, et al., 2019a; Pearson, et al.,
2019). These operations demonstrate how ground-based
eradication techniques can be utilised on both inhabited
and uninhabited islands around the UK.
Five major eradications directed by WMIL have
occurred in the UK since 1999; Ramsey Island, Wales
(brown rat) in 1999/2000, Lundy Island, England (black
and brown rat) in 2002–2004, Isle of Canna, Scotland
(brown rat) in 2005/2006, St Agnes and Gugh, Isles of
Scilly, England (brown rat) in 2013/2014 and the Shiant
Isles, Scotland (black rat) in 2015/2016. In addition to
these five sites, eradication attempts have also been made
on Looe Island in 2006, the Calf of Man in 2012 and
Caldey Island in 2015, which have not been included here
because Looe Island was reinvaded by rats three years later
and the Calf of Man and Caldey Island eradications are still
on-going (Thomas, et al., 2017a).
Pre-1998: the early eradication operations
These five eradications used ground-based techniques
with bait stations placed out across the islands on either 25
m × 25 m, 25 m × 50 m, 50 m × 50 m, 90m × 90 m or 100
m × 100 m grids depending on the target species and type
of habitat or risk areas. The smaller grid sizes (between 25
and 50 metres spacing) were used to target black rats and
the larger grid sizes (between 50 and 100 metres spacing)
used to target brown rats, with the smallest spacings used
in high risk areas (such as around properties, seabird
colonies, wharves, farms and restaurants).
Despite an early attempt to eradicate rats from
Ailsa Craig in 1925, the first documented successful rat
eradication did not actually occur in the UK until 1968
on Cardigan Island in Wales (RSPB, 1924; RSPB, 1925a;
RSPB, 1925b; Johnstone, et al., 2005; Thomas, et al.,
2017a). This makes the UK the first country to intentionally
undertake a rat eradication operation anywhere in the world.
Four other rat eradications occurred between 1968 and
1998; Inchgarvie (Firth of Forth), Scotland in 1990, Ailsa
Craig, Scotland in 1991, Handa Island, Scotland in 1997
and Puffin Island, Wales in 1998 (Ratcliffe & Sandison,
2001; Zonfrillo, 2001; Zonfrillo, 2002; Johnstone, et al.,
2005; Stoneman & Zonfrillo, 2005; Thomas, et al., 2017a).
Warfarin was the primary active ingredient used in each
of these eradications with difenacoum used as a secondary
option in the Puffin Island rat eradication (Ratcliffe &
Sandison, 2001; Zonfrillo, 2001; Zonfrillo, 2002; Stoneman
& Zonfrillo, 2005). All of these early eradications used
ground-based methods, but focused on applying bait in
holes, burrows, under rocks and vegetation and in isolated
wooden bait stations or under inverted fish bins, rather than
in a systematic grid pattern (Ratcliffe & Sandison, 2001;
Zonfrillo, 2001; Zonfrillo, 2002; Stoneman & Zonfrillo,
2005; Thomas, et al., 2017a).
This method of baiting made it difficult to monitor bait
consumption by rats and non-target species. There were
no accurate records of bait take by rats or other species
from any of these operations (Ratcliffe & Sandison, 2001;
Zonfrillo, 2001; Zonfrillo, 2002; Stoneman & Zonfrillo,
2005; Thomas, et al., 2017a). Monitoring was limited: in
most cases it didn’t occur; used chewsticks across the island
immediately following the eradication (it has been noted
that chewsticks can be difficult to interpret sign accurately);
or was determined by the recovery of the seabird or rat
populations without any quantifiable measures (Zonfrillo,
2001; Ratcliffe, et al., 2009; Thomas, et al., 2017a). In the
case of Inchgarvie and Puffin Islands eradication was not
80
The later operations (post-1999)
The use of toxins and the risks these presented to nontarget species and the environment led to the development
of Best Practice and Standard Operating Procedures
for eradication operations in New Zealand in the 1990s
and these documents are revised as new techniques and
tools are developed (Cromarty, et al., 2002; Broome, et
al., 2011). Robust protocols for eradication operations
included detailed planning, operational requirements,
implementation protocols, monitoring guidelines and
biosecurity requirements (Cromarty, et al., 2002; Broome,
et al., 2011). These best practice and standard operating
techniques developed in New Zealand were followed and
adapted during the UK eradications undertaken by Wildlife
Management International Ltd (WMIL).
A simple yet effective bait station design has been used
in each of these five eradications in the UK. Although a
range of commercially available lockable stations have
been used in selected locations (e.g. residential homes,
farm buildings, schools, etc.) during these eradications, and
for on-going biosecurity to reduce the risk to the public,
particularly children and the possibility of tampering
with these long-term stations, the main bait stations
were made from corrugated drainage pipe. This design
is cost-effective and widely available. For the 1999/2000
Ramsey Island rat eradication, 500 mm lengths were used.
However, these stations were found to be too short as they
allowed carrion crows (Corvus corone) access to the bait.
The stations were made longer by adding 250 mm lengths
to one end. The standard length for each bait station in all
subsequent eradications was 750 mm long with an access
hole cut in the centre for placement of the bait (Fig. 1).
This access hole is covered with a short section of drainage
pipe. During the 2002–2004 Lundy Island rat eradication,
crows learnt to flick the lids off the stations to reach the
bait, therefore another length of wire was put around the
centre of the station to hold the lid tightly in place. This
“crow clip” became standard on all bait stations on any
island with either carrion crows, hooded crows (C. cornix)
or ravens (C. corax) present (Fig. 1).
Technological advances in GPS and GIS-linked
systems helped streamline the positioning of bait stations
during the grid establishment stage of eradications, as well
Bell: Ground-based rat eradication in the UK
Fig. 1 Example of the main bait station in position with the
crow clip holding the central lid in place, as used in the
five ground-based eradications in the United Kingdom
that were directed by Wildlife Management International
Ltd. [Credit: Elizabeth Bell, WMIL]
as monitoring the level of bait take by rats and non-target
species. Detailed maps can be produced for the eradication
team that can give additional information such as sensitive
sites like archaeological structures, location of rare plants,
seabird colonies and white-tailed sea eagle (Haliaeetus
albicilla) nesting sites and locations of access points for
steep terrain where more care needs to be taken (Fig. 2).
This intimate knowledge of the stations and island make
it easier for the eradication team to monitor bait take by
rats as the data can be linked to the specific station and
activity levels can be recorded on the spatial map. These
maps created by the GIS-linked database offer the team
the opportunity to monitor the decline in rat numbers
throughout the operation and could allow the eradication
personnel to react instantly to hotspots or problem areas
on the island. Specific bait take information can lead to
detailed activity which shows bait take by rats throughout
the operation and detailed heat maps showing complete
bait take by rats at the end of an eradication (Fig. 3).
Cereal-based bait blocks (containing the anticoagulant
diphacinone, bromadiolone or difenacoum) were used.
Each eradication used one bait formulation as the main bait
with a second option available to target ‘fussy’ (i.e. those
rats that will not eat the primary bait for whatever reason)
or surviving rats. This gives the option to adapt the project
if rat behaviour or taste preference becomes an issue during
the eradications. This has been shown to be important in
certain eradications as demonstrated by the Isle of Canna
brown rat eradication where the last surviving rat was
targeted successfully using the alternative bait. Bait was
placed loose in the bait stations for the first three weeks of
the eradication operation. This allowed rats to cache bait in
burrows for feeding themselves and any breeding females.
Once bait take has reduced, bait is wired securely into the
stations (Fig. 4) and any rat sign on these blocks is used to
identify the presence of a surviving rat or monitor high risk
areas, such as seabird colonies or farm buildings.
Fig. 2 An example of a detailed bait station map as used
by eradication teams during the Isle of Canna operation.
Where alphanumeric codes related to bait station positions
(e.g. WP = West Plateau, A = line A, 9 = bait station 9; Z
= Boundary line Z (two lines of stations at the top of the
cliff section above the coastal slopes), 19 = bait station
19; NN = Nunnery, B = line B, 6 = bait station 6), double
ended red arrows = safe access routes up or down to
the coastal slope areas, pink shaded areas = important
archaeological site (e.g. The Nunnery).
Fig. 3 Example of a heat map of bait take (g) by rats using
the results from the St Agnes & Gugh brown rat (Rattus
norvegicus) eradication.
Although aerial application operations generally
have a range of higher implementation costs compared
to ground-based operations due to the requirement of
a helicopter, sowing bucket and need for ground crew,
engineer and other legal requirements for use of aircraft,
the implementation time of the operation is often reduced
compared to a ground-based operation. Except for the
Lundy Island eradication (which took a second winter),
Fig. 4 Example of the main bait station in position showing
the open bait station (with the lid off) with the bait wired
in place, as used in the five ground-based eradications
in the United Kingdom that were directed by Wildlife
Management International Ltd. [Credit: Elizabeth Bell,
WMIL]
81
Island invasives: scaling up to meet the challenge. Ch 1B Rodents: Review
rats were successfully targeted within 21 to 64 days (Bell,
et al, 2011; Bell, et al., 2019a; Bell, et al., 2019b; Main, et
al., 2019).
The use of rodenticide baits to complete eradications has
enabled strong relationships to be established between bait
manufacturers, eradication operators and local agencies in
the UK. This has enabled open and in-depth discussions
about the bait in regard to problems with formulation,
taste and longevity that were identified during eradication
operations. Issues such as the bait blooming (i.e. swelling
and splitting after moisture on the bait) or rapidly going
mouldy in the Lundy eradication were relayed to the
manufacturers who altered the wax content for later
operations. This meant the bait became much more robust
in the damp winter conditions during the later Scottish
operations, reducing the overall bait quantity required for
those eradications.
European Union (EU) regulations require Bitrex™
(denatonium benzoate) or an alternative bittering agent to
be added to all rodenticides to deter human consumption.
Rats are not intended to be put off by Bitrex™, although
research suggests that some rats can detect it even at very
low concentrations, and preferentially choose bait that does
not contain Bitrex™ (Veitch, 2002). Three rats actively
avoided bait containing Bitrex™ on Lundy Island and,
by working with the UK bait manufacturer, dispensation
for a small amount of Bitrex™-free bait was obtained and
was used to successfully target the rats at those sites (Bell,
2004). Despite the bait manufacturer disputing the fact
that rats could detect and avoid Bitrex™ bait, they were
open to experiment, assess the issue and work together
with WMIL and Royal Society for the Protection of Birds
(RSPB) towards a solution. Without the engagement of the
bait manufacturer from the beginning of the project and
open and frank discussions about the possibility of this
issue with Bitrex™, the Lundy operation could have failed.
There have been recent regulatory changes to the
purchase and use of rodenticides in UK. The Health and
Safety Executive (HSE) require reassurance that biocide
products can be used without unacceptable risk to wildlife
and other non-target species and in July 2015 implemented
the UK Rodenticide Stewardship Scheme. This scheme
covers all rodenticide products sold to, and used by,
professionals when applied outside buildings and in open
areas and operates under a Code of Best Practice developed
by the Campaign for Responsible Rodenticide Use (CRRU)
group (CRRU, 2015). All professionals must have proof of
competence at the point-of-sale for rodenticide baits (i.e.
have completed certification for rodenticide control and/
or eradication by completing an approved training course)
as well as comply with the best practice. These regulations
generally relate to urban control operations, pest control
operators and farmers, but eradication programmes must
also follow these regulations. RSPB, in conjunction with
CRRU, have developed an eradication-specific registered
training course under the UK Stewardship Scheme.
Ground-based operations facilitate longer, wide-scale
monitoring compared to aerial operations; not only using
the bait itself, but also using a range of monitoring tools
such as flavoured wax blocks, soap, tracking tunnels and
trail cameras. Monitoring can be established at the same
locations as the bait stations, as well as between the bait
stations to intensify the scope of monitoring and ensure every
micro-habitat is covered. Having non-toxic monitoring
devices out in the open (pegged to the ground) and using
a range of options gives more chance for any surviving
rats to interact with at least one type of monitoring tool.
This can also identify if a percentage of the rat population
or rats at a specific location are avoiding the bait stations
for any reason. This intensive effort can be used to detect
82
any survivors and the operator can adapt the eradication to
successfully target those last individuals. WMIL developed
a range of flavoured wax blocks that have proved to be
very effective in detecting the presence of surviving rats at
the final stages of eradication (Fig. 5). These blocks have
been freshly produced by the eradication team on-site to a
standard recipe as the operation progresses. This flavoured
wax recipe has been widely shared amongst the eradication
industry.
This period of intensive island-wide monitoring allows
the eradication operators to be much more confident that
the eradication has been successful prior to leaving the
island. By being able to detect and respond to surviving
rats immediately, this reduces the likelihood of eradication
failure (as any rat that is detected during this period can be
targeted) and thus the need for a second eradication attempt
(which can cost as much as the original operation). This
intensive monitoring period in these five UK operations
occurred for up to four months, depending on the size of
the island and time required to initially target the rats during
the baiting phase. Additional monitoring is completed at
least quarterly for two years prior to the intensive final
check phase and rat-free declaration following standard
international eradication protocols for temperate operations
(Broome, et al., 2011).
The use of volunteers has been an asset to these five
UK eradications by giving passionate conservationists the
chance to be involved in a project they feel strongly about,
increasing the national (e.g. RSPB) capacity in eradication
methodology, and engagement with the local communities.
However, the use of volunteers can reduce the awareness
of managers, decision makers and funders of the true cost
and effort required to complete ground-based eradications.
The costs of these five ground-based eradications
ranged from £76,000 up to £900,000, including planning,
implementation, key species pre-and post-eradication
monitoring, monitoring for survivors or incursions for two
Fig. 5 Examples of flavoured wax as used for monitoring
in the five ground-based eradications in the United
Kingdom that were directed by Wildlife Management
International Ltd. [Credit: Jaclyn Pearson, RSPB]. Where
the left (blue) block is aniseed flavour, centre (brown)
block is chocolate flavour and right (fawn) block is peanut
flavoured and each block is pegged to the ground with
a piece of fencing wire and marked with a short piece of
flagging tape for visibility.
Bell: Ground-based rat eradication in the UK
years post-eradication and confirmation monitoring (‘final
check’) prior to the declaration of rat-free status (Dr R.
Luxmoore, NTS, pers. comm.; P. St Pierre, RSPB, pers.
comm.; Lock, 2006).
There can be difficulty associated with accurately
recording the entire costs of eradications; in many cases
reported costs do not include in-kind or match funded
expenses by the agencies involved (National Trust for
Scotland, RSPB, etc.). In many cases, it can be difficult
to accurately record these costs against the eradication
operation as they relate to administration and corporate
expenses.
BEST PRACTICE FOR ERADICATIONS
It has long been recognised that every island is different
when it comes to planning and implementing an eradication
operation. As such, although the NZ best practice gave an
important starting point for the UK operations, it needed to
be adapted for the local situation to become more relevant
and effective, particularly in regard to local legislation and
animal welfare regulations.
The RSPB, in partnership with UK-based governmental
and non-governmental organisations working in island
restoration, with input from international experts in this
field produced The UK Rodent Eradication Best Practice
Toolkit which is hosted on the Great Britain Non-Native
Species Secretariat website (Thomas, et al., 2017b).
This toolkit was developed as an advisory resource to
provide systematic planning and implementation protocols
for ground-based rodent eradications and biosecurity
in the UK (Thomas, et al., 2017b). It aims to give UK
organisations technical advice on eradication methodology
as well as an eradication project management framework
to enable greater confidence in achieving island restoration
goals in invasive rodent management projects in the UK
(Thomas, et al., 2017b).
THE ROLE OF COMMUNITIES
The majority of eradications around the world have
occurred on uninhabited islands and it is thought that
islands with significant human populations, unreceptive
communities or occurrence of livestock and domestic
animals are unlikely to be feasible for rat eradication
(Campbell, et al., 2015). However, because invasive
species are also a problem on inhabited islands, such
eradications must be considered. A lack of public awareness
about invasive species impacts and misunderstanding
of eradication techniques from island communities are
thought to have been responsible for the opposition of
proposed eradications on inhabited islands around the world
(Bryce, et al., 2011). The importance of the engagement
and inclusion of local communities has been highlighted
in a number of recent eradication and research projects,
especially in regard to risk and benefit analysis and to ensure
a suitable environment for eradication projects to proceed
can occur (Bryce, et al., 2011; Eason, et al., 2008). Respect
for the attitudes, and safety, of local communities needs
to be a priority in any eradication planned for inhabited
islands. The support and agreement by the community
to proceed with an eradication is vital for any project
on an inhabited island. This is particularly important as
access into all properties is vital to effectively carry out
an eradication. Involving the residents in the concept,
planning, implementation and on-going biosecurity of the
island was recognised as the only way such an eradication
could have occurred on the islands in the UK.
Considerations to how the community view the
environment, how they think the proposed eradication will
affect them and other social science considerations need to
be assessed for eradications planned for inhabited islands.
Most importantly, all aspects of the eradication should be
discussed with the community in the early stages of the
proposal. Unlike eradication operators, most members of
the public do not have any knowledge of the principles
and techniques of eradication, particularly in regard
to rodenticide choice and operational procedures. It is
important that each community member understands these
aspects and how they will personally be affected by the
day-to-day operational requirements.
As there were staff or small communities present at
four of the five previously mentioned UK eradications,
almost all recent operations undertaken in the UK have
had to work closely within these communities and have
had to adapt to the issues and technical challenges the
presence of people has on the eradications. During each
of these eradications, WMIL and the local project partner
worked closely with the landowner, staff and residents
to understand and address concerns and questions about
the operations. Where the operation occurred on staffed
islands, the decision to complete an eradication had
already been made by the main project partner concerned
and much of the consultation with staff on the islands had
already been completed by the management prior to the
operation. Resident staff were generally supportive of the
eradication and often viewed the eradication operational
team as temporary, but separate, staff members. In
comparison to those islands with resident staff, WMIL and
RSPB recognised the importance of the engagement of
the 85-person resident community on St Agnes and Gugh
in the Isles of Scilly and started this engagement process
early for the eradication of brown rats (Bell, et al., this
issue a, Pearson, et al., 2019). The success of the St Agnes
and Gugh eradication (Isles of Scilly Seabird Recovery
Project, IOSSRP) showed how the community-based
approach that was designed to develop local networks and
use existing community structures to build support for the
project worked extremely well. The vision and benefits of
the project were shared by the community and the residents
were part of the decision-making process and management
of the project.
An open and transparent operating system has worked
well in all these five previously mentioned eradications in
the UK. Information covering details on rodenticide type,
bait station design, anticoagulant poisoning symptoms
and treatment, contact numbers and project management
was provided to all residents, stakeholders and interested
parties. The project team was permanently present on each
of the islands throughout the eradication to implement
the operation, answer any questions and deal with any
issues. Project updates were provided to the community
and stakeholders each week, which gave the residents the
opportunity to observe the operational procedures and
results as the eradication proceeded. Real-time bait-take
maps were provided as part of this process. A 24-hour
contact telephone number was provided for immediate
response to any issues that a resident may have.
BIOSECURITY
With the eradication of rats from islands, the priority is to
ensure that they do not become re-established. Biosecurity
is a critical aspect of any eradication and should be
designed, implemented and tested prior to the completion
of the eradication and departure of the eradication team.
Prevention of an accidental rat reintroduction should be
the primary aim. Precautions need to be taken not only in
obvious situations such as with visitors or boat movements,
or when high-risk items like stock feed or hay are being
delivered to the island, but also when the risk may be
mistakenly thought to be negligible.
83
Island invasives: scaling up to meet the challenge. Ch 1B Rodents: Review
The long-term legacy of these five UK eradication
projects was important to the implementing agencies
involved as well as the communities and agency staff on
the island. As such, practical biosecurity strategies were
established for the community and supporting agencies;
measures that have been designed to reduce the risk of
rats being reintroduced to a minimum, without being
a hindrance to the daily lives of the staff, community or
visitors to the island. A range of biosecurity strategies were
proposed to the residents or agency staff on each island
and, following discussions about the protocols of each
strategy, suitable measures for each island were selected
and implemented. Public awareness and education leaflets
have been developed for every eradication to ensure that
the public are aware of the rat-free status of each island
and ways they can assist in keeping the islands rat-free.
Residents and staff members from the project partners have
been trained in all relevant biosecurity measures and they
will maintain regular monitoring checks on the islands
in perpetuity. Funding for on-going biosecurity has been
provided by partner agencies and completed by staff or in
the case of St Agnes and Gugh, funds will be provided by
the community through fundraising and grants (Pearson,
et al., 2019). In some instances, such as on St Agnes and
Gugh, community coordinators will maintain liaison
between the residents and the supporting partner agencies
(Pearson, et al., 2019).
DISCUSSION
Rat eradications have been undertaken on islands around
the world for the past 65 years and in the UK for the past 50
years. International rat eradication projects over this time
have used a range of methods but most recently focused
on the aerial application of rodenticides. However, due to
legislative limitations upon the outdoor-use of rodenticides
and application methods, and although derogations can be
issued to allow aerial operations, ground-based methods are
likely to remain the predominant rat eradication technique
in the UK (and other European counties). Developments
from five eradications in the UK have streamlined operating
procedures and eradication techniques for the next
eradication. Using plastic corrugated drainage pipe as the
main bait station type has enabled the design to be adapted
to exclude large or problematic non-target species such as
rabbits and crows. The positioning of bait stations using
GPS and GIS-linked systems has streamlined recording
bait take by rats and non-target species and enabled this to
be monitored in real time. Constant monitoring throughout
the operations starting with the bait take and progressing
through to using a range of monitoring devices, such as
flavoured wax, allowed for each operation to adapt to
deal with high risk areas or ‘fussy’ rats to maximise the
likelihood of eradication success. This intensive level of
monitoring allows any issues that may arise with bait to
be addressed directly with the manufacturers and rectified
early in the operational timeframe.
Ground-based eradications have been completed on
islands ranging in size from <1 ha to 3,100 ha (Taylor,
et al., 2000; DIISE, 2015). Although an island’s size and
terrain may prevent a ground-based bait station operation
being completed, it would be perfectly feasible to eradicate
rats from even larger inhabited and uninhabited islands
assuming there were enough resources (including staff
and funding) and commitment and support from all
involved. The feasibility assessment for any proposed
eradication needs to investigate the costs and benefits of all
possible methods before deciding on the final operational
techniques. In many cases, a combination of aerial and
ground-based operations may also be suitable or preferred
by communities on large inhabited islands, as shown by
84
recent eradication plans such as for Lord Howe Island
(Wilkinson & Priddel, 2011; Walsh, 2019).
Over 85% of rat eradications around the world have
been completed on uninhabited islands (n = 721 out of
820 eradications; DIISE, 2015). However, many are now
either being investigated or planned for islands with
resident communities (Oppel, et al., 2011; Russell &
Broome, 2016; Stanbury, et al., 2017). Eradications on
inhabited islands raise social, economic, conservation
and technical challenges for the operation (Moon, et al.,
2015). The experience in the UK shows that to ensure an
island restoration project runs successfully the support
and agreement from the community must be secured. The
community must share the project’s vision and feel that
they are one of the beneficiaries. To do this, they will need
to be included and play an integral role in the decisionmaking process, planning preparation and implementation
and management of the project. In this way, the legacy of
the project will be much stronger. Those proposing the
eradication need to ensure that the community is aware of
the effects of invasive rats on the native biodiversity of their
island and how the proposed eradication can benefit those
species as well as explaining the process of the eradication
operation itself. However, project partners and eradication
operators also need to realise that for a number of residents
the biodiversity and environmental reasons to eradicate rats
may be of no interest; as such, social and economic benefits
should also be outlined during the planning stages as these
may be more important to the communities themselves. It
is important for operators to realise that communities may
not have the same understanding of eradication processes
and each aspect of the project may have to be explained.
The larger the community the longer, potentially, the
project managers will need to ensure that the residents
are all at the same position of understanding through the
various stages of the project. Archipelagos or groups of
islands bring additional stakeholders and interested parties
that need to be engaged compared to single islands. From
my experience, ten years is not an unreasonable timescale
depending upon the starting point, the value placed upon
seabirds by the community and the strength of the project
partnership. In my view, and in agreement with others
such as Moon et al. (2015), the ongoing consultation
and communication with the local community and wider
stakeholder groups during any eradication is essential.
As the need to prioritise islands for restoration
has increased, the requirement of understanding and
quantifying the costs of eradications has also increased
(Martins, et al., 2006; Holmes, et al., 2015). Although
general costs for eradications can be estimated if the size of
the island and target species are known, and it appears that
costs increase with the size of island, there are other costs
from application method, permits, non-target mitigation,
and biodiversity monitoring that need to be factored into
an eradication operation (Martin, et al., 2006; Holmes, et
al., 2015). This information is vital to be able to accurately
determine the complete costs for future eradications and it
is important that project costs are reported.
The defining factors underpinning the success of the
eradication operations on inhabited UK islands were the
professional management of the eradication, dedicated and
passionate volunteer involvement, efficient and systematic
monitoring, adapting to local conditions and ensuring a
community-inclusive approach.
This model of consultation, engagement and
community-involvement developed on these inhabited
islands eradications in the UK can offer valuable
information, advice and direction for eradication operations
planned on islands with larger communities in the UK and
Bell: Ground-based rat eradication in the UK
around the world. The eradication of brown rats from St
Agnes and Gugh could be used as a valuable education
tool to show other communities that it is possible to safely
eradicate rats and implement suitable biosecurity measures
to reduce the risk of reinvasion without impacting on the
lives of the residents, as reported by Pearson, et al. (2019).
This model, and future techniques developed during
other eradications on inhabited islands, will be even more
important if restrictions on application measures and
outdoor-use of rodenticides expand to countries outside of
the UK. It is important for eradication operators to realise
that even if aerial application methods are possible at the
location, the community on the island may not approve or
permit that type of method. As such, the use of groundbased bait station techniques will have a vital part to play
and this option should be assessed as part of any original
feasibility assessment.
Island restoration on UK islands has led to the dramatic
recovery of seabird populations. Manx shearwaters on
Ramsey and Lundy Islands have increased nearly tenfold in the ten to fifteen years since the eradication of
brown and black rats and the recolonization of European
storm petrels and other small burrowing species has been
recorded after long absences (Brown, et al., 2011; Morgan,
2012; Booker & Price 2014; Bell, et al., 2019b). These
types of results have helped develop a legacy for many of
the projects, with the residents and agency personnel on
the islands committing to and doing their part to maintain
important biosecurity measures. These results can also be
used to help explain the benefits of completing this type of
eradication project on other islands, even those with larger
communities or a complex of target species. Providing safe
breeding habitat and creating and then maintaining rodentfree status at important island sites, will be an important
part of the long-term legacy of protection for UK seabirds.
It is important that when eradication projects are being
designed and assessed that operators and project partners
factor in on-going biosecurity after the completion of the
project, particularly in relation to equipment, capacity
and long-term funding requirements. It is one of the most
vital aspects of an eradication project and agencies must
recognise the requirement that biosecurity is required in
perpetuity. For eradications that occur on inhabited islands,
this makes the engagement of, and commitment from, the
communities to undertake biosecurity measures, even more
important to ensure the legacy of any eradication project.
Detailed prioritisation exercises such as Brooke,
et al. (2007), Ratcliffe, et al. (2009) and Stanbury, et al.
(2017) have identified a number of UK and UK Overseas
Territories’ islands as being pre-eminent sites for rat
eradication because of their importance to seabirds. Twenty
of the 25 islands identified in the most recent prioritisation
exercise have resident human populations which increases
the challenges for any eradication proposed for those sites
(Stanbury, et al., 2017). One of the most important lessons
identified by completing eradication operations on inhabited
islands is that the community needs to be engaged as early
as possible, preferably in the concept and development
process. As important, all stages of the eradication need
to be completely open and transparent, with community
members involved throughout the implementation of the
project and into the future to ensure the sustainability of the
on-going biosecurity for the island. The newly developed
Best Practice for UK islands (Thomas, et al., 2017b) which
has built on all the lessons learnt from these eradications
that have occurred over the past 50 years in the UK should
help make these future eradication operations more likely
to succeed on both uninhabited and inhabited islands.
ACKNOWLEDGEMENTS
Special thanks to all the volunteers that have worked on
all these UK eradications: the success and recovery of these
islands is a credit to their hard work and enthusiasm. These
eradications were funded and supported by RSPB, The
Landmark Trust, The Lundy Company, The National Trust,
Natural England, Scottish Natural Heritage, The National
Trust for Scotland, EU LIFE, Defra, Heritage Lotteries
Fund, WMIL and private donors. The communities and
residents of St Agnes & Gugh, Isle of Canna, Lundy Island
and Ramsey Island for their hospitality and support while
the eradication teams were on their islands and for their
on-going commitment to island biosecurity. Tom Nicolson
and Adam Nicolson for their support of the Shiant Isles
eradication and their commitment to on-going biosecurity.
The Steering Groups for each eradication for their support
and advice throughout these projects. Thanks to Paul St
Pierre (RSPB) and Dr Richard Luxmoore (National Trust
for Scotland) for providing the costs for the St Agnes
and Gugh and Isle of Canna eradications. Thanks to Bell
Laboratories, PelGar International Ltd and Syngenta Ltd
for the supply of the rodenticide bait. Thanks also to Dr
Alan Buckle, Dr Colin Prescott and David Rymer, Reading
University, for their assistance with the resistance testing
for many of these eradications. Kelvin Floyd produced
Figs 2 and 3. This paper is dedicated to the memory of
my father, Brian D. Bell (1930–2016), whose vision for
rat-free islands to protect and enhance seabird populations,
and his training in eradication techniques, has taken me
around the world.
REFERENCES
Appleton, D., Booker, H., Bullock, D., Cordrey, L. and Sampson, B.
(2006). ‘The seabird recovery project: Lundy Island’. Atlantic Seabirds
8: 51–59.
Atkinson, I.A.E. (1985). ‘The Spread of Commensal Species of Rattus
to Oceanic Islands and their Effects on Island Avifaunas’. In: P.J.
Moors (ed.) Conservation of Island Birds, pp. 35–81. ICBP Technical
Publication 3. Cambridge, UK: International Council for Bird
Preservation (ICBP).
Bell, E. (2004). Lundy Seabird Recovery Project 2002–2004: The
Eradication of Black and Brown Rats from Lundy Island. New Zealand:
Wildlife Management International Ltd.
Bell, E., Boyle, D., Floyd, K., Garner-Richards, P., Swann, B.,
Luxmoore, R., Patterson, A. and Thomas, R. (2011). ‘The Groundbased Eradication of Norway Rats (Rattus norvegicus) from the Isle
of Canna, Inner Hebrides, Scotland’. In: C.R. Veitch, M.N. Clout and
D.R. Towns (eds.) Island invasives: eradication and management, pp.
269–274. Occassional Paper SSC no. 42. Gland, Switzerland: IUCN
and Auckland, New Zealand: CBB.
Bell, E., Bell, B. and Merton, D. (2016). ‘The legacy of Big South Cape:
rat irruption to rat eradication’. New Zealand Journal of Ecology 40:
212–218.
Bell, E., Floyd, K., Boyle, D., Pearson, J., St Pierre, P., Lock, L., Mason,
S., McCarthy, R. and Garratt, W. (2019a). ‘The Isles of Scilly seabird
restoration project: the eradication of brown rats (Rattus norvegicus)
on the inhabited islands of St Agnes and Gugh, Isles of Scilly’. In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West
(eds.) Island invasives: scaling up to meet the challenge, pp.. 88–94.
Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Bell, E., Bell, M., Morgan, G. and Morgan, L. (2019b). ‘The recovery
of seabird populations on Ramsey Island, Pembrokeshire, Wales,
following the 1999/2000 rat eradication’. In: C.R. Veitch, M.N. Clout,
A.R Martin, J.C Russell and C.J. West (eds.) Island invasives: scaling
up to meet the challenge, pp. 539–544. Occasional Paper SSC no. 62.
Gland, Switzerland: IUCN.
Bentley, E. (1959). ‘The distribution and status of Rattus rattus L. in the
United Kingdom in 1951 and 1956’. Journal of Animal Ecology 28:
299–308.
Booker, H. and Price, D. (2014). ‘Manx shearwater recovery on Lundy:
population and distribution change from 2001 to 2013’. Journal of the
Lundy Field Society 4: 105–116.
85
Island invasives: scaling up to meet the challenge. Ch 1B Rodents: Review
Brooke, M. (1990). The Manx Shearwater. Calton, UK: T. & A.D. Poyser.
Brooke, M., Hilton, G. and Martins, T. (2007). ‘Prioritising the world’s
islands for vertebrate-eradication programmes’. Animal Conservation
10: 380–390.
Broome, K., Brown, D, Cox, A., Cromarty, P., McClelland, P., Golding,
C., Griffiths, R. and Bell, P. (2011). Current Agreed Best Practice for
Rat Eradication - poison bait in bait stations (Version 1.3). New Zealand
Department of Conservation internal document DOCDM-839096.
Wellington, New Zealand: Department of Conservation.
Brown, A., Price, D., Slader, P., Booker H., Lock, L. and Deveney, D.
(2011). ‘Seabirds on Lundy: their current status, recent history and
prospects for the restoration of a once-important bird area’. British
Birds 104: 139–158.
Bryce, R., Oliver, M., Davies, L., Gray, H., Urquhart, J. and Lambin,
X. (2011). ‘Turning back the tide of American mink invasion at an
unprecedented scale through community participation and adaptive
management’. Biological Conservation 144: 575–583.
Campbell, J. (1892). ‘On the appearance of the brown rat (Mus
decumanus, Pallas) on Ailsa Craig’. Annals of Scottish Natural History
1892: 132–134.
Campbell, K., Beek, J., Eason, C., Glen, A., Godwin, G., Gould, F.,
Holmes, N., Howald, G., Madden, F., Ponder, J., Threadgill, D.,
Wegmann, A. and Baxter, G. (2015). ‘The next generation of rodent
eradications: Innovative technologies and tools to improve species
specificity and increase their feasibility on islands’. Biological
Conservation 185: 47–58.
Corbet, G. and Southern, H. (1977). ‘The Handbook of British Mammals’.
Oxford: Blackwell Scientific Publications.
Courchamp, F., Chapius, J-L. and Pascal, M. (2003). ‘Mammal invaders
on islands: impacts, control and control impact’. Biological Review 78:
347–383.
Cromarty, P., Broome, K., Cox, A., Empson, R., Hutchinson, W.
and McFadden, I. (2002). ‘Eradication planning for invasive alien
animal species on islands – the approach developed by New Zealand
Department of Conservation’. In: C.R. Veitch and M.N. Clout (eds.).
Turning the tide: the eradication of invasive species, pp. 85–91. IUCN,
Gland, Switzerland and Cambridge, UK.
CRRU. (2015). CRRU UK Code of Best Practice: Best Practice and
Guidance for Rodent Control and the Safe Use of Rodenticides.
Yorkshire, UK: CRRU UK.
Dawson, J. Oppel, S., Cuthbert, R., Homes, N., Bird, J., Butchart, S.,
Spatz, D. and Tershey, B. (2015). ‘Prioritizing islands for the eradication
of invasive vertebrates in the United Kingdom overseas territories’.
Conservation Biology 29: 143–153.
Jones, H., Tershy, B., Zavaleta, E., Croll, D., Keitt, B., Finkelstein, M.
and Howald, G. (2008). ‘Review of the global severity of the effects of
invasive rats on seabirds’. Conservation Biology 22: 16–26.
Lock, L. (2006). ‘Eradication of brown rats Rattus norvegicus and black
rats Rattus rattus to restore breeding seabird populations on Lundy
Island, Devon, England’. Conservation Evidence 3: 111–113.
Long, J. (2003). Introduced Mammals of the World: Their History,
Distribution and Influence, Wallingford, UK: CABI.
Lorvelec, O. and Pascal, M. (2005). ‘French attempts to eradicate
non-indigenous mammals and their consequences for native biota’.
Biological Invasions 7: 135–140.
Main, C., Bell, E., Floyd, K., Tayton, J., Ibbotson, J., Whittington, W.,
Taylor, P., Reid, R., Varnham, K., Churchyard, T., Bambini, L., Douse,
A., Nicolson, T. and Campbell, G. (2019). ‘Scaling down (cliffs) to meet
the challenge: the Shiants black rat eradication’. In: C.R. Veitch, M.N.
Clout, A.R Martin, J.C. Russell and C.J. West (eds.) Island invasives:
scaling up to meet the challenge, pp. 138–146. Gland, Switzerland:
IUCN.
Martins, T., Brooke, M., Hilton, G., Farnsworth, S., Gould, J. and Pain, D.
(2006). ‘Costing eradications of alien mammals from islands’. Animal
Conservation 9: 439–444.
Matheson, C. (1962). Brown Rat. London: Sunday Times Publications.
(Animals of Britain, No. 16).
Mavor, R., Heubeck, M., Schmitt, S. and Parsons, M. (2008). Seabird
Numbers and Breeding Success in Britain and Ireland, 2006.
Peterborough: Joint Nature Conservation Committee. (UK Nature
Conservation, No. 31).
McCann, J. (2005). ‘The introduction of the brown rat (Rattus
norvegicus)’. Somerset Archaeology and Natural History Journal 2005:
139–141.
Mitchell, P., Newton, S., Ratcliffe, N. and Dunn, T. (2004). Seabird
Populations of Britain and Ireland: Results of the Seabird 2000 Census
(1998–2002). London: T. & A.D. Poyser.
Moon, K., Blackman, D. and Brewer, T. (2015). ‘Understanding and
integrating knowledge to improve invasive species management’.
Biological Invasions 17(9): 2675–2689.
Moors, P.J. and Atkinson, I.A.E (1984). ‘Predation on Seabirds by
Introduced Animals and Factors Affecting its Severity’. In: J.P. Croxall,
P.G.H. Evans and R.W. Schreiber (eds.) Status and Conservation
of the World’s Seabirds, pp. 667–690. ICBP Technical Publication 2.
Cambridge, UK: International Council for Bird Preservation (ICBP).
Moors, P.J., Atkinson, I.A.E. and Sherley, G.H. (1992). ‘Reducing the rat
threat to island birds’. Bird Conservation International 2: 93–114.
DIISE. (2015). ‘The Database of Island Invasive Species Eradications,
developed by Island Conservation, Coastal Conservation Action
Laboratory UCSC, IUCN SSC Invasive Species Specialist Group,
University of Auckland and Landcare Research New Zealand’. <http://
diise.islandconservation.org>.
Morgan, G. (2012). ‘The bird populations of Ramsey and Grassholm’.
British Birds 105: 716–732.
Eason, C., Ogilvie, S., Miller, A., Henderson, R., Shapiro, L., Hix, S.,
MacMorran, D. and Murphy, E. (2008). ‘Smarter Pest Control Tools
with Low-residue and Humane Toxins’. In: R.M. Timm and J.M.
O’Brien (eds.) Proceedings of the 23rd Vertebrate Pest Conference, pp.
148–153. Davis, California, USA: University of California.
Parks and Wildlife Service. (2008). Macquarie Island Pest Eradication
Plan. Part A: Overview. Hobart, Tasmania: Department of the
Environment, Parks, Heritage and the Arts. <http://www.parks.tas.gov.
au/index.aspx?base=6186>.
Hayhow, D., Bond, A., Douse, A., Eaton, M., Frost, T., Grice, P., Hall,
C., Harris, S., Havery, S., Hearn, R., Noble, D., Oppel, S., Williams, J.,
Win, I. and Wotton, S. (2017). The State of the UK’s Birds 2016. Sandy,
Bedfordshire: The RSPB, BTO, WWT, DAERA, JNCC, NE, NRW and
SNH.
Heaney, V., Ratcliffe, N., Brown, A., Robinson, P. and Lock, L. (2002).
‘The status and distribution of storm-petrels Hydrobates pelagicus and
Manx shearwaters Puffinus puffinus on the Isles of Scilly’. Atlantic
Seabirds 4: 1–16.
Holmes, N., Campbell, K., Keitt, B., Griffiths, R., Beek, J., Donlan,
C. and Broome, K. (2015). ‘Reporting costs for invasive vertebrate
eradications’. Biological Invasions 17: 2913–2925.
Howald, G., Donlan, C., Galvan, J., Russell, J., Parkes, J., Samaniego, A.,
Wang, Y., Veitch, D., Genovesi, P., Pascal, M., Saunders, A. and Tershy,
B. (2007). ‘Invasive rodent eradication on islands’. Conservation
Biology 21: 1258–1268.
IUCN. (2017). ‘The IUCN Red List of Threatened Species. Version 20171’. <http://www.iucnredlist.org>. Accessed 12 May 2017.
Johnstone, I., Gray, C. and Noble, D. (2005). The State of Birds in Wales
2004. Cardiff, UK: RSPB Cymru.
86
Oppel, S., Beaven, B., Bolton, M., Vickery, J. and Bodey, T. (2011).
‘Eradication of invasive mammals on islands inhabited by humans and
domestic animals’. Conservation Biology 25: 232–240.
Parks and Wildlife Service. (2014). Evaluation Report: Macquarie Island
Pest Eradication Project, August 2014. Hobart, Tasmania: Department
of Primary Industries, Parks, Water and Environment. <http://www.
parks.tas.gov.au/file.aspx?id=31160>.
Pearson, J., St Pierre, P., Lock, L., Buckley, P., Bell, E., Mason, S.,
McCarthy, R. Garratt, W., Sugar, K. and Pearce, J. (2019). ‘Working with
the local community to eradicate rats on an inhabited island: securing
the seabird heritage of the Isles of Scilly’. In: C.R. Veitch, M.N. Clout,
A.R. Martin, J.C. Russell and C.J. West (eds.) Island invasives: scaling
up to meet the challenge, pp. 670–678. Occasional Paper SSC no. 62.
Gland, Switzerland: IUCN.
Ratcliffe, J. and Sandison W. (2001). ‘Puffin Island - will removing rats
bring back the puffins?’. Natur Cymru 1: 23–26.
Ratcliffe, N., Mitchell, I., Varnham, K., Verboven, N. and Higson, P.
(2009). ‘How to prioritise islands for the benefit of petrels: a case study
of the UK, Channel Islands and Isle of Man’. Ibis 151: 699–708.
RSPB. (1924). ‘Ailsa Craig’. Bird Notes and News.11: 27–28.
RSPB. (1925a). ‘Ailsa Craig’. Bird Notes and News 11: 99–100.
RSPB. (1925b). ‘Ailsa Craig’. Bird Notes and News 11: 150.
Bell: Ground-based rat eradication in the UK
Russell, J. and Broome, K. (2016). ‘Fifty years of rodent eradications in
New Zealand: another decade of advances’. New Zealand Journal of
Ecology 40: 197–204.
Stanbury, A., Thomas, S., Aegerter, J., Brown, A., Bullock, A., Eaton, M.,
Lock, L., Luxmoore, R., Roy, S., Whitaker, S. and Oppel, S. (2017).
‘Prioritising islands in the United Kingdom and crown dependencies
for the eradication of invasive alien vertebrates and rodent biosecurity’.
European Journal of Wildlife Research 63: 31. <https://doi.org/10.1007/
s10344-017-1084-7>.
Stoneman, J. and Zonfrillo, B. (2005). ‘The eradication of brown rats
from Handa Island, Sutherland’. Scottish Birds 25: 17–23.
Swann, R., Alton, D., Foster, S., Graham, K., Graham, A., Ramsay, A.
and Young, A. (2007). ‘The impact of brown rats on Canna’s breeding
seabirds’. Proceedings of the Conference: Tackling the Problem of Alien
Mammals on Seabird Colonies. Edinburgh, Scotland: National Trust for
Scotland.
Taylor, R., Kaiser, G. and Drever, M. (2000). ‘Eradication of Norway rats
for recovery of seabird habitat on Langara Island, British Columbia’.
Restoration Ecology 8: 151–160.
Thomas, C. (1985). Exploration of a Drowned Landscape: Archaeology
and History of the Isles of Scilly. London: Batsford Ltd.
Thomas, B. and Taylor, R. (2002). ‘A history of ground-based rodent
eradication techniques developed in New Zealand, 1959–1993’. In:
C.R. Veitch and M.N. Clout (eds.) Turning the tide: the eradication of
invasive species, pp. 301–310. Occasional Paper SSC no. 28. IUCN
SSC Invasive Species Specialist Group. IUCN, Gland, Switzerland and
Cambridge, UK.
Thomas, S., Brown, A., Bullock, D., Lock, L., Luxmoore, R., Roy, S.,
Stanbury, A. and Varnham, K. (2017a). ‘Island restoration in the UK –
past present and future’. British Wildlife 28: 231–242.
Thomas, S., Varnham, K., and Havery, S. (2017b). Current Recommended
Procedures for UK (Bait Station) Rodent Eradication Projects (Version
4.0). Sandy, Bedfordshire: Royal Society for the Protection of Birds.
Towns, D. and Broome, K. (2003). ‘From small Maria to massive
Campbell: Forty years of rat eradications from New Zealand islands’.
New Zealand Journal of Zoology 30: 377–398.
Towns, D., Atkinson, I. and Daugherty, C. (2006). ‘Have the harmful
effects of introduced rats on islands been exaggerated?’. Biological
Invasions 8: 863–891.
Twigg, G. (1992). ‘The black rat Rattus in the United Kingdom in 1989’.
Mammal Review 22: 33–42.
Walsh, A. (2019). ‘Winning the hearts and minds – proceeding to
implementation of the Lord Howe Island rodent eradication project: a
case study’. In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and
C.J. West (eds.) Island invasives: scaling up to meet the challenge, pp.
522–530. Occasional Paper SSC no. 62. IUCN, Gland, Switzerland.
Wilkinson, I. and Priddel, D. (2011). ‘Rodent Eradication on Lord Howe
Island: challenges posed by people, livestock, and threatened endemics’.
In: C.R. Veitch, M.N. Clout and D.R. Towns (eds.) Island invasives:
eradication and management, pp. 508–514. Occassional Paper SSC no.
42. Gland, Switzerland: IUCN and Auckland, New Zealand: CBB.
Zonfrillo, B. (2001). Wildlife Conservation on Ailsa Craig. The
Thomas Duncan Memorial Lectures. Girvan, Scotland: Friends of the
McKechnie Institute.
Zonfrillo, B. (2002). ‘Puffins return to Ailsa Craig’. Scottish Bird News
66: 1–2.
87
E. Bell, K. Floyd, D. Boyle, J. Pearson, P. St Pierre, L. Lock, P. Buckley, S. Mason, R. McCarthy, W. Garratt, K. Sugar and J. Pearce
Bell, E.; K. Floyd, D. Boyle, J. Pearson, P. St Pierre, L. Lock, P. Buckley, S. Mason, R. McCarthy, W. Garratt, K. Sugar and J. Pearce. The Isles of Scilly
seabird restoration project: the eradication of brown rats (Rattus norvegicus) from the inhabited islands of St Agnes and Gugh, Isles of Scilly
The Isles of Scilly seabird restoration project: the eradication of
brown rats (Rattus norvegicus) from the inhabited islands of
St Agnes and Gugh, Isles of Scilly
E. Bell1, K. Floyd1, D. Boyle1, J. Pearson2, P. St Pierre2, L. Lock2, P. Buckley2, S. Mason3, R. McCarthy4,
W. Garratt5, K. Sugar6 and J. Pearce7
Wildlife Management International Limited, PO Box 607, Blenheim 7240, New Zealand. <biz@wmil.co.nz>. 2Royal
Society for the Protection of Birds, UK Headquarters, The Lodge, Sandy, Bedfordshire SG19 2DL, United Kingdom.
3
Isles of Scilly Wildlife Trust, Trenoweth, St. Mary’s, Isles of Scilly TR21 0NS, UK. 4Lowertown Cottage, St Agnes,
Isles of Scilly, TR22 0PL, United Kingdom. 5Duchy of Cornwall, Hugh House, St Mary’s, Isles of Scilly. TR21 0HU,
UK. 6Natural England, Polwhele, Truro, Cornwall. TR4 9AD, UK. 7Isles of Scilly Area of Outstanding Natural Beauty,
Council of the Isles of Scilly, Town Hall, St Mary’s, Isles of Scilly. TR21 0LW, UK.
1
Abstract As part of the Isles of Scilly Seabird Recovery Project, and directed by Wildlife Management International
Ltd, the eradication of brown rats (Rattus norvegicus) from the inhabited islands of St Agnes & Gugh, Isles of Scilly was
completed between October 2013 and April 2014 with the assistance of volunteers, and staff from the Royal Society for
the Protection of Birds, Isles of Scilly Wildlife Trust and Natural England. Bait stations with cereal-based wax blocks
containing bromadiolone at 0.005% w/w were established on a 40–50 metre grid over the island. With the presence of
85 residents on the 142 ha islands, this is the largest community-based brown rat eradication globally to date. Given the
fact that a community is based on these islands, community engagement and advocacy was a vital and fundamental part
of the eradication. Consultation for eradication began three years prior to the operation to explain the requirements for
the proposed project and to assess support, but this built on many years of wider community engagement with seabird
conservation. All of the residents supported the eradication of rats and vision of the project. The consultation and inclusion
of the community in decision-making and management of the Isles of Scilly Seabird Recovery Project was a critical part of
the operation and key to the success of the eradication. The community took ownership of the project and has committed
to the on-going biosecurity requirements following the eradication of rats. The removal of brown rats from St Agnes and
Gugh was a major achievement and provided the opportunity to restore the islands’ communities of seabirds and other
native species. This project provided an example of the effectiveness of ground-based rodent eradication techniques on
an inhabited island and the lessons learnt during this operation can be used to help proposed eradications on other islands
with communities and with terrain suitable for ground-based techniques.
Keywords: brown rat, community, eradication, Isles of Scilly, Rattus norvegicus, St Agnes and Gugh
INTRODUCTION
The eradication of invasive species from islands has
become one of the most important tools in conservation
in recent times. It offers the opportunity that, following
an initial investment, significant long-term benefits
can be achieved. The eradication of rats is a recognised
prerequisite for the restoration of many seabird colonies on
islands. Rodents have been successfully eradicated from
over 700 islands around the world, including at least 10
UK islands (Moors & Atkinson, 1984; Atkinson, 1985;
Taylor, et al., 2000; Zonfrillo, 2001; Towns & Broome,
2003; Appleton, et al., 2006; Howald, et al., 2007; Jones,
et al., 2008; Bell, et al., 2011; Parks & Wildlife Service,
2014; DIISE, 2015; Thomas, et al., 2017; Bell, 2019; Bell,
et al., 2019; Pearson, et al., 2019), However, most of
these islands have been uninhabited. Many consider that
islands with significant human populations, unreceptive
communities or occurrence of livestock and domestic
animals are unlikely to be feasible for eradication (Oppel, et
al., 2011; Campbell, et al., 2015; Russell & Broome, 2016;
Stanbury, et al., 2017). However, an increasing number of
eradications are being considered on inhabited islands and
the importance of the engagement and inclusion of local
communities has been highlighted in a number of recent
eradication and research projects, especially in regard
to risk and benefit analysis (Eason, et al., 2008; Bryce,
et al., 2011; Oppel, et al., 2011). It should be noted that
the greatest conservation benefit to be gained from future
eradications in the UK, and in other parts of the world,
is predominantly from inhabited islands (Stanbury, et al.,
2017). As such, it is vital that techniques and protocols
developed during eradications on islands with even small
communities should be assessed, utilised or adapted for
these islands with larger communities.
The Isles of Scilly are a nationally and internationally
important location for seabirds, particularly Manx shearwater
(Puffinus puffinus), European storm petrel (Hydrobates
pelagicus) and black-backed gull (Larus fuscus) (Lock,
et al., 2006). Both Manx shearwaters and European storm
petrels are amber listed under the United Kingdom Birds
of Conservation Concern threat categorisation (Eaton, et
al., 2015). A partnership of organisations (Royal Society
for the Protection of Birds (RSPB), Natural England (NE),
Isles of Scilly Wildlife Trust (IOSWT) and Isles of Scilly
Bird Group (IOSBG)) produced the Isles of Scilly Seabird
Conservation Strategies 2005–2008 and 2009–2013
which described the national and international status and
context of the seabird populations on the Isles of Scilly
and identified priority actions and strategic goals for
management. These included current and future measures
to improve the available habitat for seabirds through rat
control and eradication (Lock, et al., 2006; Lock, et al.,
2009). St Agnes and Gugh have a number of important
land areas designated for seabirds as Special Protected
Areas (SPA), Sites of Special Scientific Interest (SSSI)
and Ramsar (Lock, et al., 2009). The eradication of
brown rats (Rattus norvegicus) from St Agnes and Gugh
was identified as a priority in these strategies as it would
remove predation pressure on Manx shearwaters and storm
petrels and provide the opportunity for other seabirds to
colonise the islands (Lock, et al., 2006; Lock, et al., 2009).
These strategies also recognised the social, economic and
health benefits for the local community (Lock, et al., 2006;
Lock, et al., 2009).
The Isles of Scilly Seabird Recovery Project (IOSSRP)
was established in 2010 and was managed by a coalition of
In:
88C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 88–94. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Bell, et al.: Rats off inhabited St Agnes and Gugh, UK
groups including RSPB, IOSWT, NE, Duchy of Cornwall
(DC), the Isles of Scilly Area of Outstanding Natural Beauty
(AONB) partnership and a representative from St Agnes
and Gugh, with support from the IOSBG. The IOSSRP
partnership identified the need to assess the possibility of
eradicating brown rats from St Agnes and Gugh to protect
and enhance the islands’ seabirds and protect Annet from
re-invasion. Annet is the most important uninhabited island
for seabirds in the Isles of Scilly as it has always been
rat-free (excluding an incursion in 2004, probably from
neighbouring St Agnes) and holds the main populations
of Manx shearwaters and European storm petrels (Lock,
et al., 2006). The partnership commissioned a feasibility
assessment in 2010 (Bell, 2011). A formal IOSSRP
Steering Group made up of representatives from all Project
Partners was established in 2012. Wildlife Management
International Ltd. (WMIL) directed the eradication with
the assistance of volunteers and RSPB, IOSWT and NE
staff. The eradication was completed between October
2013 and April 2014 (Bell, et al., 2014). This paper covers
the technical aspects of the St Agnes and Gugh brown
rat eradication and complements the Pearson, et al., (this
issue) paper on the community aspect of the eradication.
STUDY AREA AND METHODS
St Agnes and Gugh
St Agnes and Gugh (49.89267°N, 6.34073°W) are two
islands in the Isles of Scilly archipelago off the Cornish
coast, in south-west England (Fig. 1). St Agnes (105 ha)
and Gugh (37 ha) are connected by a rock and sand bar
at low tide (Fig. 1). St Agnes and Gugh are separated
from St Mary’s by a deep channel (St Mary’s Sound) that
is 1.1 kilometres at the closest point (via stepping stone
islands) or 1.3 km from shore to shore (Fig. 1). There are
85 residents, only two of whom live on Gugh. Brown
rats were accidentally introduced to the Isles of Scilly
from shipwrecks in the 1700s, and were widespread and
abundant across both islands, as well as many other islands
in the archipelago (Matheson, 1962; McCann, 2005).
Tourism is one of the islands’ major sources of income,
particularly between April and October.
There are approximately 40 homes on the island, but
at least 150 buildings (holiday lets, farm buildings, sheds,
etc.) scattered across the whole island. There are six farms
(including a chicken farm and dairy), a campground, a
school, a restaurant, a pub, two cafes, a post office and
store. There are cattle, chickens, ducks, geese, two ponies
and pigs on St Agnes. Many families have pet cats and
dogs. There is a main quay where passengers and freight
are landed, and a smaller slipway used mainly by residents.
These factors increased the number of challenges such as
providing alternative food and shelter for rats, risk to nontarget species and biosecurity.
The main habitats on St Agnes are farmland, mainly
flower farms and low intensity cattle grazing, characterised
by small fields with extensive hedges and stone walls,
ponds, maritime grassland, invasive Pittosporum, rocky
shores and sandy beaches (Parslow, 2007). St Agnes and
Gugh are home to the only known populations in the British
Isles of a number of rare plants, including least adder’stongue fern (Ophioglossum lusitanicum) (Parslow, 2007).
Rabbits (Oryctolagus cuniculus), Scilly shrews
(Crocidura suaveolens cassiteridum) and pipistrelle bats
(Pipistrellus spp.) are the only other known species of
mammal found on St Agnes and Gugh, apart from livestock
and pets. House mice (Mus musculus domesticus) were
present on St Agnes and Gugh, but have not been seen in
at least 15 years, though mice are still present on most of
the other main islands in the Scillies (Howie, et al., 2007).
Eradication operation
The eradication programme ran from 11 October 2013
to 11 April 2014 and included establishing the bait station
grid, poisoning, monitoring and biosecurity establishment.
This phase took 1,593 person days. Long-term monitoring
ran monthly between May 2014 and December 2015. The
final check, species monitoring, and rat-free declaration
ran from 6 January to 18 February 2016. This phase took
250 person-days. All IOSSRP personnel wore blaze-orange
hats (with the IOSSRP logo) to be easily recognisable to
the community and visitors. Each operational task was
undertaken and completed as follows:
Pre-eradication
Due to the presence of a community on the island
and the selected method of bait stations, different preeradication preparation tasks were required compared to
aerial baiting methods. Preparation tasks included, but
were not limited to: consultations with the community
about operational techniques; timing of each aspect of the
project and confirming access to land and buildings; testing
rats for resistance to rodenticides; getting the community
to cease using rodenticides on the island six months prior
to the eradication (i.e. to prevent bait aversion, avoid rats
becoming accustomed to bait and to prevent resistance);
removal of waste, alternative food and harbourage
(including cleaning up farm sheds and other buildings
on the island); establishing waste management systems
for each household and business (including provision of
rodent proof wheelie-bins and compost bins); application
for an extension-of-use for rodenticide use from the UK
Health and Safety Executive (HSE); construction of bait
stations; and delivery of all equipment to the islands.
The University of Reading completed resistance testing
and DNA screening of 26 rats trapped on the islands. Of
these samples, resistance (L120Q mutation) was detected
in one individual (Rymer, 2013). This resistance evidence
confirmed the requirement for multiple toxin and bait
formulations to ensure any problem rats could be targeted
successfully. An extension-of-use permission from HSE
was obtained to use specific rodenticides (difenacoum and
brodifacoum) at specific locations outdoors if it became
necessary to target any resistant rats towards the end of the
eradication.
Fig. 1 Location of St Agnes and Gugh, Isles of Scilly, United
Kingdom.
Over 1,500 bait stations were constructed by RSPB
staff and volunteers in Penzance and these and all other
equipment was delivered to St Agnes in September 2013.
89
Island invasives: scaling up to meet the challenge. Ch 1B Rodents: Review
Bait station grid
The bait station grid was established between 12
October and 7 November 2013. Bait stations were made
from 750 mm lengths of 100 mm diameter corrugated black
plastic drainage pipes, wired into the ground to prevent
movement by animals and/or wind. Bait was placed in the
centre of the station through the access hole that is covered
by an additional short section of pipe and held in place by
a ‘crow clip’ (a short piece of wire wrapped around the
centre of the station devised during the Lundy Island rat
eradication operation which prevents the crows and gulls
removing the lids (Bell, 2019)).
Bait stations were placed out on a 40 m × 50 m grid.
Positions were determined by electronic Geographic
Information System (GIS) and loaded onto a hand-held
GPS unit. Each station was marked by a bamboo cane or
flagging tape to ensure visibility in thick vegetation or poor
weather.
The entire grid of 962 tube stations was positioned
across the island (with an additional 74 commercial
Protecta™ lockable bait stations inside all private homes,
holiday rentals, public buildings and on the quay) before
being individually numbered and mapped using GPS and
added to a GIS-linked database (Fig. 2).
Poisoning
The main toxicant used was bromadiolone, Contrac™
(manufactured by Bell Laboratories), a 28 g, cereal-based
wax block bait with 0.005% active ingredient. This bait was
used between 8 November 2013–12 January 2013 and 27
January–8 March 2014 (Table 1). There were two alternative
baits, both manufactured by PelGar International, available
if any rats were detected that seemed to be avoiding or
appeared to be resistant to the main bait: Roban Excel™,
a 20 g cereal-based block bait with active ingredient
difenacoum at 0.005% w/w that was used between 13–26
January 2014 (Table 1); and Vertox Oktablok II™, a 20 g
cereal-based block with active ingredient brodifacoum at
0.005% w/w that was not required. Contrac™ and Roban
Excel™ are dyed blue (or green/blue) to be less attractive
to birds (Caithness & Williams, 1971; Hartley, et al., 1999;
Weser & Ross, 2013), thus helping to further reduce risks
to non-target species.
The poisoning operation commenced on 8 November
2013 and continued through to 8 March 2014. Baits
were present in each station throughout the poisoning
programme and replaced as required; when eaten by rats,
by non-target species such as invertebrates and/or damaged
by weather. Between 8 and 18 November 2013 there were
eight blocks of bait in each station. This was reduced to four
blocks between 19 and 25 November 2013 and reduced
again to two blocks from 26 November 2013 to 26 January
2014 (Table 1). After 27 January 2014, only one block of
bait was placed in each station. Existing undamaged bait
blocks were left in the stations and the extra blocks were
removed. All waste and partially eaten bait was collected
Fig. 2 Bait station grid on St Agnes and Gugh, Isles of
Scilly. Bait station positions are marked by a black dot.
and incinerated in a high temperature incineration facility
at the end of the operation.
Bait was loose in the stations between 8 and 25
November 2013 (so that rats can take bait back to their
burrows to feed nursing females or young) and after 26
November all bait was wired into the stations (which could
be used to confirm the presence of rats due to teeth marks
being recorded on partially eaten blocks in the stations)
(Table 1).
Excluding the stations in the houses (which were
checked once a week), all other bait stations on St Agnes
and Gugh were checked and serviced at intervals between
one to seven days (a total of 56 bait checks over 120
days) depending on the stage of the operation (Table 2).
To present the data on bait-take gained from these varied
bait station checks we grouped the data into 27 periods or
checks (mean (±SEM) = 1.9 ± 0.2 days between checks,
range 1–7 days) shown as days from baiting (Fig. 3).
Bait-take was recorded in field notebooks by bait
station number and the species believed to have consumed
Fig. 3 Amount (in kg) of bait consumed by rats at each
bait check (marked by black dot) during the brown rat
(Rattus norvegicus) eradication on St Agnes and Gugh,
Isles of Scilly.
Table 1 Baiting regime during the brown rat (Rattus norvegicus) eradication on St Agnes and Gugh, Isles of Scilly, United
Kingdom.
Date
8–18 Nov 2013
19–25 Nov 2013
26 Nov 2013 to 12 Jan 2014
13–26 Jan 2014
27 Jan to 8 Mar 2014
90
Bait type
Contrac™ (bromadiolone)
Contrac™ (bromadiolone)
Contrac™ (bromadiolone)
Roban Excel™ (difenacoum)
Contrac™ (bromadiolone)
No of blocks
8
4
2
2
1
Bait loose or wired into station
Loose in station
Loose in station
Wired into station
Wired into station
Wired into station
Bell, et al.: Rats off inhabited St Agnes and Gugh, UK
Table 2 Number of bait station checks during the brown
rat (Rattus norvegicus) eradication on St Agnes and
Gugh, Isles of Scilly, United Kingdom.
Date
8–20 November 2013
21 Nov to 13 Dec 2013
14 Dec 2013 to 8 Mar 2014
Checks per week
6
5
3
or removed the bait. These data were entered into a GISlinked database and maps showing active stations were
produced in real-time to enable the team to effectively
monitor bait-take activity and target any “hot spots”.
Searches for carcasses were completed during all
checks. Any carcasses that were found, were collected,
necropsied to determine cause of death (where possible) and
incinerated to reduce risk for non-target scavengers. It was
expected that very few rat carcasses would be found on the
surface as most rats die underground in their burrows. Five
rat carcasses were found on the surface during the Lundy
Island rat eradication and three during the Isle of Canna rat
eradication (Bell, 2004; Bell, et al., 2006). Any non-target
species that were collected during the operation were also
necropsied and assessed for anticoagulant poisoning (i.e.
blood in body cavity, bruising, discolouration of organs).
Non-target species have been affected during other
eradications: 77 non-target species’ carcasses (greater
black-backed gull Larus marinus, carrion crow Corvus
corone, house sparrow Passer domesticus, short-eared
owl Afio flammeus and rabbit Oryctolagus cuniculus)
were found on the surface during the Lundy Island rat
eradication and seven non-target species carcasses (wood
mouse Apodemus sylvaticus, and pygmy shrew Sorex
minutus) were found during the Isle of Canna operation. Of
these, only 15 showed evidence of anticoagulant poisoning
and the remainder had died of starvation (rabbit, shrew) or
either natural (short-eared owl, crow) or unknown causes
(greater black-backed gulls) (Bell, 2004; Bell, et al., 2011).
Monitoring
Three distinct periods of monitoring were undertaken
as the project progressed. Intensive monitoring using 2,500
stations at 25 m spacing was carried out from 19 November
2013 to 8 March 2014 to detect rats surviving through the
poisoning phase. This was followed by a 21-month period
of long-term monitoring using 87 biosecurity stations and
six rodent motels (wooden boxes designed to provide an
attractive, alternative ‘burrow’ for rats during an incursion)
from 9 March 2014 to 5 January 2016. These biosecurity
stations were established at high risk areas on the island;
around the coast, at the quay and other boat landing sites
and at seabird breeding sites (Bell, et al., 2014). The final
monitoring check, using 448 stations, was carried out
between 6 January and 18 February 2016 (Bell & Cropper,
2016). WMIL and RSPB staff and volunteers carried out the
intensive and final checks and IOSSRP staff, St Agnes and
Gugh residents and volunteers maintained the long-term
monitoring. Monitoring stations consisted of materials
attractive to rats that would also clearly show teeth marks
(e.g. chocolate, peanut or coconut flavoured wax, candles
and soap), tracking tunnels and trail cameras (Bushnell™).
All were individually numbered and any evidence of
activity (e.g. teeth marks or foot prints) was recorded in
field notebooks by station number and the species believed
to have consumed or marked the monitoring item.
Monitoring items were placed inside and outside each
station as well as halfway between each station during the
intensive monitoring phase and final monitoring check.
During these monitoring phases, each monitoring site
was checked regularly 3–5 times a week (depending on
weather), either separately or – during the poisoning phase
– together with the poisoning bait station grid. Monitoring
items were placed inside the biosecurity stations only
during the long-term monitoring phase and these were
checked monthly. Checks for active rat runs and activity
at high-risk sites (i.e. stone walls, farms, seabird colonies,
etc.) were also undertaken throughout all three monitoring
phases. Any rat and non-target species sign found on any
monitoring detection device at any stage of the monitoring
phase was recorded and added to the database.
RESULTS
Bait acceptance and take
Bait acceptance was excellent with no evidence of bait
avoidance. Green/blue rat droppings appeared within three
days and rats accounted for 203.6 kg of Contrac™ bait
taken (estimated 1,600–2,500 rats).
The bait-take pattern was typical of other rat eradication
campaigns (Thomas & Taylor 2002; Bell, et al., 2011). It
was very high in the immediate days after original baiting
(checks 1–3) and dropped to a relatively low level eight
days after original baiting (check 8) (Fig. 3). A small
increase was recorded at day 21 after the original baiting
(check 15) but dropped away, reaching zero bait-take on
day 23 after the original baiting (check 17) (Fig. 3).
Throughout the poisoning phase, 62% of bait stations
were visited by rats, with 42.7% active within the first
three days of the original baiting. This level of activity
was similar to the Lundy and Isle of Canna eradications
which had 42.5% and 62% of bait stations visited by rats,
respectively (Bell, 2004; Bell, et al., 2011). The high
number of active bait stations during the first two bait
checks shows that the rats quickly accepted the bait across
St Agnes and Gugh. It is likely that the small grid size and
intensive baiting regime targeted the rats effectively within
a short timeframe.
The average number of blocks taken by rats was 4.3 (±
0.1) blocks per active station (range 0–41 blocks). Again,
this level of activity was similar to the Lundy and Isle of
Canna eradications which had 3.2 and 8 blocks taken by
rats by per active station, respectively (Bell, et al., 2004;
Bell, et al., 2011). This also indicates that rats were quickly
removed from most sites across St Agnes and Gugh. As
shown by Fig. 4, bait-take was not evenly distributed
over both islands, with the greatest level of bait-take on
the coastal areas of both islands and each of the offshore
Fig. 4 Distribution of total bait-take (g) by rats consumed
per station during the brown rat (Rattus norvegicus)
eradication on St Agnes and Gugh, Isles of Scilly.
91
Island invasives: scaling up to meet the challenge. Ch 1B Rodents: Review
rock stacks connected to the main islands at low tide. The
distribution of rats and density on Gugh was likely to be
having an impact on Manx shearwaters and other seabirds
and land bird and invertebrate populations present on St
Agnes and Gugh.
There were 19 rat carcasses collected on the surface
during the operation. These were collected and incinerated
to prevent availability to non-target species.
There were low levels of interference by non-target
species with nearly 54 kg of bait being consumed; cattle
kicked up stations and ate a small amount of bait (1.4
kg), slugs and other insects consumed 51.9 kg and shrews
consumed 0.4 kg. The weather conditions also complicated
the operation and accounted for 3.4 kg of bait that had to
be replaced due to the loss of 54 bait stations in storms.
Carcasses of a water rail (Rallus aquaticus), a song thrush
(Turdus philomelos), a blackbird (T. merula) and nine
Scilly shrews were found. There was no evidence that any
of these non-target species was affected by the rodenticide.
Monitoring
Monitoring for rat presence continued island-wide
for two years after the end of the poisoning operation.
The last rat was detected on chocolate flavoured wax
on 29 November 2013 during the overlap between the
poisoning and intensive monitoring phases and this rat was
successfully targeted using the main bait, Contrac™, by 2
December 2013. No rats or sign were detected during any
phase of the long-term or final check monitoring. St Agnes
and Gugh were declared rat-free in February 2016.
Cattle, shrews and birds interfered with 899 monitoring
stations (by eating the flavoured wax or soap, marking
tracking plates or, in the case of cattle, by removing the
monitoring wires) a total of 12,156 times between 21
November 2013 and 26 February 2014. There were 127
stations affected 1,384 times by cattle, 60 (82 times) by
birds, 5 (8 times) by insects, 9 (9 times) by rabbits and
454 (2084 times) by shrews. Interference by birds, shrews
and rabbits was limited to teeth or beak marks on the
soap or flavoured wax or footprints on tracking plates.
Cattle removed wires and ate flavoured wax and soap, so
monitoring points had to be moved or hidden in those areas
with cattle.
DISCUSSION
The success of the St Agnes and Gugh brown rat
eradication shows that a well-planned, adequately
resourced, well-executed programme, with the complete
support of the community, local agencies and government
and directed by an experienced operator with dedicated
workers, can eradicate rats from inhabited islands using
a ground-based bait station operation. The project on St
Agnes and Gugh is the largest community-led (with 85
residents) brown rat eradication anywhere in the world.
Most other eradications on inhabited islands either have
smaller communities (e.g. Isle of Canna, 12 residents;
Bell, et al., 2011; Rakino in New Zealand, 16 residents;
Bassett, et al., 2016) or have staff or a military population
(e.g. Bird, Denis, Curieuse and Fregate Islands in the
Seychelles, Merton, et al., 2002; Lundy Island, Bell,
2004; Wake Island, Brown, et al., 2013) and have not had
direct involvement of the community during and after the
eradication or leaving the community responsible for all
biosecurity measures (Pearson, et al., this issue).
However, the success of the eradication was dependent
on the participation and support of the entire local
community. The community maintained an integral role
and was consulted extensively in the planning, preparation
and implementation of the eradication programme. As
92
such, it is vital that techniques and protocols developed
during eradications on islands with even small communities
should be assessed, utilised or adapted for islands with
larger communities. The opinions and safety of local
communities need to be a priority in any eradication
planned for inhabited islands.
Stock and chicken feed provided a possible alternative
food source for rats, but all the farmers were fully
supportive of the project and stored all the unopened feed
on pallets (with bait stations and/or traps underneath) or in
rodent-proof containers and any opened feed was stored
in large plastic, metal or wooden sealed bins. Where
possible, farm buildings were kept clean to ensure fresh
sign was quickly noted. All these methods meant that the
sheds were cleared of rats and any roaming rats which reinvaded the area could be noted quickly. The presence of
a large chicken farm could have been a major problem
as their runs provide excellent rat habitat and alternative
food. The owner of the chicken farm strictly managed his
chickens and feeding regime throughout the rat eradication
operation which made targeting rats and monitoring for
any survivors on this farm easier.
Rubbish can be the most serious issue on an inhabited
island wanting to eradicate rats. This was discussed
comprehensively with the community before the project
commenced. As a result, rat-proof wheelie bins and Green
Johanna compost bins were provided to the residents
and all rubbish was stored in these prior to removal to
St Mary’s. Rubbish was removed regularly (generally
weekly) from St Agnes to St Mary’s by the Isles of Scilly
Council. In October and early November 2013, with the
permission and assistance of residents, a number of sheds,
farm buildings and outhouses were cleared and tidied by
the IOSSRP team to ensure bait stations could be placed
along all the walls.
St Agnes and Gugh were cleared of rats within three
weeks (23 days from original baiting). Bait-take showed
that the rat population appeared to be low (approximately
2,000 rats) and was not evenly distributed across the
islands. There were high concentrations of rats on Gugh
and around the coastal areas on St Agnes where the
burrow-nesting seabird colonies are present, meaning rats
were likely to have been having an effect on these breeding
seabirds (Moors & Atkinson, 1984; Atkinson, 1985; Jones,
et al., 2008).
The interference by cattle was another major factor
affecting the operation, with cattle kicking up or crushing
stations, but cooperation by the farmers to move stock
around different paddocks, as well as altering the bait
station positions, wiring the bait or lids into position in
addition to the crow clip or weighting the stations down
with rocks, meant this problem was quickly dealt with.
Many of the monitoring stations were removed from, and
then replaced back into, certain areas (such as Covean and
Wingletang) as the cattle were rotated between paddocks.
Importantly, there were no known non-target species
affected by this operation. Although a small number of
Scilly shrews (n = 9) were found dead and necropsied
during the eradication, proof of poisoning could not be
confirmed (i.e. no symptoms of anticoagulant poisoning
such as blood in body cavity, bruising or discolouration
of organs). However, no liver or tissue samples were
taken from non-target species for further analysis. It
should be noted that, in certain cases, bait-take by shrews
subsequently stopped in nearby stations suggesting these
animals had died due to primary poisoning. Although there
is no information on the LD50 for shrews, using LD50 data
from other small mammals (voles and mice), it is likely
that shrews would have to eat between 0.2–1.25 mg/kg to
be affected by bromadiolone. This amounts to 0.001 blocks
Bell, et al.: Rats off inhabited St Agnes and Gugh, UK
of bait and this level of bait take by shrews occurred at
83 different stations between 22 November 2013 and 5
March 2014 suggesting that approximately 83 shrews may
have been affected by the baiting phase (totalling to 0.4
kg of bait). However, it is thought that as Scilly shrews
have small home ranges (< 50 m2; Spencer-Booth, 1963;
Rood, 1965), excluding those with a bait station in their
immediate home range, most shrews would not encounter
bait stations or poisoned invertebrates using the 40 m × 50 m
grid. This means that even if a small number of individuals
was killed, the overall population would survive. The risk
to the shrew population was considered minimal, but the
potential for a small number of individuals to be affected
was acknowledged (Bell, 2011). Calculations of baittake indicate that more shrews than anticipated may have
been at risk, but extensive searches for carcasses and the
necropsies performed do not support this; there was no
definitive evidence of any shrew death being attributable
to the rodenticide. Scilly shrew numbers have increased to
population levels higher than those before the eradication
(IOSSRP, unpublished data).
A large quantity of bait was consumed or damaged
by slugs and other insects. Bait was changed often to
ensure there was always the most attractive and palatable
bait available to rats. Contrac™ was more durable than
expected, compared to earlier experience on Lundy
Island where it deteriorated within one to two days (Bell,
2004), meaning it lasted better in the St Agnes and Gugh
environment. Occasionally it was difficult to interpret sign
on the blocks during the important monitoring phase of
the operation, owing to the nature of the block and ridges,
but the Contrac™ bait successfully targeted all rats on St
Agnes and Gugh within three weeks.
There was no evidence that any other non-target species
were affected by the rodenticide, traps or monitoring tools
used in the operation. Following necropsy of shrews and
other non-target species carcasses (water rail, thrush and
blackbird), there was no bait found in the stomach or
symptoms of anticoagulant poisoning (i.e. blood in the
body cavity, bruising or haemorrhaging or discoloured
organs). Although 19 dead rats were found on the surface
(1.1% of estimated rat population on St Agnes and Gugh),
there was no evidence of any other animal scavenging these
carcasses. There were no observations of pet cats, crows,
gulls or raptors eating dead or dying rats on St Agnes and
Gugh.
Weather also affected the eradication when storms
removed or dislodged stations, but this generally was
limited to coastal areas.
The eradication of invasive species such as rats from
islands has become one of the most important tools in
avian conservation worldwide. It was recognised that
for the restoration and protection of seabird colonies on
St Agnes and Gugh, the eradication of rats was required.
This operation has already benefited key seabird species
on the islands as well as the Scilly shrew as shown by
comparisons between the pre- and post-eradication
biodiversity monitoring. Manx shearwaters were recorded
successfully breeding within one year of the eradication
and 73 pairs were recorded in 2016 compared to 22 pairs
and no fledged chicks in 2013 (Pearson, 2016). European
storm petrels were first recorded on St Agnes in 2015,
with 9 pairs in 2016, and the Scilly shrew population has
increased to levels higher than the pre-eradication levels
since rats have been eradicated (IOSSRP, unpublished
data; Pearson, 2016; Thomas, et al., 2017).
Although eradicating rats from St Agnes and Gugh is
a considerable and significant achievement, it is important
to stress that keeping these islands rat-free will require
constant vigilance and commitment from the whole
community, partner agencies and visitors in order to
prevent, detect and respond to any incursions. Prevention
of an accidental rat re-introduction should be the primary
aim. The greatest risk is via service and private vessels
traveling between all of the inhabited islands in the Isles
of Scilly, especially if delivering farming equipment, hay,
stock feed, equipment or food to St Agnes. There is also
a small risk from visiting yachts and general tourism.
Permanent biosecurity stations have been established on
St Agnes and Gugh; these will be maintained indefinitely
by trained community members and IOSSRP personnel. A
detailed biosecurity plan has been developed to prevent,
detect and respond to possible incursions. Residents have
been trained in these biosecurity measures, identification
of rodents and rodent sign, and methods to reduce the risk
of accidentally introducing rodents, demonstrating the
commitment of the St Agnes and Gugh community to the
restoration of their islands.
It is important to stress that the eradication of brown
rats from St Agnes and Gugh is a valuable education tool to
show other island communities that it is possible to safely
eradicate rats without unduly impacting on the lives and
habits of the local residents. The successful eradication
of brown rats from St Agnes and Gugh demonstrates how
the techniques of ground-based bait station operations can
be utilised on inhabited islands throughout the UK and
the world where this technique is feasible and where the
community is involved and supportive.
ACKNOWLEDGEMENTS
This project was carried out with funding from EU
LIFE (Scilly Isles LIFE Project 11 UK NAT 387), UK’s
Heritage Lottery Fund (HG-11-06880) and the Defrafunded Isles of Scilly Area of Outstanding Natural
Beauty (AONB) Partnership Sustainable Development
Fund. The Contrac™ bait, T-Rex™ and Protecta™ bait
stations were generously donated by Bell Laboratories.
The Roban Excel™ and Vertox Oktablok II™ bait were
donated by PelGar International Ltd. Our heartfelt
thanks go to the entire St Agnes and Gugh community
for their support and enthusiasm throughout the project,
supplying goods and services to the team, providing rat
sightings, and for their willingness to implement new
waste management and ongoing biosecurity procedures.
We would also like to thank Sophie Thomas (RSPB),
David Appleton (Natural England), Michelle Hawkins
(Natural England), Peter Exley (RSPB), Tony Whitehead
(RSPB), Trevor Kirk (AONB), Rebecca Steggles (AONB),
Dave Mawer (IOSWT), Darren Hart (IOSWT), Dr. Alan
Buckle (Reading University), Dr. Colin Prescott (Reading
University) and David Rymer (Reading University). St
Agnes Boating, The Isles of Scilly Steamship Company,
St Mary’s Boating Association, Bryher Boats and St
Martin’s Boating Association all provided boat transport.
Samantha Sadighi, Great Green Systems Limited, supplied
the compost bins for the St Agnes & Gugh community
at a discounted rate. Special thanks to all the IOSSRP
Seabird Taskforce Volunteers – Helen Brandes, Haydn
Brooks, Alex Cropper, Bob Dawson, Lindsey Death, Neil
Duffield, Paul Garner-Richards, Richard Halstead, Sarah
Havery, Vickie Heaney, Jack Ibbotson, Cal James, Helene
Jessop, Darren Mason, Tegan Newman, Will (Billy)
Renny, Jack Roper, Rebecca Steggles, Colin Taylor, Toby,
Rachel and Otto Taylor, Karen Varnham, Alastair Wilson,
Will Whittington and Lewis Yates for their hard work,
enthusiasm and support throughout the project. Extra
thanks to Jenny Parker (RSPB) and David Flumm (RSPB)
and their team of mainland-based volunteers who made the
1,500 bait stations.
93
Island invasives: scaling up to meet the challenge. Ch 1B Rodents: Review
REFERENCES
Appleton, D., Booker, H., Bullock, D., Cordrey, L. and Sampson, B.
(2006). ‘The seabird recovery project: Lundy Island’. Atlantic Seabirds
8: 51–59.
Atkinson, I.A.E. (1985). ‘The Spread of Commensal Species of Rattus
to Oceanic Islands and their Effects on Island Avifaunas’. In: P.J.
Moors (ed.) Conservation of Island Birds, pp. 35–81. ICBP Technical
Publication 3. Cambridge, UK: International Council for Bird
Preservation (ICBP).
Bassett, I., Cook, J., Buchanan, F. and Russell, J. (2016). ‘Treasure
Islands: biosecurity in the Hauraki Gulf Marine Park’. New Zealand
Journal of Ecology 40: 250–266.
Bell, E. (2004). Lundy Seabird Recovery Project 2002–2004: The
Eradication of Black and Brown Rats from Lundy Island. Unpublished
client report to the Royal Society for the Protection of Birds, SouthWest Regional Office, Devon, UK. Blenheim, New Zealand: Wildlife
Management International Ltd.
Bell, E., Boyle, D., Garner–Richards, P. and Floyd, K. (2006). Canna
Seabird Recovery Project: Phase 1: Eradication of Brown Rats
2005–06. Unpublished client report to the National Trust for Scotland,
Edinburgh, Scotland. Blenheim, New Zealand: Wildlife Management
International Ltd.
Bell, E (2011). Isles of Scilly Seabird Recovery Project: Part 3: Assessing
the Feasibility of Removing Brown Rats from St Agnes and Gugh, Isles
of Scilly. Unpublished client report to Isles of Scilly Seabird Recovery
Project Partnership. Blenheim, New Zealand: Wildlife Management
International Ltd.
Bell, E., Boyle, D., Floyd, K., Garner–Richards, P., Swann, B.,
Luxmoore, R., Patterson, A. and Thomas, R. (2011). ‘The groundbased eradication of Norway rats (Rattus norvegicus) from the Isle
of Canna, Inner Hebrides, Scotland’. In: C.R. Veitch, M.N. Clout and
D.R. Towns (eds.) Island invasives: eradication and management, pp.
269–274. Occassional Paper SSC no. 42. Gland, Switzerland: IUCN
and Auckland, New Zealand: CBB.
Bell, E, Boyle, D., Tayton, J. and Floyd, K. (2014). The Isles of Scilly
Seabird Recovery Project: Technical Report for the Rat-removal Phase
on St Agnes and Gugh, October 2013 to April 2014. Unpublished client
report to the Isles of Scilly Seabird Recovery Project Partnership.
Blenheim, New Zealand: Wildlife Management International Ltd.
Bell, E. and Cropper, A. (2016). The Isles of Scilly Seabird Recovery
Project: Final Check Report Following the Eradication of Brown Rats
(Rattus norvegicus) from St Agnes & Gugh, 6 January to 18 February
2016. Unpublished client report to the Isles of Scilly Seabird Recovery
Project Partnership. Blenheim, New Zealand: Wildlife Management
International Ltd.
Bell, E.A. (2019). ‘It’s not all up in the air: the development and use of
ground-based rat eradication techniques in the UK’. In: C.R. Veitch,
M.N. Clout, A.R Martin, J.C Russell and C.J. West (eds.) Island
invasives: scaling up to meet the challenge, pp. 79–87. Occasional
Paper SSC no. 62. Gland, Switzerland: IUCN.
Bell, E., Bell, M., Morgan, G. and Morgan, L. (2019). ‘The Recovery
of Seabird Populations on Ramsey Island, Pembrokeshire, Wales
Following the 1999/2000 Rat Eradication’. In: C.R. Veitch, M.N. Clout,
A.R Martin, J.C Russell and C.J. West (eds.) Island invasives: scaling
up to meet the challenge, pp. 539–544. Occasional Paper SSC no. 62.
Gland, Switzerland: IUCN.
Brown, D., Pitt, W. and Tershy, B. (2013). A Review of the Wake Island Rat
Eradication Project. Honolulu, Hawaii: US Fish and Wildlife Service.
Bryce, R., Oliver, M., Davies, L., Gray, H., Urquhart, J. and Lambin,
X. (2011). ‘Turning back the tide of American mink invasion at an
unprecedented scale through community participation and adaptive
management’. Biological Conservation. 144: 575–583.
Caithness, T. and Williams, G. (1971). ‘Protecting birds from poisoned
baits’, New Zealand Journal of Agriculture 122: 38–43.
Campbell, K., Beek, J., Eason, C., Glen, A., Godwin, G., Gould, F.,
Holmes, N., Howald, G., Madden, F., Ponder, J., Threadgill, D.,
Wegmann, A. and Baxter, G. (2015). ‘The next generation of rodent
eradications: Innovative technologies and tools to improve species
specificity and increase their feasibility on islands’. Biological
Conservation 185: 47–58.
DIISE. (2015). The Database of Island Invasive Species Eradications,
developed by Island Conservation, Coastal Conservation Action
Laboratory UCSC, IUCN SSC Invasive Species Specialist Group,
University of Auckland and Landcare Research New Zealand. <http://
diise.islandconservation.org>.
Eason, C., Ogilvie, S., Miller, A., Henderson, R., Shapiro, L., Hix, S.,
MacMorran, D. and Murphy, E. (2008). ‘Smarter Pest Control Tools
with Low-residue and Humane Toxins’. In: R.M. Timm and J.M.
O’Brien (eds) Proceedings of the 23rd Vertebrate Pest Conference pp.
148–153. Davis, California, USA: University of California.
Eaton, M., Aebischer, N., Brown, A., Hearn, R., Lock, L., Musgrove, A.,
Noble, D., Stroud, D. and Gregory, R. (2015). ‘Birds of conservation
concern 4: the population status of birds in the UK, Channel Islands and
Isle of Man’. British Birds 108: 708–746.
Hartley, L., O’Connor, C., Waas, J. and Matthews, L. (1999). ‘Colour
preferences in North Island robins (Petroica australis): Implications for
deterring birds from poisonous baits’. New Zealand Journal of Ecology
23: 255–259.
94
Howald, G., Donlan, C., Galvan, J., Russell, J., Parkes, J., Samaniego, A.,
Wang, Y., Veitch, D., Genovesi, P., Pascal, M., Saunders, A. and Tershy,
B. (2007). ‘Invasive rodent eradication on islands’. Conservation
Biology 21: 1258–1268.
Howie, A., Jelbert, K. and Doyle, J. (2007). ‘Mammals in Cornwall and
the Isles of Scilly: Project Findings’. In: S. Board (ed.) Celebrating
Cornwall’s Mammals – from Dormice to Dolphins, pp. 1–70. Cornwall,
UK: Environmental Records Centre for Cornwall and Isles of Scilly
(ERCCIS).
Jones, H., Tershy, B., Zavaleta, E., Croll, D., Keitt, B., Finkelstein, M.
and Howald, G. (2008). ‘Review of the global severity of the effects of
invasive rats on seabirds’. Conservation Biology 22: 16–26.
Lock, L., Brown, A., Clitherow, J., Mawer, D. and St Pierre, P. (2006).
Isles of Scilly Seabird Conservation Strategy. Unpublished client report
to the Isles of Scilly Seabird Conservation Group. Exeter, UK: Royal
Society for the Protection of Birds (RSPB).
Lock, L., Brown, A., Clitherow, J., Mawer, D. and St Pierre, P. (2009).
Isles of Scilly Seabird Conservation Strategy. Unpublished client report
to the Isles of Scilly Seabird Conservation Group. Exeter, UK: Royal
Society for the Protection of Birds (RSPB).
Matheson, C. (1962). Brown Rats. London: Sunday Times Publications.
(Animals of Britain, No 16).
McCann, J. (2005). ‘The introduction of the brown rat (Rattus
norvegicus)’. Somerset Archaeology and Natural History Journal 2005:
139–141.
Merton, D., Climo, G., Laboudallon, V., Robert, S. and Mander, C. (2002).
‘Alien mammal eradication and quarantine on inhabited islands in the
Seychelles’. In: C.R. Veitch and M.N. Clout (eds.). Turning the tide:
the eradication of invasive species, pp. 182–198. Occasional Paper SSC
no. 28. IUCN SSC Invasive Species Specialist Group. IUCN, Gland,
Switzerland and Cambridge, UK.
Moors, P.J. and Atkinson, I.A.E. (1984). ‘Predation on Seabirds by
Introduced Animals and Factors Affecting its Severity’. In: J.P. Croxall,
P.G.H. Evans and R.W. Schreiber (eds.) Status and Conservation
of the World’s Seabirds, pp. 667–690. ICBP Technical Publication 2.
Cambridge, UK: International Council for Bird Preservation (ICBP).
Oppel, S., Beaven, B., Bolton, M., Vickery, J. and Bodey, T. (2011).
‘Eradication of invasive mammals on islands inhabited by humans and
domestic animals’. Conservation Biology 25: 232–240.
Parks and Wildlife Service. (2014). Evaluation Report: Macquarie Island
Pest Eradication Project, August 2014. Hobart, Tasmania: Department
of Primary Industries, Parks, Water and Environment. <http://www.
parks.tas.gov.au/file.aspx?id=31160>.
Pearson, J. (2016). ‘Manx shearwater and storm petrel numbers increase
on St Agnes and Gugh’. Seabirds South West 8: 2.
Pearson, J., St Pierre, P., Lock, L., Buckley, P., Bell, E., Mason, S.,
McCarthy, R. Garratt, W., Sugar, K. and Pearce, J. (2019). ‘Working with
the local community to eradicate rats on an inhabited island: securing
the seabird heritage of the Isles of Scilly’. In: C.R. Veitch, M.N. Clout,
A.R Martin, J.C Russell and C.J. West (eds.) Island invasives: scaling
up to meet the challenge, pp. 670–678. Occasional Paper SSC no. 62.
Gland, Switzerland: IUCN.
Parslow, R. (2007). The Isles of Scilly. London, UK: HarperCollins.
Rood, J. (1965). ‘Observations on the home range and activity of the
Scilly shrew’. Mammalia 29(4): 507–516.
Russell, J. and Broome, K. (2016). ‘Fifty years of rodent eradications in
New Zealand: another decade of advances’. New Zealand Journal of
Ecology 40: 197–204.
Rymer, D. (2013). Genetic Resistance to Anticoagulant Rodenticides in
Norway Rats (Rattus norvegicus): Findings from St. Agnes and Gugh,
the Isles of Scilly, June 2013. Unpublished technical report prepared for
IOSSRP partners. Reading, UK: University of Reading.
Spencer-Booth, Y. (1963). ‘A coastal population of shrews C. s.
cassiteridum’. Proceedings of the Zoological Society of London 140:
322–326.
Stanbury, A., Thomas, S., Aegerter, J., Brown, A., Bullock, A., Eaton, M.,
Lock, L., Luxmoore, R., Roy, S., Whitaker, S. and Oppel, S. (2017).
‘Prioritising islands in the United Kingdom and crown dependencies
for the eradication of invasive alien vertebrates and rodent biosecurity’.
European Journal of Wildlife Research 63: 31. < https://doi.org/10.1007/
s10344-017-1084-7>.
Taylor, R., Kaiser, G. and Drever, M. (2000). ‘Eradication of Norway rats
for recovery of seabird habitat on Langara Island, British Columbia’.
Restoration Ecology 8: 151–160.
Thomas, B. and Taylor, R. (2002). ‘A history of ground-based rodent
eradication techniques developed in New Zealand, 1959–1993’. In:
C.R. Veitch and M.N. Clout (eds.) Turning the tide: the eradication of
invasive species, pp. 301–310. Occasional Paper SSC no. 28. IUCN
SSC Invasive Species Specialist Group. IUCN, Gland, Switzerland and
Cambridge, UK.
Thomas, S., Brown, A., Bullock, D., Lock, L., Luxmoore, R., Roy, S.,
Stanbury, A. and Varnham, K. (2017). ‘Island restoration in the UK –
past present and future’. British Wildlife April 2017: 231–242.
Towns, D. and Broome, K. (2003). ‘From small Maria to massive
Campbell: Forty years of rat eradications from New Zealand islands’.
New Zealand Journal of Zoology 30: 377–398.
Weser, C. and Ross, J.G. (2013). ‘The effect of colour on bait consumption
of kea (Nestor notabilis): Implications for deterring birds from toxic
baits’. New Zealand Journal of Zoology 40(2): 137–144.
Zonfrillo, B. (2001). Wildlife Conservation on Ailsa Craig. The
Thomas Duncan Memorial Lectures. Girvan, Scotland: Friends of the
McKechnie Institute.
M.de L. Brooke
Brooke, M.de L. Rat eradication in the Pitcairn Islands, South Pacific: a 25-year perspective
Rat eradication in the Pitcairn Islands, South Pacific:
a 25-year perspective
M.de L. Brooke
Department of Zoology, University of Cambridge, Downing Street, Cambridge CB2 3EJ, UK.
<m.brooke@zoo.cam.ac.uk>.
Abstract This essay offers a 25-year overview of efforts to remove Pacific rats (Rattus exulans) from the four islands
of the Pitcairn group. Following the 1991–1992 discovery that rats were severely reducing breeding success of gadfly
petrels (Pterodroma spp.), Wildlife Management International proposed eradication. Eradication success was achieved
using ground-based baiting on the small atolls of Ducie and Oeno in 1997, and there is now evidence of petrel recovery on
Oeno, but two eradication attempts on inhabited Pitcairn (1997 and 1998) failed. By the early 2000s, the development of
aerial baiting through the 1990s placed an eradication operation on the fourth island, Henderson, within reach. Preparatory
fieldwork in 2009 allayed doubts in two key areas: the feasibility of maintaining a captive “back-stop” Henderson rail
(Porzana atra) population, and bait uptake by crabs (Coenobita spp.). Royal Society for the Protection of Birds (RSPB)
expertise secured the necessary funding of £1.5 million, and 75 tonnes of brodifacoum-containing bait were dropped in
August 2011. Despite extensive mortality of free-living rails, the population, supplemented by released captive birds,
returned to pre-operational levels in 2–3 years. Meanwhile those tending captive rails saw no rat sign before leaving
Henderson in November 2011. Unfortunately, a rat was sighted in March 2012, and continuing rat presence confirmed in
May 2012. Subsequently rat numbers have returned to pre-operational levels without any sign of population ‘overshoot’
as observed on Pitcairn. Genetic analysis suggests around 80 rats, roughly 1 in 1,000, survived the bait drop. With no
evidence of imperfect bait coverage or deficiencies in bait quality or brodifacoum resistance, it seems some animals
chose not to eat bait. Choice tests on Henderson Island rats suggest some rats prefer natural foods over bait. This adverse
situation may have been exacerbated because, in August 2011, natural fruits were more abundant than anticipated due to
drought earlier in the year. To overcome rat preference for natural food, any second Henderson attempt might benefit from
more attractive bait. Without such developments, a second attempt risks another failure. Henderson’s biota will survive
the delay.
Keywords: brodifacoum, Ducie, Henderson, Henderson rail, Oeno, Pterodroma
INTRODUCTION
The Sir Peter Scott Commemorative Expedition to the
Pitcairn Islands of 1991–1992 involved 35 personnel in
the field over a span of 15 months. While short periods
were spent on the sole inhabited island of Pitcairn (500 ha)
and the low atolls of Oeno (c. 60 ha) and Ducie (c. 75 ha),
Henderson Island (4300 ha) was the principal study site.
Since Henderson had been designated a World Heritage
Site in 1988 “as one of the last near-pristine limestone
islands of significant size in the world” (<http://whc.
unesco.org/en/list/487>), it had been appreciated that the
natural history of the island was incompletely documented.
The expedition aimed to rectify this omission, bringing
together expertise in archaeology, geology and many
branches of natural history.
One of the Expedition’s unexpected findings was the
very low breeding success of gadfly petrels (Pterodroma
spp.) on Henderson: ca. 5% among Murphy’s petrels (P.
ultima), 10% in Kermadec petrels (P. neglecta), and 15–
20% in Herald (P. heraldica) and Henderson petrels (P.
atrata) (Brooke, 1995). This was especially concerning in
the case of Henderson Petrels, split from Herald Petrels
as a result of expedition work (Brooke & Rowe, 1996),
endemic to Henderson and therefore without any source
of immigrants to rescue the situation, and potentially on a
downward trajectory to extinction within a few centuries
(Brooke, et al., 2010a).
Field observations showed that the cause of this low
breeding success was predation by Pacific rats (Rattus
exulans), introduced to the island by Polynesians settlers
about 700–800 years ago (Weisler, 1994). Hatching
success was apparently not substantially reduced by rats.
Rather, the problem arose in the first week after hatching,
especially when the chick moved from under to beside the
parent. Then the rats approached, pulled the chick away
from the nest site, even in the presence of a brooding
parent, and ate it (Brooke, 1995).
Observations on the atolls of Oeno and Ducie were too
intermittent to establish whether rats there had a similar
impact on the breeding success of petrels. However, the
fact that petrel densities were 1–2 orders of magnitude
higher on Oeno and Ducie than on Henderson prior to the
eradications on the atolls suggested that rat impact was
less, if not negligible. Probably because of the presence
of rats and feral cats (Felis catus), petrels do not breed on
Pitcairn.
After these findings had entered the public domain
via the expedition report (Pitcairn Islands Scientific
Expedition, 1992) and a special volume of the Biological
Journal of the Linnean Society (Benton & Spencer, 1995),
the late Brian Bell of Wildlife Management International
contacted the author to propose rat eradication in the
Pitcairn Islands (Bell & Bell, 1998). At this time, the mid1990s, an eradication on Henderson was not feasible using
ground-based methods. Therefore, the proposal was for
eradications on Oeno and Ducie using tested ground-based
methods to benefit three gadfly petrel species but, crucially,
not the Henderson Petrel which was not confirmed as a
nesting species on either atoll.
ACTIONS
Oeno and Ducie
The modest extent and flat accessible topography of the
atolls meant that the proposed eradication campaigns were
likely to be successful, given prior achievements elsewhere
(Towns & Broome, 2003). The eventual source of funding
was the UK’s Department for International Development
(DfID) whose interest lay principally in Pitcairn Island
and its people. For this reason, the programme linked
eradications on Oeno and Ducie, offering clear biodiversity
gains with limited risk of failure, to an eradication
attempt on Pitcairn where the risks of failure were higher
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 95–99. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
95
Island invasives: scaling up to meet the challenge. Ch 1B Rodents: Review
because of the rugged and heavily vegetated topography
and the complications associated with human presence.
Nonetheless the project proceeded in late 1997 with
approximately £100,000 of funding for Pitcairn and Oeno
from DfID and a further £20,000 for Ducie from the World
Wide Fund for Nature (Bell & Bell, 1998).
Success was duly achieved on Oeno and Ducie by
hand-laying of bait (baiting rate unspecified) on a 25 m
grid (Bell & Bell, 1998). The Oeno eradication has been
followed by growth of the population of the seabird species
most easily censused, Murphy’s petrel, at an annual rate of
6% (Brooke, et al., 2017). There are no post-eradication
census data from Ducie.
Pitcairn
Eradication was not achieved on Pitcairn in 1997.
There, preceding bait laying, the endeavour of cutting a
25 m grid of paths through the dense scrub cloaking the
island’s extremely severe terrain taxed the endurance of
the WMIL team, especially since, in the absence of prior
reconnaissance, the severity of the task ahead had not
been appreciated. Coverage of the cliffs was probably
incomplete. A lesson was learnt: future operations of this
magnitude must involve prior on-site reconnaissance by
key personnel.
The WMIL team departed shortly after the completion
of bait laying (overall baiting rate not specified), entrusting
the task of follow-up monitoring to the Pitcairn Islanders
(Bell & Bell, 1998). Given the many calls on the islanders’
time, and their lack of appropriate expertise, this strategy
was probably a mistake. With the benefit of hindsight, it
would have been better if extra costs had been incurred and
logistical difficulties overcome to allow some dedicated
team members to remain on Pitcairn to detect any residual
rat presence. While this change in protocol would not
have guaranteed a successful outcome, it could only have
increased the probability of success.
WMIL returned in 1998 to attempt to rectify the 1997
eradication failure. Unfortunately, the outcome reprised
that of 1997 despite more intensive monitoring after the
initial baiting, coupled with spot-laying of bait wherever
rat sign was detected (Bell, 1998).
A striking feature of these failures was not simply
the rapidity with which rats recovered to their prebait levels which, the reports of Pitcairners suggested,
happened within 18–24 months. There was also a universal
impression among the islanders and indeed myself on a
visit in 2000 that numbers overshot the status quo ante,
to a startling extent. For example, rats were frequently
encountered in homes, even in cooking ovens left ajar. A
possible explanation of this ‘overshoot’, that cannot be
confirmed by any formal existing trapping or density data,
is that, after the reduction in rat numbers due to baiting,
a large amount of food accumulated, for example on or
below Pitcairn’s abundant fruit trees. This surfeit possibly
nourished the extreme increase in rat numbers.
Henderson
Following the successful eradication of rats from
several large New Zealand islands using aerial baiting
techniques during the 1990s (Towns & Broome, 2003)
and from 113 km2 Campbell Island in 2001 (McClelland
& Tyree, 2002), the possibility of an eradication project
on Henderson Island using aerial baiting moved up the
agenda. A feasibility report delivered a favourable verdict,
subject to two caveats (Brooke & Towns, 2008). The
first was that, in the areas of high land crab (Coenobita
spp.) density behind Henderson’s beaches, it should be
demonstrated that sufficient bait could be scattered so that,
even after substantial bait removal by crabs, enough bait
remained to permit all rats to consume a fatal quantity.
The second concerned the endemic flightless Henderson
rail (Porzana atra). Given the recorded susceptibility of
rails to brodifacoum in cereal bait (Eason, et al., 2002) –
as would be used in a Henderson operation – there was a
need to demonstrate that Henderson rails could be caught
and then kept healthy in captivity. In the worst-case
scenario, the elimination of the wild population during the
eradication operation, the captives, once released after the
disappearance of bait, would become the founders of the
new wild population.
Both these issues were successfully addressed by
a field expedition in August/September 2009 (Brooke,
et al., 2010b; Cuthbert, et al., 2012), paving the way for
an eradication operation in 2011. The feasibility report
(Brooke & Towns 2008) suggested the late winter months
of September/October as the period of lowest food
availability and therefore the most suitable for bait-laying.
This suggestion was based on a 1-year study of plant
phenology (Brooke, et al., 1996), and drew on the fact that
Rattus exulans includes a proportion of vegetable material
in its diet. In the absence of any data whatsoever on the
intra-annual variation in the availability of invertebrates
and their contribution to the rats’ diet, this potential factor
Table 1 Summary table of rat eradication operations on the four Pitcairn Islands. Details from Bell & Bell (1998), Bell
(1998), Torr & Brown (2012) and E. Bell (pers. comm.).
Island
Type
Pitcairn
Volcanic Hand broadcast
Pitcairn
1998
April –
July
Oeno
Volcanic First two: hand
broadcast. Then
bait stations and
spot-laying
Atoll
Hand broadcast
1997
July –
August
Ducie
Henderson
Atoll
Hand broadcast
Makatea Aerial
1997
2011
November
August
96
Method
Year
Month(s)
baited baited
1997
June –
August
Bait type
Pestoff 20R; wax-covered
chocolate bait for 3rd
baiting
Pestoff 20R. Later
baitings supplemented by
wax-covered chocolate
bait
Pestoff 20R
Pestoff 20R
Pestoff 20R
No.
baitings
3
Successful?
No
3+
No
2
Yes
2
2
Yes
No
Brooke: Rat eradication in the Pitcairn Islands
could not be addressed in project planning. In the event,
late August 2011 became the provisional project date.
Fund-raising for the £1.5 million budget proceeded apace
under the aegis of the Royal Society for the Protection of
Birds (RSPB).
The operation was logistically complex involving
the 298-tonne Alaskan crab-fishing vessel, the Aquila,
sailing from the United States. Carrying two helicopters,
the Aquila undertook other rat eradications in the central
Pacific (Palmyra Atoll followed by Enderbury and Birnie
in the Phoenix Islands) before loading the 76 tonnes of bait
required for Henderson in Samoa. She then sailed east to
Henderson.
Meanwhile the rail-catching team were landed on
the island on 8 July 2011. The team immediately noticed
that fruit was more abundant than expected – of which
more anon. Catching of rails proceeded satisfactorily but
adapting birds to captivity proved more problematical
than in 2009, and 22 died before the solution was found,
enticing the birds to the food bowls with live bait such as
immobilised moths (Oppel, et al., 2016). In retrospect, it
appears that, by chance, the smaller 2009 batch of rails (26
caught: two died) simply included few birds reluctant to
adapt to captivity (Brooke, et al., 2010b; Brooke, et al.,
2012).
The losses meant that the number of captive rails, 75,
at the time of the Aquila’s arrival on 14 August, was lower
than the target of 100 birds, but not so much lower as to
cause a postponement or cancellation of baiting. The details
of bait spreading are covered in the report of the project
leaders (Torr & Brown, 2012). Overall the process went
remarkably smoothly, with bait buckets filled on board
the Aquila, obviating the need for any onshore storage of
bait. GPS mapping of the island, prior to the first bait drop,
revealed the area to be 43 km2, an enlargement over the
37 km2 that had been the basis for planning. Fortunately
there was sufficient contingency bait that this unexpected
expansion necessitated no adjustment of planned bait
densities.
Excluding enhanced bait application in the areas
of high crab density (Cuthbert, et al., 2012) and in the
coconut groves, the application rate was 10 kg/ha of pellets
(brodifacoum concentration of 20 ppm) over the majority
of the island for the first drop carried out between 15
and 17 August, and 6 kg/ha during the second bait drop
on 21 and 22 August. The 5-day interval between drops
was slightly less than originally planned because settled
weather prompted a decision to proceed immediately,
rather than delay until the planned interval of seven days
(Torr & Brown, 2012).
The immediate impact of the bait drop on the wild freeliving rails was dramatic – as it was on rats. Sixteen of 16
rails that were radio-tagged, and whose fate could therefore
be determined with certainty, died. However, mortality
island-wide was not total. The best estimate is that 93
percent of free-living rails died, leaving c. 500 survivors
(Oppel, et al., 2016). A few weeks after the drop, these
birds began breeding. Their numbers were supplemented in
October and November by the release of the captive birds,
and the population has since completely recovered (Oppel,
et al., 2016). Although, in the event, the captive birds were
not essential for the species’ persistence, the outcome was
in doubt in the anxious days after the bait drops, and there
is no question that a similar captive rail population must
be established, should there be another eradication attempt
in the future. This recommendation only gains force if, for
example, the bait drops occur over a longer time period, or
there are three drops instead of two. No other bird species
is known to have been adversely affected by the bait drops
on Henderson.
At the time the team caring for the captive rails left
Henderson in November, three months after the bait drop,
no signs of surviving rats had been noticed. Disastrously,
a surviving rat was seen and captured on video by a visitor
in March 2012. A follow-up visit, in May, confirmed
continuing rat presence and, as expected, rat numbers had
returned to ‘normal’ about two years later with no sign of
the overshoot noted on Pitcairn (Bond, et al., 2019).
The eradication failure immediately prompted a review
of the operation and a search for possible operational
errors. None has been discovered (Internal RSPB
documents). There were no apparent gaps in bait coverage,
and none of the batches of bait, deliberately retained for
post-operational testing, was shown to have incorrect toxin
loading. Such post-hoc testing cannot absolutely exclude
the remote possibility that some bags of bait did not have
toxic baits, a factory error. Finally, fieldwork on Henderson
in 2013 tested the rats, presumably animals descended
by several generations from the actual survivors, for
resistance to brodifacoum. No such resistance was found
(Churchyard, et al., 2015).
Genetic studies after failure excluded the possibility
that Pitcairn or other islands elsewhere in the Pacific had
been a source of rats that had somehow reached Henderson
and re-populated the island. In any case, knowledge of boat
traffic made this scenario extremely unlikely. Thus, there
had been a failure of eradication and not a re-introduction.
Because rat samples had been secured before the operation,
and were then obtained afterwards, it was possible to use
the change in microsatellite allele frequency to estimate
how many rats survived (Amos, et al., 2016). The answer
was about 80 individuals, very roughly one in a thousand
of the rats present on Henderson before the operation
(Brooke, et al., 2010b). It is a total compatible with the
absence of observations of living rats for around seven
months after the bait drops.
Can this total, neither indicating a tiny number of
survivors that might be ascribed to chance nor several
hundreds, even thousands, indicating serious deficiencies
in operational protocol, suggest improvements that might
be made for a second attempt?
Mention has already been made of the fact that the rail
team encountered more fruit than expected on Henderson
in July 2011. This was probably a delayed consequence
of a drought that afflicted Pitcairn, and presumably also
Henderson, from November 2010 to March 2011. When
this drought broke, it is likely that the trees became greener,
flowered and then fruited, at a time that was inopportune
for the rat eradication, especially if flowering and fruiting
were accompanied by increased numbers of invertebrates.
Although there has been one year-long study of the
leafing, flowering and fruiting phenology of Henderson’s
plants (Brooke, et al., 1996), this is clearly inadequate to
understand how plant phenological schedules may change
from year to year, and how they are altered by annual
variations in weather. That would require around 20 years
of study, an impossible task on isolated Henderson. Thus,
tailoring a rat eradication to a particular window of plant
food scarcity will always be difficult, if not impossible.
And no subsequent findings have altered the cautious
recommendations of the feasibility study (Brooke &
Towns, 2008), derived from the Brooke, et al. (1996) plant
phenology study, that September or a month either side is
the most suitable period.
Compounding this problem is that the operation must be
set in train – boats chartered, bait ordered and so forth – at
least six months before baiting (Parkes & Fisher, 2017). It
would, in theory, be possible to cancel an operation at a late
stage, for instance if there were reports of a surge in fruit
97
Island invasives: scaling up to meet the challenge. Ch 1B Rodents: Review
abundance, but the penalties for such a late cancellation
could well approach £500,000.
Following their helicopter flights across the island in
2011, the pilots reported, to universal surprise, a few tens of
coconut trees (Cocos nucifera) emerging from the canopy
growing on the raised atoll lagoon. Since the ground is
about 30 m above sea level, these trees must have involved
human intervention. They were certainly not planted by
members of the Sir Peter Scott Commemorative Expedition
of 1991–1992. There are two other known possibilities. The
first is that the Pitcairners who, during World War II, cut a
network of paths across the island, some several kilometres
from the coast, were responsible. Another possibility is that
the helicopter presence associated with the visit of the USS
Sunnyvale in 1966 provided an opportunity for coconuts to
be ‘bombed’ from overhead.
However the coconuts arrived, it is not surprising that
they have been growing unknown for decades since most
parts of this impenetrable island have remained unvisited
for centuries. The relevance of these observations is
that the research visit of 2013 (Churchyard, et al., 2015)
conducted captive trials to test which natural foods, if any,
were preferred by rats to bait pellets. Given a four-way
choice between coconut (removed from its shell), Myrsine
fruits, Pandanus nuts and Pestoff bait pellets, coconut
was preferred, with pellets second. Moreover 11 of 30
rats ate no pellets whatsoever in a 3-day trial (details in
Churchyard, et al., 2015). These findings were confirmed
by further similar research in 2015 that also indicated
the preference for natural food could not be overcome
by increasing the relative abundance of bait pellets, an
experimental adjustment equivalent to increasing the bait
application rate during helicopter operations (Lavers, et al.,
2016).
Although the coconut groves behind the North
and North-West Beaches received deliberately high
applications of bait pellets (Torr & Brown, 2012), this was
not the case for the unknown isolated trees in mid-island.
However, there are no data bearing on where on the island
the 80 surviving rats lived and whether their home ranges
were in the vicinity of coconuts.
It is evident that an absence of coconuts is not a sine qua
non of a successful rat eradication. Success was achieved
on Oeno (coconuts present) and Ducie (no coconuts).
Projects failed on Henderson and Pitcairn, both with
coconuts. More generally, numerous islands with coconuts
have been cleared of rats, including the island of Palmyra
(</www.fws.gov/refuges/news/PalmyraAtollRatFree.
html>) visited by the Aquila two months before it reached
Henderson.
Although Henderson’s coconuts could have contributed
to the project’s failure (Holmes, et al., 2015), removing this
possible cause would not be easy. Reaching every midisland coconut would require a helicopter to insert a small
group of “coconut destroyers” close to each tree, perhaps
via a winch. Their task would be to destroy all the nuts and
possibly the tree as well. That would still leave the coastal
coconuts. It is unlikely that their total destruction would
be countenanced by the Pitcairn Islanders and, in any case,
their flowers are a significant food of the endemic Stephen’s
lorikeet (Vini stepheni) (Trevelyan, 1995). Even destroying
or removing off-island all the fallen nuts, weighing several
tens of tonnes, would not be easy. But the practicalities
should be explored.
The discussion has reached the stage where the 2011
eradication appears to have failed, not because of any
operational blemishes and not because of any brodifacoumresistance but because a small number of rats failed to
consume a fatal dose, approximately one pellet, of bait.
98
Instead they chose to eat natural food in preference to bait
(Keitt, et al., 2015). This picture is entirely compatible
with the more general observation that tropical rodent
eradications are less likely to be successful than those on
temperate islands (Russell & Holmes, 2015)
If a second eradication attempt is to have an improved
chance of success, some aspects of the protocol may
have to change. The impracticalities of guaranteeing that
a bait drop occurs at a time of minimal food abundance
have already been discussed. The challenge of reducing
the availability of coconuts needs further thought. Finally,
I strongly advocate consideration of a further option, the
development of a more attractive bait formulation that will
entice even those rats that might have shunned the pellets
used in 2011 to eat bait. It will probably never be known
whether these crucial rats did not eat bait pellets because a
more palatable natural food was available, and/or whether
illness or pregnancy affected their appetite for novel foods
(neophobia). Altering the formulation of bait pellets by the
addition of such flavours as chocolate or peanut has already
been trialled by Orillion, the manufacturers of PestOff
pellets (Bill Simmons, pers. comm.). However, it remains
uncertain whether these changes would demonstrably
reduce the risk to an operation of such rat behaviours as
neophobia.
Although modest alteration of pellets may not engender
regulatory problems in UK Overseas Territories (Bill
Simmons, pers. comm.), the development of pellets of
enhanced attractiveness could pose technical problems.
For example, any additives must not make the pellets
more ‘sticky’ and liable to clog the hoppers underslung
from bait-distributing helicopters. But, optimistically, such
developments will occur as New Zealand develops the
expertise to rid itself of alien predators by 2050, as other
countries follow New Zealand’s lead, and as the relative
intractability of tropical islands is addressed.
Meanwhile, from my 25-year perspective, Henderson
will probably not change greatly in the next decade. A
patient approach will hugely increase the likelihood that
any second rat eradication attempt on Henderson is made
when the chances of success are demonstrably higher. It
will also avoid the mistake made on Pitcairn, of undertaking
an eradication project because money was available rather
than because a rational, even hard-nosed, assessment
confirmed that the chances of success and the biodiversity
gains of success outweighed the costs and risks of failure.
ACKNOWLEDGEMENTS
This personal account arises from eight separate
visits to the Pitcairn Islands, hugely facilitated by the Sir
Peter Scott Commemorative Expedition, the Foreign &
Commonwealth Office, the Royal Society for the Protection
of Birds, and the Pitcairn Islanders. Thanks to referees for
many helpful comments. Opinions expressed are my own.
REFERENCES
Amos, W., Nichols, H.J., Churchyard, T. and Brooke, M. de L. (2016).
‘Rat eradication comes within a whisker! A case study of a failed project
from the South Pacific’. Royal Society Open Science. DOI: 10.1098/
rsos.160110.
Bell, B.D. and Bell, E. (1998). Habitat restoration: Pitcairn Islands, South
Pacific. Eradication of rats and feral cats, April to December 1997.
Unpublished report. Wellington, New Zealand: Wildlife Management
International.
Bell, D. (1998). Habitat restoration: Pitcairn Islands, South Pacific.
Eradication of rats and feral cats, April to September 1998. Unpublished
report. Wellington, New Zealand: Wildlife Management International.
Benton, T.G. and Spencer T. (eds.) (1995). The Pitcairn Islands:
Biogeography, Ecology and Prehistory. London, U.K.: Linnean Society
and Academic Press.
Brooke: Rat eradication in the Pitcairn Islands
Bond, A.L., Cuthbert, R.J., McClelland, G.T.W., Churchyard, T., Duffield,
N., Havery, S., Kelly, J., Lavers, J.L., Proud, T., Torr, N., Vickery, J.A.,
and Oppel, S. (2019). ‘Recovery of introduced pacific rats following
a failed eradication attempt on Subtropical Henderson Island, South
Pacific Ocean’. In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell
and C.J. West (eds.) Island invasives: scaling up to meet the challenge,
pp. 167–174. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Brooke, M.de L. (1995). ‘The breeding biology of the gadfly petrels
Pterodroma spp. of the Pitcairn Islands: Characteristics, population
sizes and controls’. Biological Journal of the Linnean Society 56:
213–231.
Brooke, M. de L., Jones, P.J., Vickery, J.A. and Waldren, S. (1996).
‘Seasonal patterns of leaf growth and loss, flowering and fruiting on a
sub-tropical Central Pacific Island’. Biotropica 28: 164–179.
Brooke, M.de L. and Rowe, G. (1996). ‘Behavioural and molecular
evidence for specific status of dark and light morphs of the Herald petrel
Pterodroma heraldica’. Ibis 138: 420–432.
Russell, J.C. and Holmes N.D. (2015). ‘Tropical island conservation: Rat
eradication for species recovery’. Biological Conservation 185: 1–7.
Torr, N. and Brown, D. (2012). Henderson Island Restoration Project:
Post-operational Report. Unpublished report. Sandy, UK: Royal
Society for the Protection of Birds.
Towns, D.R. and Broome, K.G. (2003).’ From small Maria to massive
Campbell: Forty years of rat eradications from New Zealand islands’.
New Zealand Journal of Zoology 30: 377–398.
Trevelyan, R. (1995). ‘The feeding ecology of Stephen’s lory and nectar
availability in its food plants’. Biological Journal of the Linnean Society
56: 185–197.
Weisler, M.I. (1994). ‘The settlement of marginal Polynesia: New
evidence from Henderson Island’. Journal of Field Archaeology 24:
83–102.
Brooke, M. de L. and Towns, D.R. (2008). A Feasibility Study for
the Eradication of Kiore Rattus exulans from Henderson Island.
Unpublished report. Sandy, UK: Royal Society for the Protection of
Birds.
Brooke, M. de L., O’Connell, T.C., Wingate, D., Madeiros, D., Hilton,
G.M. and Ratcliffe, N. (2010a). ‘The potential for rat predation to cause
decline of the globally threatened Henderson petrel Pterodroma atrata:
Evidence from population modelling, the field and stable isotopes’.
Endangered Species Research 11: 47–59.
Brooke, M. de L., Cuthbert, R. J., Henricson, A., Torr, N., Warren, P. and
O’Keefe, S. (2010b). Towards Rat Eradication on Henderson Island:
Fieldwork Report August–September 2009. Unpublished report. Sandy,
UK: Royal Society for the Protection of Birds.
Brooke, M. de L., Harrison, G. and Cuthbert, R. J. (2012). The Capture,
Husbandry and Captive Breeding of Henderson Rails During the
Henderson Island Rat Eradication and the Impact of the Project on
Native Landbirds. Unpublished report. Sandy, UK: Royal Society for
the Protection of Birds.
Brooke, M. de L., Bonnaud, E., Dilley, B.J., Holmes, N.D., Jones, H.P.,
Provost, P., Rocamora, G., Ryan, P.G., Surman, C. and Buxton R.T.
(2017). ‘Seabird population changes following mammal eradications on
islands’. Animal Conservation 21: 3–12.
Churchyard, T., Proud, T., Brooke, M. de L., O’Keefe, S., Warren, P. and
Rodden, M. (2015). Henderson Island Trip Report: 30th June–31st
August 2013. Unpublished report. Sandy, UK: Royal Society for the
Protection of Birds.
Cuthbert, R.J., Brooke, M. de L. and Torr, N. (2012). ‘Overcoming hermit
crab interference during rodent baiting operations: A case study from
Henderson Island, South Pacific’. Wildlife Research 39: 70–77.
Eason, C.T., Murphy, E.C., Wright, G.R.G. and Spurr E.B. (2002).
‘Assessment of risks of brodifacoum to non-target birds and mammals
in New Zealand’. Ecotoxicology 11: 35–48.
Holmes, N.D., Griffiths, R., Pott, M., Alifano, A., Will, D., Wegmann, A.S.
and Russell, J.C. (2015). ‘Factors associated with rodent eradication
failure’. Biological Conservation, 185: 8–16.
Keitt, B., Griffiths, R., Boudjelas, S., Broome, K., Cranwell, S., Millett,
J., Pitt, W. and Samaniego-Herrera, A. (2015). ‘Best practice guidelines
for rat eradication on tropical islands’. Biological Conservation, 185:
17–26.
Lavers, J.L., McClelland, G.T.W., MacKinnon, L., Bond, A.L., Oppel, S.,
Donaldson, A.H., Duffield, N.D., Forrest, A.K., Havery, S.J, O’Keefe,
S., Skinner, A., Torr, N., and Warren, P. (2016). ‘Henderson Island
Expedition Report: May–November 2015’. RSPB Research Report 57.
Sandy, UK: Royal Society for the Protection of Birds.
McClelland, P. and Tyree, P. (2002). ‘Eradication – the clearance of
Campbell Island’. New Zealand Geographic. 58: 86–94.
Oppel, S., Bond, A.L., Brooke, M. de L., Harrison, G., Vickery, J.A.
and Cuthbert, R.J. (2016). ‘Temporary captive population and rapid
population recovery of an endemic flightless rail after a rodent
eradication operation using aerially distributed poison bait’. Biological
Conservation 204: 442–448.
Parkes, J. and Fisher, P. (2017). ‘Review of the Lehua Island Rat
Eradication Project 2009’. Pacific Cooperative Studies Unit Technical
Report 195. Honolulu, HI: University of Hawai’i at Mānoa, Department
of Botany.
Pitcairn Islands Scientific Expedition. (1992). Sir Peter Scott
Commemorative Expedition to the Pitcairn Islands, 1991–1992.
Unpublished expedition report. University of Cambridge, UK:
Department of Zoology.
99
K. Broome, D. Brown, K. Brown, E. Murphy, C. Birmingham, C. Golding, P. Corson, A. Cox and R. Griffiths
Broome, K.; D. Brown, K. Brown, E. Murphy, C. Birmingham, C. Golding, P. Corson, A. Cox and R. Griffiths.
House mice on islands: management and lessons from New Zealand
House mice on islands: management and lessons from New Zealand
K. Broome1, D. Brown2, K. Brown1, E. Murphy1, C. Birmingham1, C. Golding1, P. Corson1, A. Cox3 and R. Griffiths4
Department of Conservation, P.O. Box 10-420 Wellington New Zealand. <kbroome@doc.govt.nz>. 2Picton,
New Zealand. 3Ashburton, New Zealand. 4Island Conservation, Auckland, New Zealand.
1
Abstract The impacts of house mice (Mus musculus), one of four invasive rodent species in New Zealand, are only clearly
revealed on islands and fenced sanctuaries without rats and other invasive predators which suppress mouse populations,
influence their behaviour, and confound their impacts. When the sole invasive mammal on islands, mice can reach high
densities and influence ecosystems in similar ways to rats. Eradicating mice from islands is not as difficult as previously
thought, if best practice techniques developed and refined in New Zealand are applied in association with diligent planning
and implementation. Adopting this best practice approach has resulted in successful eradication of mice from several
islands in New Zealand and elsewhere including some of the largest ever targeted for mice; in multi-species eradications;
and where mouse populations were still expanding after recent invasion. Prevention of mice reaching rodent-free islands
remains an ongoing challenge as they are inveterate stowaways, potentially better swimmers than currently thought,
and prolific breeders in predator-free habitat. However, emergent mouse populations can be detected with conventional
surveillance tools and eradicated before becoming fully established if decisive action is taken early enough. The invasion
and eventual eradication of mice on Maud Island provides a case study to illustrate New Zealand-based lessons around
mouse biosecurity and eradication.
Keywords: biosecurity, eradication, impacts, invasive rodents, Maud Island
INTRODUCTION
The house mouse (Mus musculus) established in New
Zealand (NZ) around 1830, about 550 years after the first
rodent to arrive, the Pacific rat or ‘kiore’ (Rattus exulans),
60 years after Norway rats (R. norvegicus) and 30 years
before ship rats (R. rattus) (Atkinson, 1973). Mice in New
Zealand have traces of ancestry from three subspecies –
Mus musculus domesticus, M. m. castaneus and M. m.
musculus – however M. m. domesticus is the dominant
subspecies (King, et al., 2016; Veale, et al., 2018). The
hybridisation of subspecies could have occurred before or
after the mice arrived in NZ (Veale, et al., 2018).
In this paper, we explore three questions related to the
management of mice on islands for biodiversity protection:
Today mice are widespread and common throughout NZ
but not as common as ship rats. Mice increase in numbers
quickly in response to pulses of food and reductions in ship
rat abundance (Elliott & Kemp, 2016).
We use the invasion of Maud Island by mice in 2013
and their successful eradication in 2014 as a case study to
illustrate our lessons.
Rodent colonisations of smaller islands in the NZ
archipelago have different histories influenced by past
human visitation and proximity to the largest islands ‘North’
and ‘South’ considered ‘mainland’ by New Zealanders. Of
the 1065 islands >1 ha (excluding the mainland), mice
established on about 42 of them (Ruscoe & Murphy 2005;
Department of Conservation (DOC), unpublished data).
Action against mice for biodiversity protection goals
began with efforts by NZ Wildlife Service with rodentproof packaging of stores destined for rodent-free islands.
The first eradication of mice in NZ occurred in 1984 on
2 ha Whenuakura Island, although the project targeted
Norway rats, not mice (Veitch & Bell 1990).
In 1989 the first deliberate attempts to eradicate
mice from islands occurred on Mana 217 ha (Hook &
Todd, 1992), Rimariki 22 ha (Veitch & Bell, 1990),
and Allports 16 ha, (Brown, 1993). We can identify 36
attempts to remove mice from NZ islands larger than 1
ha, 28 of them succeeded and eight failed (Appendix 1).
Mice have reinvaded seven of the 28 from which they
were eradicated. Some of the eradication failures could
possibly be attributed to reinvasion. These figures update
NZ data presented by MacKay, et al., (2007) and Howald,
et al. (2007) who included eradication attempts worldwide
where the eradication of mice was not always a stated
goal and where the presence of mice on the island prior to
eradication remained unproven.
1. What do we know about the impacts of mice on NZ
island ecosystems?
2. What have we learnt about eradicating mice from
islands and what do we now consider best practice
in NZ?
3. What have we learnt about preventing mice from
establishing new populations on NZ islands?
IMPACTS OF MICE
Mice often inhabit islands with other invasive species
which can confound efforts to quantify mice impacts.
Predators, particularly rats, can have a marked influence on
the behaviour and densities of mice while simultaneously
reducing and masking mice impacts (Bridgman, 2012).
Removal of mice in these situations often requires
simultaneous removal of other invasive mammals, thereby
continuing the confusion over how to attribute recovery to
the absence of mice and not the other species involved.
On islands where mice are the only invasive mammal
present they usually attain higher densities, exhibit different
behaviours and therefore have more conspicuous impacts
on native biodiversity (Angel, et al., 2009).
Mice as bird predators
Mice eat small bird’s eggs. Frogley (2013) filmed them
eating quail (Coturnix japonica) (30 × 24 mm), zebra finch
(Taeniopygia guttata) (14 × 9 mm) and canary (Serinus
canaria) eggs (16 × 11 mm) from unattended used nests
placed on the forest floor. Fewer of the quail eggs tested
were eaten, suggesting they are near the size limit for mice
to break into. Over 400 hours of filming six natural forest
bird nests in podocarp-broadleaved forest at Maungatautari
resulted in observation of only a single mouse visit (Watts,
et al., 2017).
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
100
up to meet the challenge, pp. 100–107. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Broome, et al.: Mice – management and lessons from NZ
Smaller seabirds such as some storm petrel species
appear more vulnerable to egg and sometimes chick
predation by mice although some studies suggest this
has little effect on productivity (Campos & Granadeiro,
1999). Shore plover (Thinornis novaeseelandia) breed
very successfully on Waikawa Island with mice at high
densities. There is no evidence of egg predation on shore
plover (egg size 37 × 26 mm) or white-faced storm petrels
(Pelagodroma marina) (egg size 36 × 26 mm) on Waikawa
Island in the presence of high mouse numbers (H. Jonas &
J. Dowding pers. comm.).
The evidence for other impacts on birds in NZ is more
circumstantial, for example differences in abundance of
snipe (Coenocorypha aucklandica) and black-bellied
storm petrels (Fregetta tropica) on Antipodes Island with
mice and rodent-free islands such as Adams and Bollons
(Miskelly, et al., 2006; Imber, et al., 2005).
Mice as reptile predators
On Mana Island removal of grazing livestock led to an
increased mouse population due to improved habitat from
rank grass. The McGregor’s skink (Oligosoma macgregori)
population declined and mice were seen eating skinks in
pitfall monitoring traps. Following the eradication of mice
in 1989 McGregor’s skink numbers increased and they
became more conspicuous (Newman, 1994).
Norbury, et al. (2014) followed the fate of translocated
Otago skinks (Oligosoma otagense) in a fenced site which
contained mice as the only mammalian predator. They
observed mice attacking 25 cm adult skinks but noted skink
survival rates were adequate for population persistence.
Romijn (2013) compared the capture rates of ornate
skinks (Oligosoma ornata) between sites with and without
mice present (without other predators). The site with mice
had periodic control of mice to maintain densities below 21
per100 trap-nights. He found population increases at both
sites but significantly higher rates in the site with no mice.
Mice were implicated in the suppression of recruitment
in a shore skink (O. smithi) population at Tawharanui
fenced sanctuary (Wedding, 2007).
Mice as invertebrate predators
Invertebrates are an important part of the broad
diet of mice (Ruscoe & Murphy, 2005). St Clair (2011)
compiled the known impacts of invasive rodents on island
invertebrates including a range of NZ species influenced
by mice.
Watts, et al. (2017) conducted a large-scale treatment
switch experiment at Maungatautari in 2011–2016. Two
fenced enclosures in forest had all mammalian pests
removed except mice. At one site they eradicated mice and
at the other allowed mice to increase. Results suggested
mice suppressed beetles, spiders, earthworms and weta in
both abundance and size.
Mice impacts on vegetation
Williams, et al. (2000) found mice destroy all seed
they eat, rather than acting as seed dispersers. On the New
Zealand mainland, seed predation by mice may affect
regeneration of kauri (Agathis australis) (Badan, 1986),
pingao (Desmoschoenus spiralis) and sand tussock (Poa
triodioides) (Miller & Webb, 2001). Mouse predation
on mountain beech (Fuscospora cliffortioides) and rimu
(Dacrydium cupressinum) seeds not only reduces rates of
seedling establishment, but may also alter the composition
of forests over time (Wilson, et al., 2007). Seed predation
by mice may also impede ecological restoration efforts, for
example inhibiting a tree planting programme on Mana
Island (Hook & Todd, 1992).
Watts, et al. (2017) found no significant impact of mice
on forest seedling establishment over their five-year study.
However, they noted their (predator fenced) mainland
study site has been subject to modification by a range of
introduced mammals for hundreds of years prior to the
beginning of the study.
Other biodiversity impacts by mice
Two studies reported observations of mice eating the
eggs of a NZ native fish, inanga (Galaxias maculatus)
(Baker, 2006; Hickford, et al., 2010).
Besides the direct impacts discussed above, mice also
influence other predators who use them as a food source.
For example, stoats (Mustela erminea) will include mice
in their diet. In beech (Fuscospora spp.) dominated forest,
mast seeding events lead to high populations of mice
followed by increased stoat populations with consequent
impacts on native species (King & Murphy, 2005).
Mice may also provide an important year-round food
resource for larger predators on islands with strongly
seasonal primary food resources such as colonial nesting
seabirds. They may therefore ‘artificially’ sustain higher
predator populations through the non-seabird nesting
periods.
MOUSE ERADICATION
Since 1989 developments in mouse eradication
methodologies in New Zealand mirrored those of rat
eradications (Towns & Broome, 2003; Broome, 2009;
Russell & Broome, 2016). Aerial broadcast baiting was
consistently chosen for eradications targeting mice on
islands larger than 40ha (Appendix 1).
Mouse susceptibility to brodifacoum is highly variable.
For example, Cuthbert, et al. (2011) had two Gough Island
mice survive doses of 2.44 and 5.41 mg/kg, respectively.
These individuals were subsequently offered more bait in
no-choice tests and died after ingesting 12.2 and 7.14 mg/
kg. Three (of 10) mice from Lord Howe Island survived
doses of 5.2 mg/kg in a no-choice bait test (D. Priddel
pers. comm.). A subsequent trial using 30 wild-caught
Lord Howe mice allowed to feed ad libitum for three days
resulted in 100% mortality (A. Walsh pers. comm.).
Mice usually die from about five days following the first
application. For example, MacKay, et al. (2007) found no
sign of surviving mice on Adele Island eight days after bait
application. However, they can survive much longer (see
case study) and in one laboratory trial, a warfarin-resistant
mouse survived a total of 65 days after first feeding on
brodifacoum laced bait (Rowe & Bradfield, 1976).
Bridgman (2012) studied the behaviour of mice in
the presence of ship rats. She found ship rats strongly
influenced the movements of mice, reducing home ranges
and nutrition levels. This has implications for eradication
projects targeting both rats and mice, reinforcing the need
for comprehensive bait coverage and well-spaced multiple
bait applications to allow for the dominant rats to die off
and theoretically ‘free up’ the movement of any mice
remaining.
Some projects failed to eradicate mice because they
did not explicitly target them. For example, on Mokoia
Island in 1989 an eradication project targeting Norway
rats using bait stations spaced at 50 × 50 m subsequently
found mice on the island (P. Jansen, pers. comm.). Because
the eradication was designed around the home range of
Norway rats, mice survived and became detectable after
the rat population had crashed.
101
Island invasives: scaling up to meet the challenge. Ch 1B Rodents: Review
Eradications of mice on islands in NZ progressed
through the 1990s with mixed success (MacKay, et
al., 2007). The review of mouse eradication projects
by MacKay, et al., in 2007 could not find a consistent
operational factor contributing to eradication failure but
recommended robust planning of future projects to rule out
operational errors, thereby providing better insight into the
cause of failures.
Following this recommendation, a project to eradicate
mice from three islands (Adele, Tonga, Fisherman)
in Tasman Bay in 2007 strictly adhered to the current
agreed best practice methodology for mouse eradications
(Golding, 2010). The Island Eradication Advisory
Group (IEAG), a technical advisory group of the NZ
Department of Conservation, updates and maintains a
document providing technical advice to project managers
in the planning, implementation and monitoring of rat
eradications on islands (Broome, et al., 2017a).
The IEAG consider best practice for mouse eradications
to be similar to that used for rats with the following changes:
Bait applications use 50% overlap on both the first
and second application (cf. for rats where 50% overlap
is recommended for the first application and 25% for the
second) (Fig. 1).
Bucket flow rates remain at or above 4 kg/ha (cf. for
rats where bucket flow rates of 3 kg/ha are permissible).
With 50% overlaps as in 1 above, this means applying a
minimum of 8 kg/ha on the ground in each application.
The interval between applications is extended to a
minimum of 14 days (cf. for rats where more flexibility in
timing of the second application is permissible).
The IEAG has recently developed a best practice
document incorporating these elements with other advice
borrowed from the rat best practice (Broome, et al.,
2017b). Since the Tasman Bay project, all subsequent
mouse eradications following this advice have succeeded.
including one of the largest (Macquarie 12,800 ha); multispecies eradications (Macquarie, and Rangitoto/Motutapu
3,809 ha) and a still-establishing mouse population (Maud
309 ha – see case study).
Changes 1 and 2 recognise the smaller territories of
mice than rats and strive to ensure all mice encounter bait.
Relatively few mouse home range studies have occurred
on NZ islands (Ruscoe & Murphy, 2005). MacKay, et al.
(2011) measured home ranges varying from 0.15–0.48 ha
on Saddle Island. Radio-tracking found animals living in
areas with dense shrub and grass cover had smaller ranges
and mean nightly movements than those living in areas with
tall canopy and minimal ground cover. Elsewhere on the
NZ mainland in the absence of other mammalian predators
and competitors, Goldwater, et al. (2012) estimated
densities of 160 mice/ha in rank kikuyu grass (Pennisetum
clandestinum) immediately after other mammals were
eradicated, but density has since greatly declined.
Eradication designs must cater for not only the smallest
home range (rather than the mean) but also the smallest
foraging movements by mice over the limited period that
bait is available in palatable condition. At 8 kg/ha the 2 g
baits used in NZ would in theory be on the ground at 0.4
baits/m2 providing ample opportunity for mice to encounter
baits, especially after a second application.
Keeping bucket flow rates relatively high (possum
control operations using the same equipment routinely use
rates around 1 kg/ha), reduces the risk of interruptions in
bait flow out of the bucket. Such interruptions in flow are
potentially fatal to eradication success as they would not
be mapped by the helicopter’s GPS navigation recording
system, and therefore could go unnoticed.
Change 3 acknowledges mice as light and erratic
feeders compared to rats (Clapperton, 2006). Extending the
period of bait availability, compared to a rat eradication,
is desirable to ensure all mice have access to lethal doses
before bait is consumed by other fauna or environmentally
degraded. Brown (1993) found mice initially reluctant to
take bait presented in bait stations on Allports and Motutapu
Islands. They often ‘sampled’ small portions of baits over
several nights before full-scale consumption ensued. He
described a gradual spread of consumption from a focal
point, speculating that social interactions between mice
encouraged more to try the new food resource presented.
To counter the risk of mice being present but undetected
in the presence of rats, some projects have deliberately
designed their baiting strategy to mice eradication
standard. For example, the rodent eradication (ship rats
and kiore) on Great Mercury Island was designed to mouse
eradication best practice standards despite no confirmed
evidence of mice. The island operated as a pastoral farm
with minimal biosecurity precautions for over 50 years
so it was difficult to believe mice had not arrived during
this time. The project sponsors found it cost effective risk
management to assume mice were present and design the
project accordingly (Corson & Hawkins, 2016).
MOUSE BIOSECURITY
Keeping islands free of mice presents ongoing
challenges in quarantine, surveillance and responding to
arrivals. Pathways for invasion include cargo and personal
luggage landed on the island, vessels and aircraft of all
sizes, and swimming or rafting to islands.
Fig. 1 50% overlap when aerially sowing bait. Arrows
indicate centres and direction of two consecutive sowing
lines. The dark shaded area shows the area of overlap
between the first and the (half-completed) second line.
102
Vulnerabilities to these pathways differ between
islands but some islands may also be less susceptible to
establishment of a mouse population following incursion.
For example, Secretary, Kapiti, Stewart, Raoul and
Campbell Islands have records of mice arriving, without
evidence of meaningful action to respond, and yet failing
to subsequently establish populations (DOC unpublished
data). At the time all of these large islands had rats or stoats
present, potentially providing a form of biological defence
against mouse establishment. Weka (Gallirallus australis)
may also play a role where they occur on islands. For
example, on rat-free Tarakaipa Island mice were barely
detectable in the presence of weka (DB pers. obs.). Weka
held in captivity eagerly attacked mice entering their pen
(CG pers. obs.). Conversely, the subsequent eradication of
Broome, et al.: Mice – management and lessons from NZ
such predators could, in theory, increase the vulnerability
of the island to invasion by mice. Further research into this
phenomenon is warranted.
on the rodent-free island. They readily detected mice by
both tracking tunnels and wax tags, even during the initial
phases of the invasion.
The probability of establishment can relate to propagule
pressure (Lockwood, et al., 2005). Because rodent
populations fluctuate seasonally in NZ with peaks in late
summer, the risk of invasion could increase at this time of
year. Additionally, mast seeding events in some forests can
produce superabundant populations of mice which increase
propagule pressure on nearby islands. For example, mice
were successfully eradicated from Adele Island in 2007
and a biosecurity system installed. In the 2014/15 summer
a significant mast seeding event occurred in the adjacent
Abel Tasman National Park where mice became abundant.
In February 2015 they were discovered on Adele. Attempts
to eliminate them by localised trapping around points of
detection failed and a population re-established (CG pers.
obs.).
Invading mice can move large distances. For example,
pairs of mice sequentially released at opposite ends of
Saddle Island (approximately 400 m apart), increased their
nightly movements two-fold, and range sizes ten-fold,
relative to movements on this island prior to the mouse
eradication. This allowed them to rapidly and reliably
encounter each member of the opposite sex (MacKay,
2011).
Mice as stowaways
Mice are inveterate stowaways with numerous records
of their discovery in cargo destined for islands. The DOC
invasion incidents database has 24 records of mice reaching
islands amongst cargo between January 2010 and June
2017. Two more were intercepted on vessels en route to
pest-free islands. Mice have been discovered in visitor day
packs, in kayaks and nesting in under-seat dingy flotation.
Container, building and vessel openings must be <6 mm to
restrict mouse access. Of equal importance is the vigilance
required to ensure doors, lids and hatches remain closed
when not in use.
Quarantine measures to prevent mice reaching islands
require constant vigilance by people involved. Careful
checking of cargo, using rodent-proof containers for
transport and control measures on board vessels are key
components. These precautions can be enhanced by
good rodent management and habitat control at ports and
minimising the quantity of equipment transferred to islands
(e.g. by having field equipment remain on-island).
Mice swimming to islands
Mice are often thought of as poor swimmers relative
to rats (Russell & Clout, 2005). However, Evans, et al.
(1978) found mice would readily enter water and swim. A
fisherman saw a number of mice 600 m from shore in Lake
Monowai while night fishing during the 2009 mouse plague
(CG pers. comm..). Fishermen anecdotally report them in
trout guts (James & Fox, 2017) and they have been found
live in coastal flood debris (DB pers. obs.). The maximum
distance over water that mice can cross unassisted remains
unknown and therefore the pathway should not be assumed
unimportant when considering biosecurity risks for an
island.
Pomona and Rona Islands in Lake Manapouri were
both assumed a ‘safe’ distance offshore (500 m and 600
m respectively) but both were reinvaded by mice within a
decade of successful eradication, probably by swimming or
rafting on flood debris. These re-invasions coincided with
beech masting events when mice reached high abundance
on the mainland.
Detection methods
We can readily detect mice at low densities, in the
absence of other rodent species, using a range of tools
including footprint tracking tunnels, chew cards and other
bait interference methods, snap traps and trained detection
dogs. Nathan, et al. (2013) studied mouse detection on
Saddle Island (6 ha) during an experimental invasion
event in which a male and a female mouse were released
A mouse invading pest-free Moturua Island initially
tracked inked footprint tracking cards in October 2011
and was finally trapped in late 2011. On one occasion this
animal travelled at least 750 m between tracking tunnels
over a 36-hour period (KB unpublished data).
Mice established on islands in relatively high numbers
can hinder the detection of newly invading rats by
‘swamping’ detection tools. For example, they cover ink
tracking cards on Waikawa Island within a few nights
which can obscure the footprints of an invading rat. Mice
usually do not trigger DOC200 stoat and rat traps but steal
the bait, rendering the trap less attractive. These mouseinduced limitations delayed the detection of a Norway
rat incursion on Waikawa Island in 2012, indicated by a
dramatic decline in the critically endangered NZ shore
plover. The rat was never caught and only retrospectively
identified with the help of a rodent detection dog by the
discovery of a nest containing bird remains and Norway rat
fur and droppings (EM unpublished data).
Incursion response
Responding to the discovery of invading mice on a pest
free island is challenging due to the potential delay between
incursion and discovery through periodic surveillance
checks. Nathan, et al. (2015) demonstrated the urgency
of responding to a mouse invasion by experimentally
releasing one male and one female mouse on Saddle
Island. They subsequently bred and the mouse population
reached the island’s carrying capacity within five months.
Routine surveillance discovered invading mice on Adele
Island in February 2015, potentially months after arrival.
Despite intensive trapping around points of detection the
incipient population could not be eliminated.
CASE STUDY MAUD ISLAND
Biosecurity
Before 2013 rodents had never established on Maud
Island (309 ha) in the Marlborough Sounds. Consequently,
it has some highly rodent-vulnerable native species
including some not found elsewhere, such as the Maud
Island frog (Leiopelma pakeka), and others restricted to a
handful of nearby pest-free islands.
Keeping pests from reaching Maud has long been a
priority. Landing is restricted and DOC staff are present
year-round. Stoats are considered the biggest invasive
threat because they can swim the 900 m from the mainland
and have done so on at least three occasions. Traps targeting
stoats and rats are throughout the island and checked
regularly. A quarantine store at the mainland DOC ranger
station is used to check cargo destined for Maud or other
pest free islands. Extra precautions are taken to prevent
chytrid fungus – a pathogen implicated in the worldwide
decline of frog populations (Berger, et al., 1999) – from
reaching Maud.
In 2006, a mouse was killed by the Maud Island
resident ranger when turning garden compost. An incursion
103
Island invasives: scaling up to meet the challenge. Ch 1B Rodents: Review
response using mouse traps and a trained rodent detection
dog failed to find further sign of mice after several weeks.
with a Pestoff 20R pellet as used for the aerial baiting,
indicating no aversion to the bait.
In October 2013, a mouse was captured in visitor
accommodation on the island. An incursion response
immediately deployed traps, detection devices and a
rodent detection dog. Several mice were trapped around
the buildings. The dog handler reported mice in several
places across the island. Breeding was confirmed from
necropsied animals. The youngest mice were in age class
1 (0–1 months in age) and the eldest in age class 6 (8–10
months) suggesting the first invaders arrived about a year
previously and they had bred through the winter, which is
uncommon in NZ.
Testing of all four trapped mice revealed brodifacoum
liver residues in three of them of 4.65–8.82 mg/kg.
Considering liver values probably resulted from higher
doses due to losses through excretion and metabolism
(Eason & Wickstrom, 2001), these mice probably received
many times the published LD50 for mice of 0.52 mg/
kg (O’Connor & Booth, 2001). Maggots from the more
decomposed male caught 22 September contained 2.35 mg/
kg. DNA testing found these mice to be clearly from the
original Maud invasion, not a new independent invasion.
DNA analyses found the Maud Island population
highly inbred, suggesting the population arose from a
single incursion. Although the mice were a genetic subset
of the mainland population, their point of origin could not
be established (E.M. & R. Fewster, unpublished data).
With an emerging picture of an established mouse
population across the island, the incursion response team
were forced to admit their efforts had begun too late and a
whole island eradication was required.
To understand how mice had reached the island and
remained undiscovered for long enough to establish,
an independent review of biosecurity procedures was
undertaken (Kennedy & Chappell, 2013). This found
several weaknesses, including a lack of devices capable of
killing or detecting mice on the island or on the ranger’s
boat, that was pulled onto a slipway on the island when
not in use. The focus on stoats and rats allowed mice to go
unnoticed. Some staff regularly visiting the island bypassed
quarantine standards.
Extensive monitoring over the subsequent two years no
further survivors but a further incursion in 2018 has once
again established a mouse population on the island. Mouse
trapping on the island after bait application was intended as
indicative monitoring only and had limited coverage of the
island. We assume other mice survived in un-trapped areas
long after bait application. These animals presumably
acquired a lethal dose of brodifacoum and died without
reproducing.
The successful eradication of an expanding population
of mice from Maud is an indication of high bait acceptance
despite other natural food being available in relative
abundance. Camera footage from some of the buildings on
Maud showed mice taking large quantities of bait placed
in trays during the eradication and presumably caching it
(CB pers. obs.).
CONCLUSION
The review could not identify the pathway for the
mouse incursion but made many recommendations for
improvement which were actioned prior to the eradication.
The island’s biosecurity plan has recently been re-written
to capture these new practices and give more authority to
biosecurity rangers to enforce standards.
Mice remain on many large islands in New Zealand and
around the world. The techniques used in NZ to eradicate
mice have been successful and could readily be applied
to other temperate islands of similar size with a good
chance of success. Biosecurity measures to protect islands
from mouse invasion are challenging and mice must be
considered a real threat to all rodent free islands, regardless
of previous invasion history.
Eradication
Biosecurity lessons:
In 2014, mouse eradication best practice was
successfully applied to eradicating the newly established
population of mice on Maud Island. Challenges included
the abundance of natural food available to the expanding
mouse population, and the presence of residential buildings
requiring careful management of domestic foodstuffs and
waste to minimise access to alternative food after toxic
baiting.
Quarantine standards must apply to everyone to be
effective. The pre-eminence of biosecurity over other duties
of island staff and managers needs regular reinforcement
to create an organisational culture which can sustain high
biosecurity standards over time.
A helicopter applied 8 kg/ha on 23 July 2014 followed
by 8 kg/ha 23 days later (15 August) with strict adherence
to the current agreed best practice described above. Two
mice were trapped on Maud on 19 August, 27 days after
the first bait application. Both had bait in their stomachs.
A badly decayed male mouse was taken from a snap trap
on 22 September and a female trapped the next day. This
sexually mature female showed no signs of past or present
breeding and appears to have survived about 60 and 37 days
after the first and second bait applications, respectively. An
intensive trapping grid (10 m × 10 m) was installed around
each capture site covering about one hectare. No further
mice were caught.
We estimated the age (from tooth eruption and wear) of
the last mouse caught to be five months, meaning it could
have lived through all bait applications. Bait was freely
available from July to October, so these individuals must
have encountered it. Although a range of trap baits were
used, the snap traps which caught each mouse were baited
104
All potential threats and all potential pathways need to
be assessed and multiple layers of protection established:
i.e. quarantine checking, pest proof containerisation,
hygiene of transportation, targeted surveillance, capability
and readiness for incursion response.
Independent review of procedures can give valuable
insights into opportunities for improvements and should be
done proactively and routinely.
The risk of successful mouse invasions may be
influenced by island predators (or lack thereof) and mouse
abundance at potential source populations.
Eradication lessons:
The current agreed best practice used in NZ has a very
good track record of success (>90% in known outcomes)
against mice on temperate islands. This is far better than
previously published review figures which did not present
data on the quality of planning and delivery or discriminate
between operations deliberately targeting mice and those
targeting other species where mice also occur.
Broome, et al.: Mice – management and lessons from NZ
Mice can take a long time to succumb to the cumulative
effects of small doses of brodifacoum and some individuals
may require significantly higher doses than others. A
baiting strategy which prolongs the availability of toxicant
to mice has a better chance of success. In NZ this is usually
achieved with two well-spaced bait applications but a third
application is also an option.
Bait application rates need to allow for other bait
consumers when multiple target species are involved and
must not fall below the ability of sowing equipment to
spread bait 100% reliably.
Where the presence of mice is likely but unproven due
to suppression by other species, it is prudent to design the
eradication assuming their presence, rather than discover
that they have survived a rat eradication and thrived in the
absence of rats or other predators.
Eradication is feasible against newly established and
expanding populations of invading mice, especially if
current agreed best practice is followed.
REFERENCES
Atkinson, I.A.E. (1973). ‘Spread of the ship rat (Rattus r. rattus L.) in
New Zealand’. Journal of the Royal Society of New Zealand 3: 457–472.
Angel, A. Wanless, R. and Cooper, J. (2009). ‘Review of impacts of the
introduced house mouse on islands in the Southern Ocean: are mice
equivalent to rats?’ Biological Invasions 11: 1743–1754.
Badan, D. (1986). ‘Diet of the house mouse (Mus musculus L.) in two
pine and a kauri forest’. NZ Journal of Ecology 9: 137–141.
Baker, C.F. (2006). ‘Predation of inanga (Galaxias maculatus) eggs
by field mice (Mus musculus)’. Journal of the Royal Society of New
Zealand 36: 143–147.
Bancroft, B. (2004). Patiti (Banded) Island rodent eradication 2004.
Rotorua, New Zealand: Waimangu Volcanic Valley Ltd and Department
of Conservation.
Berger, L. Speare, R. and Hyatt A. (1999). ‘Chytrid Fungi and Amphibian
Declines: Overview, Implications and Future Directions’. In: Campbell,
A. (ed.) Declines and Disappearances of Australian Frogs, pp. 23–33.
Canberra, ACT: Environment Australia.
Bowie, M. Kavermann M. and Ross, J. (2011). ‘The Quail Island Story
– Thirteen Years of Multi-species Pest Control: Successes, Failures
and Lessons Learnt’. In: C.R. Veitch, M.N. Clout and D.R. Towns
(eds.) Island invasives: eradication and management, pp. 157–161.
Occassional Paper SSC no. 42. Gland, Switzerland: IUCN and
Auckland, New Zealand: CBB.
Bridgman, L.J. (2012). ‘Interactions between ship rats and house mice’.
PhD thesis. Hamilton, New Zealand: University of Waikato. Retrieved
from <http://hdl.handle.net/10289/6396>.
Broome, K. (2009). ‘Beyond Kapiti – a decade of invasive rodent
eradications from New Zealand islands’. Biodiversity 10: 14–24.
Broome, K., Golding, C., Brown, K., Corson, P. and Bell, P. (2017a).
Rat Eradication Using Aerial Baiting: Current Agreed Best Practice
Used in New Zealand (Version 3.1). Wellington, NZ: New Zealand
Department of Conservation internal document DOC-29396.
Broome, K., Golding, C., Brown, K., Horn, S., Corson, P. and Bell, P.
(2017b). Mouse Eradication Using Aerial Baiting: Current Agreed Best
Practice Used in New Zealand (Version 1.0). Wellington, NZ: New
Zealand Department of Conservation internal document DOC-3034281.
Brown, D. (1993). ‘Eradication of mice from Allports and Motutapu
Islands’. Ecological Management 1: 19–30. Wellington, NZ:
Department of Conservation.
Brown, D. (2013). Coal Island (Te Puka-Hereka) Restoration Plan.
Department of Conservation Te Anau. <http://www.doc.govt.nz/
Documents/conservation/land-and-freshwater/offshore-islands/coalisland-restoration-plan.pdf>. DOCDM-12333366.
Campos, A.R. and Granadeiro, J.P. (1999). ‘Breeding biology of the
white-faced storm-petrel on Selvagem Grande Island, north-east
Atlantic’. Waterbirds: The International Journal of Waterbird Biology
22: 199–206.
Chappell, R. (2008). Eradication of Ship Rat, Kiore, House Mouse and
Rabbit from Ohinau and Motutapere Islands – Final Report. Department
of Conservation internal report DOC-5449724.
Clapperton, B.K. (2006). ‘A Review of the Current Knowledge of Rodent
Behaviour in Relation to Control Devices’. Science for Conservation
263. Wellington, NZ: Department of Conservation. 55 p.
Clout, M.N. and Russell, J.C. (2006). ‘The Eradication of Mammals from
New Zealand Islands’. In: F. Koike, M.N. Clout, M. Kawamichi, M. De
Poorter and K. Iwatsuki (eds.) Assessment and Control of Biological
Invasion Risks, pp.127–141. Occasional Paper SSC no. 62. Gland,
Switzerland: IUCN.
Corson, P. and Hawkins, K. (2016). Post Operational Report for the
Eradication of Kiore, Ship Rats and Cats from Ahuahu – Great Mercury
Island, New Zealand. Department of Conservation internal report
DOCDM-1477863.
Cuthbert, R., Visser, P, Louw, H. and Ryan, P. (2011). ‘Palatability and
efficacy of rodent baits for eradicating house mice (Mus musculus) from
Gough Island, Tristan da Cunha’. Wildlife Research 38(3): 196–203.
Department of Conservation (2011). Indian Island Rodent Eradication.
Department of Conservation internal report DOCDM-752616.
Eason C and Wickstrom, M. (2001). Vertebrate Pesticide Toxicology
Manual (Poisons). Wellington, NZ: Department of Conservation.
Elliott, G. and Kemp, J. (2016). ‘Large scale pest control in New Zealand
beech forests’. Ecological Management and Restoration 17(3): 200–
209.
Evans, R.L., Katz, E.M., Olson, N.L. and Dewsbury, D.A. (1978). ‘A
comparative study of swimming behaviour in eight species of muroid
rodents’. Bulletin of the Psychonomic Society 11: 168–170.
Fraser, Island, Wilcken, J. and Buley, K. (2013). Rotoroa Wildlife
Management and Translocation Plan 2013-2038. Version 1.4.
Auckland, NZ: Auckland Zoo.
Frogley, K. (2013). ‘Invasive Mice and their Effects on Birds: Do
Mice (Mus musculus) Within Fenced Sanctuaries Prey on Bird Eggs?
University of Otago Wildlife Management Report 270. 30 p.
Glassey, M. (2006). ‘A Historical and Ecological Survey of Hauturu
Island, Whangamata’. BAppSc thesis. Auckland University of
Technology / Bay of Plenty Polytechnic, Tauranga.
Golding, C. (2010). ‘House mouse (Mus musculus) eradication by aerial
bait application on Adele, Tonga and Fisherman Islands, Abel Tasman
National Park, New Zealand’. Conservation Evidence 7: 62–68.
Goldwater, N., Perry, G.L.W. and Clout, M.N. (2012). ‘Responses of
house mice to the removal mammalian predators and competitors’.
Austral Ecology 37: 971–979.
Griffiths, R., Buchanan, F., Broome, K., Neilsen, J., Brown, D., Weakley,
M. (2015). ‘Successful eradication of invasive vertebrates on Rangitoto
and Motutapu Islands’. Biological Invasions 17: 1355–1369.
Hickford, M.J., Cagnon, M. and Schiel, D.R. (2010). ‘Predation, vegetation
and habitat-specific survival of terrestrial eggs of a diadromous fish,
Galaxias maculatus (Jenyns, 1842)’. Journal of Experimental Marine
Biology and Ecology 385: 66–72.
Hook, T. and Todd, P. (1992). ‘Mouse Eradication on Mana Island’. In:
C.R. Veitch, M. Fitzgerald, J. Innes and E. Murphy (eds.) National
Predator Management Workshop, pp. 33. Threatened Species Occasional
Publication 3, Wellington, NZ: Department of Conservation.
Horn, S. Russell J. and Simmons, B. (these proceedings). Antipodes
Island mouse eradication.
Howald, G., Donlan, C.J., Galvan, J.B., Russell, J.C., Parkes, J.,
Samaniego, A., Wang, Y., Veitch, D., Genovesi, P., Pascal, M., Saunders,
A. and Tershy, B. (2007). ‘Invasive rodent eradications on islands’.
Conservation Biology 21: 1258–1268.
Imber, M.J., Bell, B.D. and Bell E.A. (2005) ‘Antipodes Islands birds in
autumn 2001’. Notornis 52:125–132.
James, G. and Fox, S. (2017). ‘Mouse Eating Trout’. <http://www.
bestofnzflyfishing.com/articles/mouseeatingtrout.html>. Accessed June
2017.
Kennedy, E. and Chappell, R. (2013). Audit of DOC’s Quarantine and
Surveillance Arrangements for the Islands of the Marlborough Sounds.
November 2013. Christchurch, NZ: Department of Conservation
internal report DOC-1318761.
King, C., Alexander A., Chubb T., Cursons R., MacKay J., McCormick
H., Murphy E., Veale A. and Zhang H. (2016). ‘What can the
geographic distribution of mtDNA haplotypes tell us about the invasion
of New Zealand by house mice Mus musculus?’ Biological Invasions
18 (6):1551–1565.
King, C. and Murphy, E. (2005). ‘Stoat’. In: C. King (ed.) The Handbook
of New Zealand mammals, pp. 261–287. Oxford: Oxford University
Press.
Lockwood, J.L., Cassey, P. and Blackburn, T. (2005). ‘The role of
propagule pressure in explaining species invasions’. Trends in Ecology
and Evolution 20(5): 223–228.
MacKay, J.W.B., Russell, J.C. and Murphy, E.C. (2007). ‘Eradicating
house mice from islands: successes, failures and the way forward’. In:
G.W. Witmer, W.C. Pitt and K.A. Fagerstone (eds.) Managing Vertebrate
Invasive Species: Proceedings of an International Symposium, pp. 294–
304. Fort Collins, Colorado, USA: USDA/APHIS Wildlife Services,
National Wildlife Research Center.
MacKay, J.W.B. (2011). ‘Improving the Success of Mouse Eradication
Attempts on Islands’. PhD thesis, Auckland, NZ: University of
Auckland.
MacKay, J.W.B., Murphy, E.C., Anderson, S.H., Russell, J.C., Hauber,
M.E., Wilson, D.J. and Clout, M.N. (2011). ‘A Successful Mouse
Eradication Explained by Site-specific Population Data’. In: C.R.
Veitch, M.N. Clout and D.R. Towns (eds.) Island invasives: eradication
and management, pp. 198–203. Occassional Paper SSC no. 42. Gland,
Switzerland: IUCN and Auckland, New Zealand: CBB.
McKinlay, B. (1999). ‘Eradication of Mice from Mou Waho, Lake
Wanaka’. Ecological Management 7: 1–5. Wellington, NZ: Department
of Conservation.
Miller, A.P. and Webb, P.I. (2001). ‘Diet of house mice (Mus musculus L.)
on coastal sand dunes, Otago, New Zealand’. New Zealand Journal of
Zoology 28: 49–55.
Miskelly, C.M., Walker, K.J. and Elliott, G.P. (2006). ‘Breeding ecology
of three subantarctic snipes’. Notornis 53: 361–374.
Nathan, H.W., Clout, M.N., Murphy, E.C. and MacKay, J.W.B. (2013).
‘Strategies for detection of house mice (Mus musculus) on a recently
invaded island’. New Zealand Journal of Ecology 37(1): 26–32.
105
106
Veitch & Bell, 1990; Newman, 1985
Hook & Todd, 1992; Newman, 1994
Brown, 1993
Brown, 1993
MacKay, et al., 2007
Veitch & Bell, 1990
Clout & Russell, 2006
Torr, 2002
Glassey, 2006
Clout & Russell, 2006
Clout & Russell, 2006
Secondary
baiting method
NA
Aerial broadcast
NA
NA
NA
NA
NA
NA
Hand broadcast
NA
Bait station
Primary baiting
method
Bait station
Bait station
Bait station
Bait station
Bait station
Bait station
Bait station
Aerial broadcast
Bait station
Bait station
Aerial broadcast
Eradication
status
Successful
Successful
Successful
Successful
Failed
Successful
Successful
Successful
F or R
Failed
Successful
Eradication
start date
1983
1989
1989
1989
1989
1989
1992
1993
1993
1993
1994
DOC best
practice
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Whenuakura
Mana
Allports
Motutapu
Mokoia
Rimariki
Moturemu
Enderby
Hauturu
Te Haupa
Motutapere
Area
(ha)
2
217
16
2
136
22
5
710
10
6
46
Island
Coromandel
Cook Strait
Marlborough
Marlborough
Lake Rotorua
Kaituna Bay
Kaipara Harbour
Subantarctic
Coromandel
Hauraki Gulf
Coromandel
Mice
targeted
Y
Y
Y
Y
N
Y
Y
N
N
Y
Y
Region
Appendix 1 Mouse eradications on NZ islands >1ha. F or R = failed or reinvaded and S (R) = successful but subsequently reinvaded.
Nathan, H.W., Clout, M.N., MacKay, J.W.B., Murphy, E.C. and Russell,
J.C. (2015). ‘Experimental island invasion of house mice (Mus
musculus)’. Population Ecology 57(2): 363–371.
Newman, D.G. (1985). The Apparent Loss of the Whenuakura Island
Tuatara Population, Whangamata Islands Wildlife Sanctuary.
Wellington, NZ: Department of Conservation internal report, file WIL
35/2/13.
Newman, D.G. (1994). ‘Effects of a mouse, Mus musculus, eradication
programme and habitat change on lizard populations of Mana Island,
New Zealand, with special reference to McGregor’s skink, Cyclodina
macgregori’. New Zealand Journal of Zoology 21: 443–456.
Norbury, G., van den Munckhof, M., Neitzel, S., Hutcheon, A., Reardon,
J. and Ludwig, K. (2014). ‘Impacts of invasive house mice on postrelease survival of translocated lizards’. New Zealand Journal of
Ecology 38: 322–327.
O’Connor, C.E. and Booth, L.H. (2001). ‘Palatability of rodent baits
to wild house mice’. Science for Conservation 184. Wellington, NZ:
Department of Conservation. 11 p.
Owen, K. (1998). ‘Removal and Reintroduction of North Island Weka
(Gallirallus australis greyi) to Mokoia Island as a Result of a Talon 7/20
Cereal Based Aerial Poison Drop’. Ecological Management 6: 41–47.
Wellington, NZ: Department of Conservation.
Ritchie, J. (2000). Matakohe – Limestone Island Scenic Reserve
Restoration Plan. Auckland, NZ: Natural Logic Environmental
Management Ltd.
Romijn R.L. (2013). ‘Can Skinks Recover in the Presence of Mice? BSc
(Hons) thesis. Wellington, NZ: Victoria University of Wellington.
Rowe, F.P. and Bradfield, A. (1976). ‘Trials of the anticoagulant
rodenticide WBA 8119 against confined colonies of warfarin-resistant
house mice (Mus musculus)’. Journal of Hygiene 77: 427–431.
Ruscoe, W.A. and Murphy, E.C. (2005). ‘The House Mouse’. In: C.M.
King (ed.) The Handbook of New Zealand Mammals. Oxford: Oxford
University Press.
Russell, J.C., and Clout, M.N. (2005). ‘Rodent Incursions on New Zealand
Islands’. In: J. Parkes, M. Statham and G. Edwards (eds.) Proceedings
of the 13th Australian Vertebrate Pest Conference, pp. 324–330. Lincoln,
New Zealand: Landcare Research.
Russell, J.C. and Broome, K.G. (2016) ‘Fifty years of rodent eradications
in New Zealand: another decade of advances’. New Zealand Journal of
Ecology 40: 197–204.
Shaw, V., Torr, N. (2011). ‘Eradicating mammal pests from Pomona and
Rona Islands in Lake Manapouri, New Zealand: a focus on rodents’.
In: C.R. Veitch, M.N. Clout and D.R. Towns (eds.) Island invasives:
eradication and management, pp. 356–360. Occassional Paper SSC no.
42. Gland, Switzerland: IUCN and Auckland, New Zealand: CBB.
St Clair, J.J.H. (2011). ‘The impacts of invasive rodents on island
invertebrates.’ Biological Conservation 144: 68–81.
Torr, N. (2002). ‘Eradication of rabbits and mice from Subantarctic
Enderby and Rose Islands’. In: C.R. Veitch and M.N. Clout (eds.)
Turning the tide: the eradication of invasive species, pp. 319–328.
Occasional Paper SSC no. 28. IUCN SSC Invasive Species Specialist
Group, IUCN, Gland, Switzerland and Cambridge, UK.
Towns D.R. and Broome, K.G. (2003). ‘From small Maria to massive
Campbell: forty years of rat eradications from New Zealand islands’.
NZ Journal of Zoology 30: 377–398.
Veale, A.J., Russell, J.C., King, C.M. (2018). The genomic ancestry,
landscape genetics, and invasion history of introduced mice in New
Zealand. Royal Society Open Science 5: 170879. <http://dx.doi.
org/10.1098/rsos.170879>.
Veitch, C.R. and Bell, B.D. (1990). ‘Eradication of Introduced Animals
from the Islands of New Zealand’. In: D.R. Towns, C.H. Daugherty and
I.A.E. Atkinson (eds.) Ecological Restoration of New Zealand Islands,
pp. 137–146. Wellington, NZ: Department of Conservation.
Veitch, C.R. (2002a). ‘Eradication of Norway rats (Rattus norvegicus)
and house mice (Mus musculus) from Browns Island, New Zealand’.
In: C.R. Veitch and M.N. Clout (eds.) Turning the tide: the eradication
of invasive species, pp. 350–352. Occasional Paper SSC no. 28. IUCN
SSC Invasive Species Specialist Group, IUCN, Gland, Switzerland and
Cambridge, UK.
Veitch, C.R. (2002b). ‘Eradication of Norway rats (Rattus norvegicus)
and house mice (Mus musculus) from Motuihe Island, New Zealand’.
In: C.R. Veitch and M.N. Clout (eds.) Turning the tide: the eradication
of invasive species, pp. 353–356. Occasional Paper SSC no. 28. IUCN
SSC Invasive Species Specialist Group, IUCN, Gland, Switzerland and
Cambridge, UK.
Watts, C., Innes, J., Wilson, D., Fitzgerald, N., Bartlam, S., Thornburrow,
D., Smale, M. and Barker, G. (2017). Impacts of Mice Alone on
Biodiversity: Final Report of a Waikato Field Trial. Landcare Research
Contract Report LC2747. Prepared for Waikato Regional Council. 33 p.
Wedding, C. (2007). ‘Aspects of the Impacts of Mouse (Mus musculus)
Control on Skinks in Auckland, New Zealand’. MSc thesis. Auckland,
NZ: Massey University.
Williams, P.A., Karl, B.J., Bannister, P. and Lee, W.G. (2000). ‘Small
mammals as potential seed dispersers in New Zealand’. Austral Ecology
25: 523–532.
Wilson, D.J., Wright E.F., Canham C.D. and Ruscoe W.A. (2007).
‘Neighbourhood analyses of tree seed predation by introduced rodents
in a New Zealand temperate rainforest’. Ecography 30: 105–119.
References
Island invasives: scaling up to meet the challenge. Ch 1B Rodents: Review
Hauraki Gulf
Lake Wanaka
Whangarei Harbour
Lake Rotorua
Hauraki Gulf
Whangarei Harbour
Whangarei Harbour
Lake Rotorua
Whangarei Harbour
Canterbury
Lake Rotomahana
Marlborough
Coromandel
Marlborough
Lake Manapouri
Lake Manapouri
Tasman Bay
Tasman Bay
Tasman Bay
Hauraki Gulf
Fiordland
Hauraki Gulf
Hauraki Gulf
Canterbury
Hauraki Gulf
Dusky Sound
Hauraki Gulf
Marlborough
Subantarctic
Region
Mice
targeted
Y
Y
Y
Y
Y
Y
Y
Y
Y
N
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Browns
Mou Waho
Matakohe
Mokoia
Motuihe
Matakohe
Matakohe
Mokoia
Matakohe
Quail
Patiti
Blumine
Ohinau
Pickersgill
Rona
Pomona
Adele
Fisherman
Tonga
Te Haupa
Coal
Motutapu
Rangitoto
Quail
Te Haupa
Indian
Rotoroa
Maud
Antipodes
Island
Area
(ha)
60
140
37
136
179
37
37
136
37
85
13
377
46
96
60
262
88
4
8
6
1,163
1,509
2,311
85
6
167
140
309
2,012
DOC best
practice
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Y
Y
Y
Y
Y
N
Y
Y
Y
Y
N
Y
Y
Y
Y
Eradication
start date
1995
1996
1996
1996
1997
1997
1998
2001
2001
2002
2004
2005
2005
2005
2007
2007
2007
2007
2007
2008
2008
2009
2009
2009
2010
2010
2013
2014
2016
Eradication
status
Successful
Successful
F or R
F or R
Successful
F or R
F or R
S (R)
F or R
Failed
Failed
Successful
Successful
Successful
S (R)
S (R)
S (R)
S (R)
S (R)
Successful
Successful
Successful
Successful
F or R
Successful
Successful
Successful
Successful
Successful
Primary baiting
method
Aerial broadcast
Aerial broadcast
Aerial broadcast
Aerial broadcast
Aerial broadcast
Aerial broadcast
Aerial broadcast
Aerial broadcast
Bait station
Bait station
Bait station
Aerial broadcast
Aerial broadcast
Aerial broadcast
Aerial broadcast
Aerial broadcast
Aerial broadcast
Aerial broadcast
Aerial broadcast
Bait station
Aerial broadcast
Aerial broadcast
Aerial broadcast
Aerial broadcast
Trapping
Aerial broadcast
Aerial broadcast
Aerial broadcast
Aerial broadcast
Secondary
baiting method
NA
NA
NA
Hand broadcast
NA
NA
NA
Hand broadcast
NA
Hand broadcast
NA
NA
NA
NA
NA
NA
NA
NA
NA
Hand broadcast
NA
Bait station
Bait station
Hand broadcast
Bait station
NA
NA
NA
NA
Veitch, 2002a
McKinlay, 1999
MacKay, et al., 2007
Clout & Russell, 2006; Owen, 1998
Veitch, 2002b
MacKay, et al., 2007; Ritchie, 2000
Clout & Russell, 2006; Ritchie, 2000
MacKay, et al., 2007
Clout & Russell, 2006
Bowie, et al., 2011
Bancroft, 2004
MacKay, et al., 2007
Chappell, 2008
MacKay, et al., 2007
Shaw & Torr, 2011
Shaw & Torr, 2011
Golding, 2010
Golding, 2010
Golding, 2010
MacKay, et al., 2011
Brown, 2013
Griffiths, et al., 2015
Griffiths, et al., 2015
Bowie, et al., 2011
MacKay, et al., 2011
Department of Conservation, 2011
Fraser, et al., 2013
This paper
Horn, et al. (these proceedings)
References
Appendix 1 (continued) Mouse eradications on NZ islands >1ha. F or R = failed or reinvaded and S (R) = successful but subsequently reinvaded.
Broome, et al.: Mice management and lessons from NZ
107
P.W. Carey
Carey, P.W. Simultaneous rat, mouse and rabbit eradication on Bense and Little Bense Islands, Falkland Islands
Simultaneous rat, mouse and rabbit eradication on Bense and
Little Bense Islands, Falkland Islands
P.W. Carey
SubAntarctic Foundation for Ecosystems Research, 8 Estuary Road, Christchurch, New Zealand.
<peter@subantarctic.com>.
Abstract Bense and Little Bense Islands (144 ha total area) have, for over a century, supported populations of three
introduced pest mammals: Norway rat (Rattus norvegicus), house mouse (Mus musculus), and European rabbit
(Oryctolagus cuniculus). An operation to eradicate these mammals simultaneously was undertaken in winter 2016. Cereal
pellets laced with brodifacoum (25 ppm) were hand-broadcast on both islands in two applications with 3,900 kg of bait
applied in total. Baiting transects were spaced at 20 m intervals and bait-throwing positions located every 20 m along
each transect. The coastline was also baited at 20 m intervals. Precision bait coverage was aided by programming GPS
units to give off an audible alarm when staff reached each correct bait-throwing position. Application 1 resulted in an
average bait density of 15.3 kg/ha. Application 2 commenced 10 days later and resulted in an average baiting density of
11.7 kg/ha. Reduced availability of field staff resulted in coverage in Application 2 being less complete than in Application
1 and only the most important mammal habitats were baited a second time. These were: all tussock areas, all coastlines,
and some inland heath areas. Areas with no vegetation (e.g. burned zone on Bense) and some inland heath communities
were not treated, although all of these retained unconsumed bait from Application 1. Some non-target mortality was
recorded, with dolphin gulls (Larus scoresbii) being the most common victims. This was also the only species observed to
consume bait pellets. Consumption of poisoned mammals or gulls may have killed three turkey vultures (Cathartes aura
jota), one striated caracara (Phalcoboenus australis), and one short-eared owl (Asio flammeus). The removal of invasive
species is part of a broader ecological restoration plan for these islands and will hopefully lead to an increase in native
biodiversity, including the re-establishment of the endemic passerines Cobb’s wren (Troglodytes cobbi) and blackish
cinclodes (Cinclodes antarcticus).
Keywords: ecological restoration, invasive species
INTRODUCTION
Like the natural biodiversity on most islands, the native
plants and animals of the Falkland Islands are vulnerable
to catastrophic impacts when non-native mammals are
introduced (Tabak, et al., 2014; Carey, 2015). Prior to
the arrival of humans, the Falklands had only one species
of terrestrial mammal – the Falklands fox, or warrah
(Dusicyon antarcticus). While people quickly hunted
this sole native mammal to extinction by 1876, they also
introduced a further nine alien species which have since
established feral populations. These are: Norway rat
(Rattus norvegicus), black rat (R. rattus), house mouse
(Mus musculus), European rabbit (Oryctolagus cuniculus),
eastern cottontail rabbit (Sylvilagus sp.), brown hare
(Lepus capensis), Patagonian grey fox (Dusicyon griseus),
domestic cat (Felis catus), and guanaco (Lama guanicoe),
as well as domestic dogs (Canis lupus), poultry, and
livestock (Strange, 1992; Woods & Woods, 2006). These
invasive species have had negative impacts on the native
birds (Tabak, et al., 2015) and invertebrates (St Clair, 2011)
through direct predation and competition for food.
first to attempt the simultaneous removal of three species:
R. norvegicus, M. musculus, and O. cuniculus. These were
the only introduced mammal species on the Bense islands.
The Falkland Islands are located in the south-west
Atlantic Ocean, approximately 500 km east of Argentina.
Spanning 51°–53° S and 57°–62° W, there are 778 islands
in the archipelago (FITB, 2016). Eleven islands are
permanently inhabited, although only the two largest of
these are home to more than one family. The Falklands
are unique among subantarctic islands in that much of
the land is privately owned, and conservation-minded
landowners have been at the forefront of environmental
work in the islands (for example Strange, 2007; Poncet,
et al., 2011). Invasive species eradications began in 2001
with the removal of Norway rats from two small islands
(Brown, et al., 2001). Rats have since been successfully
cleared from a further 66 islands, while the Patagonian
grey fox was eradicated from one island in 2008 (Poncet, et
al., 2011; FIG, 2015). The project covered here is the first
Falklands attempt to eradicate mice and rabbits, and the
Site description
Although conservation gains can be made by eradicating
a single mammal species where more than one invasive
species is present (Helmstedt, et al., 2016), eradication
attempts that simultaneously target all invasive species
are desirable when logistically and financially feasible.
Simultaneous multi-species eradications can avoid
magnifying the problems caused by one pest species when
another is removed. On subantarctic Macquarie Island, the
removal of cats prior to the eradication of rabbits may have
contributed to a population increase of the latter species,
which in turn exacerbated grazing pressure on plants and
soil erosion (Bergstrom, et al., 2009; but see Springer
(2016) for a discussion of the role of rabbit population
fluctuations).
METHODS
Bense (c. 107 ha) and Little Bense (c. 37 ha) Islands are
found in Port North, in the north-west Falkland Islands at
51°29’S 61° 31’W. These two islands have been home to
Norway rats, house mice, and European rabbits for more
than 100 years. Rabbits were deliberately introduced by
whalers whereas rats and mice arrived as stowaways on
vessels anchored in the nearby harbour or used to move
livestock (R. Napier pers. comm.). The islands are joined
by a rocky reef, exposed at low tide, and therefore were
treated as a single island for eradication purposes (Fig. 1).
The vegetation is broadly similar across the two islands,
with at least 20 species of vascular plants recorded (Table
1). The coastal zone is maritime tussock formation, with
lush stands of tussock grass (Poa flabellata) growing to
3 m in height. The interior is low-growing oceanic heath
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
108
up to meet the challenge, pp. 108–113. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Carey: Mice & rabbits Bense & Little Bense, Falkland Islands
formation, dominated by diddle-dee (Empetrum rubrum)
(Moore, 1968). Bense Island has had greater grazing
pressure with horses, cattle, and sheep wintering on the
island at various times during the 20th century. These same
species were also placed on Little Bense but would quickly
migrate to Bense Island as Little Bense has no water on
it. (W. Goodwin, pers. comm.) This may explain why
palatable species such as boxwood (Hebe elliptica) are
more prevalent on Little Bense, and why there are also
greater expanses of dense tussock on the smaller island.
Both islands have been free of livestock since 1985.
Also in 1985, a fire burned about 20% of Bense Island.
The scorched area remains an unvegetated barren zone
of peat and ash, with loose peat creeping downwind and
smothering some areas of unburned vegetation.
Fig. 1 Bense Island (bottom left), Little Bense Island (top)
and West Falkland Island (bottom right). Note the dark,
burned area along much of the east coast of Bense. At
low tide, Bense and Little Bense Islands are connected
by a rocky reef. Bense is 750 m away from West Falkland
at its closest point.
The western coast of Bense Island has vertical cliffs up
to c. 25 m in height. The terrain gradually tilts lower as
one moves east, with gentle cobble or sand beaches found
on the east coast. Little Bense is lower (c. 18 m maximum
height) with a coastline of sloping rocks in the west and
north, and sand beaches in the east and south. At its closest
point, the mainland of West Falkland Island is 750 m away
from Bense Island.
Despite the presence of invasive mammals, the avifauna
of these islands is not completely extirpated (Table 1)
Table 1 Plants and birds commonly found on Bense and Little Bense Islands.
Birds
Magellanic penguin
Rock shag
Imperial shag
Black-crowned night
heron
Upland goose
Kelp goose
Ruddy-headed goose
Falklands steamer duck
Crested duck
Turkey vulture
Variable hawk
Striated caracara
Magellanic oystercatcher
Blackish oystercatcher
Two-banded plover
Magellanic snipe
Plants
Spheniscus magellanicus
Plalacrocorax magellanicus
Phalacrocorax atriceps albiventer
Nycticorax nycticorax
falklandicus
Chloephaga picta
Chloephaga hybrida
Chloephaga rubidiceps
Tachyeres brachypterus
Lophonetta specularioides
Cathartes aura jota
Geranoaetus polyosoma
Phalcoboenus australis
Haematopus leucopodus
Haematopus ater
Charadrius falklandicus
Gallinago paraguaiae
magellanica
Brown skua
Catharacta antarctica
Dolphin gull
Larus scoresbii
Kelp gull
Larus dominicanus
South American tern
Sterna hirundinacea
Dark-faced ground tyrant Muscisaxicola maclovianus
Grass wren
Cistothorus platensis
Falklands thrush
Turdus falcklandii
White-bridled finch
Melanodera melanodera
Long-tailed meadowlark Leistes loyca
Black-chinned siskin
Spinus barbatus
Tussock grass
Couch grass
Common bent grass
Hair grass
Poa flabellata
Agropyron pubiflorum
(magellanicum)
Agrostis tenuis
Aira sp.
Small fern
Chickweed
Wavy hair grass
Diddle-dee
Tufted fescue grass
Cudweed
Pig vine
Native boxwood
Mountain berry
Meadow grass
Sheep’s sorrel
Sea cabbage
Blechnum penna-marina
Cerastium arvense
Deschampsia flexuosa
Empetrem rubrum
Festuca cirrosa (erecta)
Gamochaeta nivalis
Gunnera magellanica
Hebe elliptica
Pernettya pumila
Poa sp.
Rumex acetosella
Senecio candicans
Procumbent pearlwort
Groundsel
Christmas bush
Wood rush
Sagina procumbens
Senecio vulgaris
Baccharis magellanica
Luzula alopecurus
109
Island invasives: scaling up to meet the challenge. Ch 1B Rodents: Review
and the islands were listed within the Falklands as a top
priority for mammal eradication (Miller, 2008). While 26
land and sea bird species were commonly found on the
islands, conspicuously absent were the only Falklands
endemic passerines: Cobb’s wren (Troglodytes cobbi)
and blackish cinclodes (Cinclodes antarcticus). Neither
of these species breeds on islands with rats (Tabak, et al.,
2016). The islands are also bereft of burrowing seabirds
such as sooty shearwater (Puffinus griseus) and thin-billed
prion (Pachyptila belcheri), both of which breed on nearby
rat-free islands (Woods & Woods, 1997).
Bense and Little Bense have never had a resident human
population, but because they were a desirable site for
wintering livestock, for much of the 20th century they were
occasionally home to shepherds and farmhands for a few
days at a time. A small shanty, built on Bense in 1926, was
the only building found on either island until 2002, when
a second shanty was built next to the original structure.
All farming ceased in 1996, when Bense and Little Bense
Islands (along with neighbouring Cliff Island and Bradley
Islet) were purchased by the SubAntarctic Foundation for
Ecosystems Research (SAFER) with a goal to restore the
islands’ ecology and improve them as wildlife habitat.
Index trapping
Index trapping to ascertain habitat preferences and
relative abundance of rodents was conducted on Bense
Island over eight visits, spanning 10 years and most
seasons (i.e. November 2004, October 2006, July 2007,
August 2007, September 2008, March 2010, January
2013, January 2014). Trap lines followed the methods
described in Cunningham and Moors (1996), using Victor
Easy Set wooden snap-traps (Woodstream Corp., Lititz,
Pennsylvania, USA), with an interval of 25 m between
trapping stations. A trap which caught an animal or which
was sprung with no catch, was deemed to have been
effective for half the night, and was therefore counted as
0.5 of an effective trap-night. Trap lines were placed in
two different habitats: coastal tussock formation (1,612.5
effective trap-nights), or inland heath communities (1,077
effective trap-nights).
Eradication operation
Following basic ecological studies, including surveys
of birds and invertebrates, an operation to eradicate rats,
mice, and rabbits was undertaken in winter 2016. For
bait distribution, local field staff were hired in Stanley,
the Falklands capital. None had previous experience
with hand-baiting so training was provided the day prior
to the beginning of operations. The operation ran from 8
August to 3 September and was timed to coincide with the
period when natural food on the islands is most scarce.
Cereal pellets laced with brodifacoum at 25 ppm (25-W
Conservation Pellets, manufactured by Bell Laboratories)
were hand-broadcast along parallel transects in two
applications, with an interval of 10 days between them.
A baiting map of the islands, comprising a series
of parallel transects spaced at 20 m intervals laid over a
high-resolution satellite photo, was created using QGIS
software. Along each of these transects, baiting points were
located every 20 m (Fig. 2). This resulted in an imaginary
grid with 20m squares across both islands. Baiting points
were also created at 20 m intervals along the coastlines
of both islands, following the natural contours of the
shoreline. Map data were loaded onto handheld GPS units
(Garmin GPSMAP64) with an audible alarm set to sound
whenever the unit reached a baiting point. Field personnel
could then navigate to a desired transect line and follow it
exactly, with the alarm telling them when they had reached
a baiting point. GPS units were accurate to around 2 m.
110
Fig. 2 Detail of the baiting map of Bense Island. Each white
or purple dot represents a baiting point. Baiting points
are 20 m apart.
The walking tracks of field staff were monitored using GPS
tracking and were checked each night against a base map.
Any areas not covered properly were thus identified, and
targeted for remedial attention the following day.
At each baiting point, five full scoops of bait were
flung in five different directions as per hand broadcast
best-practice (Broome, et al., 2011). Thus, coverage at
each baiting point overlapped with bait thrown from
neighbouring baiting points. Bait pellets were thrown with
plastic scoops cut to hold 100 g when full. Staff carried the
pellets in 20-litre plastic buckets, which could hold about
15 kg of bait. Rubber gloves, Tyvek coveralls, and dust
masks were available to all field personnel.
Bait was transported to the islands from Stanley. It first
went by barge to a protected bay on West Falkland Island,
and from there it was moved to Bense and Little Bense
in loads slung under a Chinook helicopter. The helicopter
deposited the bait in six depots across the approximate
midline of Bense Island, and at one location in the centre
of Little Bense Island. A total of 4,400 kg of bait was
delivered to the islands for this operation.
At the end of the operation, seven bait stations were
established along the north-eastern coast of Bense Island,
in areas thought to be the most likely zone of landfall for
any rats that might swim from West Falkland Island. Bait
was placed inside lengths of polyethylene pipe, 15 cm in
diameter. Wax baits (containing 0.0005% w/w difenacoum
and 0.001% w/w denatonium benzoate) were wired to the
inside of the pipe and a handful of brodifacoum cereal
pellets were also added. Bait stations were secured to the
ground with wire staples and rocks.
Carey: Mice & rabbits Bense & Little Bense, Falkland Islands
Post-eradication monitoring
The islands were re-visited briefly in December 2016
(three months post-baiting) and in November 2017 (14
months post-baiting) to search for survivors of the baiting
operation. During the latter visit, two hundred chewsticks
(PCR Wax Tag, Pest Control Research) with peanut butterflavoured wax attractants were installed in all coastal areas
and in vegetated interior zones, with preference given to
those areas known to be good rodent habitat. Chewsticks
were checked for bite marks from rabbits and rodents before
departure (up to 14 days after installation) and were left in
place to be checked on subsequent visits to the islands. Staff
actively searched for tracks, fresh droppings, and other
signs of mammals throughout the visit. Daytime searches
for rabbits were made by a dedicated hunter, including
extensive observations by binoculars from a camouflaged
position on high ground and by careful downwind stalking
through areas known to be favoured by rabbits. A thermal
camera (Thermapp) was used to replicate these searches at
night without the use of lights that could frighten rabbits.
Weather during eradication operation
Temperatures ranged from -3° C to +7° C, with
moderate to strong winds on all days. Snow and sleet
showers frequently swept the islands but accumulation
was slight and short-lived. No precipitation fell as rain.
Weather did not prevent baiting except for one half-day
during Application 1 and one full-day during Application
2, when wind speeds were too high to cast bait effectively.
RESULTS
Index trapping
Index trapping showed rats were much more prevalent
in coastal tussock areas, with 82 rats caught there from
1,612.5 effective trap nights, whereas on inland heath
areas, only three rats were recorded from 1,077 effective
trap nights. Mice were more evenly distributed between the
two habitats sampled, with 50 caught in coastal tussock,
and 32 caught in inland heath.
Rabbits were not targeted with snap traps but individuals
were observed on most parts of Bense, except the denuded
burn-zone. Rabbits were not thought to be present on Little
Bense until a single animal was observed there in February
2015. This was the only time in 18 visits that a rabbit was
seen on Little Bense, suggesting that if there was a resident
population on the island, it was likely much smaller than
that on Bense.
Effectiveness and coverage of Application 1
For the first application (8–16 August), a team of
five field staff covered Bense and Little Bense with bait,
resulting in a mean density of 15.3 kg/ha. However, bait
was more densely applied along the shoreline and in dense
tussock, while it was applied less densely in the burn zone,
which is devoid of vegetation. All cliffs were baited along
their top edges and on all lower ledges that were safely
accessible. Where safe access was not possible, pellets
were thrown from above. Along accessible shorelines,
particular attention was paid to the beach margin where
vegetation began and to areas just above the high tide line
where debris had accumulated.
Although Little Bense is only a third the size of Bense,
baiting there proved to be much more challenging due to
the extremely dense tussock grass and the fragmented,
convoluted northern coast. Overland access to the many
coastal chasms and rock slabs was particularly difficult
since it required climbing through or over the worst of the
tussock (over 2 m high). To apply bait to this northern coast
more efficiently, a small boat was used. In some chasms
the boat could be used as a mobile baiting platform, with
pellets broadcast into the tussock from the deck. In other
areas, personnel were landed to climb to the vegetated
margin, then re-boarded and moved to the next position.
Effectiveness and coverage of Application 2
For the second application (26 August–3 September), a
team of three field staff attempted to duplicate the coverage
achieved in Application 1. However, due to the smaller
team and staff injuries, this was not possible. Instead,
Application 2 made selective coverage, with priority given
to areas known from index trapping to be the best rodent
and rabbit habitat. On Bense Island, Application 2 covered
all tussock areas, all shorelines, and all areas north of the
island’s midline, regardless of vegetation type. Not covered
were some areas of inland heath south of the midline, and
the denuded burn zone. These latter areas still had intact
bait remaining from Application 1.
On Little Bense Island, Application 2 covered all
tussock areas and all shorelines, but did not cover inland
heath areas. As on Bense, the inland heath here still had
intact, uneaten bait remaining from the first application. On
Little Bense, staff injuries also curtailed coverage in the
tussock area: bait was applied on every second transect,
meaning there was a gap of 40 m (instead of the normal
20 m) between each baiting line. To help reduce the size of
the potentially un-baited space between transects, bait was
thrown wider on lateral throws, and a greater quantity was
thrown. The coastline was baited as in the first application,
including the use of the boat to access the north coast.
Mean baiting density on Application 2 was 11.7 kg/ha.
Over the whole operation, c. 3,900 kg of bait were applied
to the islands.
Daily reviews of the GPS tracks of workers revealed
that some areas were missed in the earliest days of baiting
but these were easily remedied the following day. After the
first two days, all workers had mastered navigation and no
further areas needed remediation.
Mammal and non-target mortality
Staff stayed on the islands from the start of Application
1 until seven days after the completion of Application 2
and during this time staff searched for animals killed in
the operation. In total, 64 dead rabbits were found on
Bense Island but none was found on Little Bense. All
intact carcasses found were placed under heavy tussock
grass or in burrows to hide them from scavenging birds.
However, many carcasses were discovered after they had
been scavenged, so some secondary poisoning is likely
to have occurred. Three dead mice were found on Bense
Island and one was found on Little Bense. No dead rats
were found on either island, presumably because they died
in their burrows.
Dolphin gulls (Larus scoresbii) were the most common
non-target casualty with a total of 23 carcasses discovered.
This species was observed to eat bait pellets directly,
often fighting conspecifics for them. Dolphin gulls were
the only species seen to eat the pellets. Three dead adult
turkey vultures (Cathartes aura jota) were found, as was
one adult striated caracara (Phalcoboenus australis) and
one short-eared owl (Asio flammeus). The owl had been
scavenged before discovery. Dissection of the striated
caracara showed no visual evidence that it had directly
ingested bait pellets, so perhaps it died from eating parts
of a poisoned animal, most likely a rabbit or dolphin gull.
Striated caracaras were observed playing with pellets but
were never observed to ingest them. Two dead flightless
steamer ducks (Tachyeres brachypterus) were found (one on
111
Island invasives: scaling up to meet the challenge. Ch 1B Rodents: Review
each island). Direct consumption of bait may explain these
deaths, but this species is known to eat offal occasionally
(Woods, 1975) so it is also possible they were victims of
secondary poisoning from eating a dead dolphin gull. Kelp
gulls (Larus dominicanus) and snowy sheathbills (Chionis
alba), two birds known for their curiosity and scavenging
habits, were both present but were not seen to touch the
pellets and no dead kelp gulls or sheathbills were recorded
during the operation.
Post-operation follow-up
During the December 2016 follow-up visit, informal
observations did not detect any live mammals, and no
footprints were found despite careful examination of areas
with soft soil or wet sand, where rabbit or rat tracks had
been commonly seen in the past. The bait stations on
Bense were also completely undisturbed with no evidence
of gnawing on the wax baits. Three freshly-dead kelp
gulls were found and evidence of pellet consumption was
discovered upon dissection: the crops of two of the birds
were discoloured with the bright green biomarker found in
the pellets. It is thought these birds consumed bait that was
inadvertently exposed during this visit when stored bait
was moved near the campsite.
The more thorough post-operation visit in November
2017 did not discover any evidence of rodents or mammals
on the islands. No live rodents or rabbits were seen,
nor were any fresh droppings or tracks discovered. No
chewsticks had been sampled by rodents, although the bite
marks of striated caracaras and other birds were found on
10 sticks. Nocturnal observations with the thermal camera
also found no mammals. However, bait blocks inside bait
stations were found to be heavily sand-blasted and in need
of replacement.
DISCUSSION
The first application of bait achieved 100% coverage
as per the project design. However, Application 2 was
less complete and several compromises were made, with
priority given to bait the areas shown by index trapping
to be the most important as habitat for invasive mammals.
However, one area of concern was the dense tussock on
Little Bense where the second application of baiting could
have left gaps between baiting lines.
Rats have proved easier to eradicate from islands than
mice, with rats successfully removed from islands in 92%
of the operations attempted (Howald, et al., 2007), whereas
early reports found success was achieved in only 62%
of mouse operations (MacKay, et al., 2007). However,
recent findings show a more optimistic picture, with mice
successfully eradicated in 77% of operations in New
Zealand, and this figure rises to 100% when considering
only operations that followed current best-practice
techniques (Broome, et al., 2019). Mice may be harder to
eradicate because of behavioural traits such as aversion
to cereal (Humphries, et al., 2000) or smaller home range
(Clapperton, 2006; MacKay, et al., 2011). This necessitates
a denser and more meticulous application of bait to ensure
that all mice encounter pellets. The possible gaps in bait
availability in dense tussock areas on Little Bense are thus
a cause for concern.
Eradication operations carry a risk of killing non-target
species through direct ingestion of poison pellets or by
eating an animal that was poisoned. At South Georgia,
brodifacoum pellets were consumed directly by skuas,
sheathbills, and pintails, while other scavengers such as
kelp gulls and giant petrels were less likely to eat baits
112
(Lee, et al., 2013). In contrast, at Campbell (McClelland,
2011) and Macquarie Islands (Springer & Carmichael,
2012) kelp gulls were found to be extremely vulnerable
to primary poisoning. In the Falklands, the death of nontarget species is not well known since most islands have
been without observers immediately after the completion
of baiting operations. However, on Great Island, the bodies
of many kelp and dolphin gulls were found following a rat
eradication operation in July 2016 (T. Poole, pers. comm.).
Dolphin gulls were the most common bird species poisoned
on Bense and Little Bense Islands and their corpses were
possibly a source of secondary poisoning of turkey vultures
and striated caracara. It is suggested that future eradication
operations in the Falklands plan for some personnel to
remain on the island after the completion of baiting in
order to improve understanding of non-target mortality.
That no evidence of mammals could be found on
the island 14 months post-baiting is cause for optimism.
However, the overall success of this operation will not
be known until late 2018 (26 months post-baiting) after
further monitoring has taken place. Elsewhere, rabbits have
proven particularly difficult to eradicate using poison alone
(Torr, 2002) and monitoring may reveal the need to use
additional techniques on Bense and Little Bense Islands.
There are no trained detection dogs in the Falklands and
snares and fumigants are not advised as they could have
an impact on burrowing penguins. In addition, biosecurity
concerns prevent the import of rabbit-specific pathogens,
thus leaving spotlight shooting as the most effective tool
available for eliminating any remaining rabbits.
This Bense and Little Bense islands operation was
intended to help restore native biodiversity with the potential
to re-establish populations of the endemic Cobb’s wren
and blackish cinclodes. However, it will also contribute
to future operations on other Falkland islands by allowing
landowners to understand which eradication techniques do,
or do not, work. As the first attempt to eradicate mice in the
Falklands, the results will be especially helpful in planning
for eradications on mouse-infested islands such as Steeple
Jason Island, which is home to many seabird species and
has been identified as an Important Bird Area. (Falklands
Conservation, 2006). In the Falkland Islands, private
landowners have been a driving force in many ecological
restoration projects, so the training and experience gained
by local residents in the course of the Bense operation
also serves to increase the pool of skilled staff who can
participate in future eradications on other islands.
ACKNOWLEDGEMENTS
Bait for this operation was generously donated by the
Falkland Islands Government. The SAFER foundation
is also grateful for support from N. Beazley, Encounter
Foundation NZ, the Environmental Planning Department of
the Falkland Islands Government, Falklands Conservation,
J. & S. Holman, The Baroness Hooper CMG and G. Roldan.
We thank P. Goodhue for creating the operational maps,
and A. Cox, E. Murphy, S. Poncet, and K. Springer for
advice on operational design. Key logistical support was
provided by Hurtigruten Cruises, Lindblad Expeditions,
K. McCallum, Ministry of Defence / British Forces
South Atlantic Islands, One Ocean Expeditions, and B.
& S. Pole-Evans. For baiting, we thank A. Cleminson, L.
Hartnoll, S. Henry, J. Henry, D. Hewitt, and T. Poole, and
for post-baiting monitoring, we thank J. Davis, R. Moore,
and K. Olds. Thanks to J. Fenton for the plant list. Critical
comments from P. Broady and two anonymous reviewers
improved this manuscript.
Carey: Mice & rabbits Bense & Little Bense, Falkland Islands
REFERENCES
Bergstrom, D.M., Lucieer, A., Kiefer, K., Wasley, J., Belbin, L., Pedersen,
T.K. and Chown, S.L. (2009). ‘Indirect effects of invasive species
removal devastate World Heritage island’. Journal of Applied Ecology
46: 73–81.
Broome, K.G., Brown, D, Cox, A., Cromarty, P., McClelland, P.,
Golding, C., Griffiths, R., and Bell, P. (2011). Current Agreed Best
Practice for Rodent Eradication – Hand Broadcasting Poison Bait
(Version 1.3). New Zealand Department of Conservation internal
document DOCDM-458533. Wellington, New Zealand: Department of
Conservation.
Broome, K., Brown, D., Brown, K., Murphy, E., Birmingham, C.,
Golding, C., Corson, P. and Cox, A. (2019). ‘House mice on islands:
management and lessons from New Zealand’. In: C.R. Veitch, M.N.
Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) Island invasives:
scaling up to meet the challenge, pp. 100–107. Occasional Paper SSC
no. 62. Gland, Switzerland: IUCN.
Poncet, S., Poncet, L., Poncet, D., Christie, D., Dockrill, C. and Brown,
D. (2011). ‘Introduced Mammal Eradications in the Falkland Islands
and South Georgia’. In: C.R. Veitch, M.N. Clout and D.R. Towns
(eds.) Island invasives: eradication and management, pp. 332–336.
Occassional Paper SSC no. 42. Gland, Switzerland: IUCN and
Auckland, New Zealand: CBB.
Springer, K. (2016). ‘Methodology and challenges of a complex multispecies eradication in the sub-Antarctic and immediate effects of
invasive species removal’. New Zealand Journal of Ecology 40: 273–
278.
Springer, K., and Carmichael, N. (2012). ‘Non-target Species Management
for the Macquarie Island Pest Eradication Project’. In: R.M. Timm (ed.)
Proceedings of the 25th Vertebrate Pest Conference, pp. 38–48. Davis,
USA: University of California, Davis.
St Clair, J.J.H. (2011). ‘The impacts of invasive rodents on island
invertebrates’. Biological Conservation 144: 68–81.
Brown, D., Chadderton, L. and Brown, K. (2001). Report on the
Falklands Conservation Rat Eradication Project. Stanley, F.I.:
Falklands Conservation.
Strange, I.J. (1992). A Field Guide to the Wildlife of the Falkland Islands
and South Georgia. London, UK: Harper Collins Publishers.
Carey, P.W. (2015). ‘Remote Ocean Outposts’. In: D. Liggett, B. Storey,
Y. Cook, and V. Meduna (eds.), Exploring the Last Continent, pp. 157–
173. Heidelberg, Germany: Springer.
Strange, I.J. (2007). New Island Falkland Islands: A South Atlantic
Wildlife Sanctuary for Conservation Management. Hook, UK: New
Island Conservation Trust.
Clapperton, K.A. (2006). A Review of the Current Knowledge of Rodent
Behaviour. Science for Conservation 263. Wellington, New Zealand:
Department of Conservation.
Tabak, M.A., Poncet, S., Passfield, K. and Martinez del Rio, C. (2014).
‘Invasive species and land bird diversity on remote South Atlantic
islands’. Biological Invasions 16: 341–352.
Cunningham, D.M. and Moors, P.J. (1996). Guide to the Identification
and Collection of New Zealand Rodents, 3rd Edition. Wellington, New
Zealand: Department of Conservation.
Tabak, M.A., Poncet, S., Passfield, K., Goheen, J.R., Martinez del Rio, C.
and Griffen, B. (2015). ‘Rat eradication and the resistance and resilience
of passerine bird assemblages in the Falkland Islands’. Journal of
Animal Ecology 84: 755–764.
Falklands Conservation (2006). Important Bird Areas of the Falkland
Islands. London, UK: Falklands Conservation, Royal Society for the
Protection of Birds, and BirdLife International.
FIG. (2015). Rat-free Islands in the Falkland Islands Spreadsheet.
Version May 2015. Falkland Islands Government.
FITB. (2016). Falkland Islands Tour Guide Handbook. 3rd Edition.
Stanley, F.I.: Falkland Islands Tourist Board.
Helmstedt, K.J., Shaw, J.D., Bode, M., Terauds, A., Springer, K.,
Robinson, S.A., Possingham, H.P. and Driscoll, D. (2016). ‘Prioritizing
eradication actions on islands: It’s not all or nothing’. Journal of Applied
Ecology 53: 733–741.
Howald, G.C., Donlan, J., Galvan, J.P., Russell, J.C., Parkes, J.,
Samaniego, A., Wang, Y., Veitch, D., Genovesi, P., Pascal, M., Saunders,
A., and Tershy, B.R. (2007). ‘Invasive rodent eradication on islands’.
Conservation Biology 21: 1258–1268.
Humphries, R.E., Sibly, R.M. and Meehan, A.P. (2000). ‘Cereal aversion
in behaviourally resistant house mice in Birmingham, UK’. Applied
Animal Behaviour Science 66: 323–333.
Tabak, M.A., Anderson, O.R.J., Robb, G., Poncet, S., Passfield, K.,
Martinez, M.G. and Martinez Del Rio, C. (2016). ‘Comparative
isotopic natural history of two native passerines (Troglodytes cobbi and
Cinclodes antarcticus) and the invasive rats (Rattus norvegicus) that
extirpate them’. Austral Ecology 41: 622–632.
Torr, N. (2002). ‘Eradication of rabbits and mice from Subantarctic
Enderby and Rose Islands’. In: C.R. Veitch and M.N. Clout (eds.).
Turning the tide: the eradication of invasive species, pp. 319–328.
Occasional Paper SSC no. 28. IUCN SSC Invasive Species Specialist
Group. IUCN, Gland, Switzerland and Cambridge, UK.
Woods, R.W. (1975). The Birds of the Falkland Islands. Owestry, UK:
Anthony Nelson.
Woods, R.W. and Woods, A. (1997). Atlas of Breeding Birds of the
Falkland Islands. Owestry, UK: Anthony Nelson.
Woods, R.W. and Woods, A. (2006). Birds and Mammals of the Falkland
Islands. Old Basing, UK: Wild Guides.
Lee, J., Black, A., Parker, G., and Rexer-Huber, K. (2013). Report on
Mortality of Non-target Species Following Year 1 Phase 2 of the South
Georgia Rodent Eradication. Stanley, F.I.: Government of South
Georgia and the South Sandwich Islands.
MacKay, J.W.B., Russell, J.C., and Murphy, E.C. (2007). ‘Eradicating
house mice from islands: successes, failures, and the way forward’. In: G.
W. Witmer, W. C. Pitt, and K.A. Fagerstone (eds.) Managing Vertebrate
Invasive Species: Proceedings of an International Symposium, pp.
294–304. USDA/APHIS/WS, National Wildlife Research Center, Fort
Collins, CO.
MacKay, J.W.B., Murphy, E.C., Anderson, S.H., Russell, J.C., Hauber,
M.E., and Clout, M.N. (2011). ‘A successful mouse eradication
explained by site-specific population data’. In: C.R. Veitch, M.N. Clout,
and D.R. Towns (eds.) Island invasives: eradication and management,
pp. 198–203. Occassional Paper SSC no. 42. Gland, Switzerland: IUCN
and Auckland, New Zealand: CBB.
McClelland, P.J. (2011). ‘Campbell Island – pushing the boundaries of rat
eradications’. In: C.R. Veitch, M.N. Clout, and D.R. Towns (eds.) Island
invasives: eradication and management, pp. 204–207. Occassional
Paper SSC no. 42. Gland, Switzerland: IUCN and Auckland, New
Zealand: CBB.
Miller, C. (2008). Report from Rat Eradication Workshop, Stanley,
Falkland Islands. Sandy, UK: Royal Society for the Protection of Birds.
Moore, D.M. (1968). The Vascular Flora of the Falkland Islands. British
Antarctic Survey Reports No. 60. London, UK: Natural Environment
Research Council.
113
P. Garden, P. McClelland and K. Broome
Garden, P.; P. McClelland and K. Broome. The history of the aerial application of rodenticide in New Zealand
The history of the aerial application of rodenticide in New Zealand
P. Garden1, P. McClelland2 and K. Broome3
1
5 Ruby Ridge, Wanaka 9803, New Zealand. <pghltd@xtra.co.nz>. 2 237 Kennington Roslyn Bush Road, Invercargill
9876, New Zealand. 3 Department of Conservation, P.O. Box 10-420, Wellington 6143, New Zealand.
Abstract Following the incursion of rats (Rattus rattus) on Taukihepa (Big South Cape Island; 93.9 km²) off southern
New Zealand in 1963, and the subsequent extirpation of several endemic species, the New Zealand Wildlife Service
realised that, contrary to general belief at the time, introduced predators do not reach a natural balance with native species
and that a safe breeding habitat for an increasing number of ‘at risk’ species was urgently needed. Offshore islands offered
the best option for providing predator free habitat but there was a limited number of predator-free islands available and
most were very small. Eradicating rodents on larger islands to provide a wider range and greater area of habitats was
required and hand treating these larger areas using trapping and hand application of toxicants, the only methods available
at the time, proved problematic and often impossible. Helicopters had been used to distribute bait for the control of
rabbits and brushtail possums in the past but eradication of any particular predator species was considered ‘not feasible’.
The development of a GPS-based aircraft guidance system, a suitable bait product, specialised bait delivery systems and
second-generation anti-coagulant toxicants changed that. Now islands as large as South Georgia (3,900 km²) have been
treated using this method.
Keywords: aerial application, brodifacoum, eradication, helicopter, Mus musculus – house mouse, Rattus exulans –
Pacific rat, Rattus norvegicus – Norway rat, Rattus rattus – ship rat
INTRODUCTION
New Zealand’s terrestrial flora and fauna evolved in
isolation from mammalian predators leading to many
species being highly susceptible to any ground-based
predators that hunt by smell and sight (e.g. Tennyson &
Martinson, 2006). Since the arrival of humans, this unique
environment has suffered from the deliberate or accidental
introduction of a range of species that have decimated
native biodiversity. This includes four species of rodent,
Norway rat (Rattus norvegicus), ship rat (R. rattus), Pacific
rat or kiore (R. exulans) and house mouse (Mus musculus),
which continue to have a devastating impact on New
Zealand’s native flora and fauna (King, 2005).
three species of endemic vertebrates sent shock waves
through conservation circles (Bell, et al., 2016). A fourth
species was saved only by transferring to it a nearby
predator-free island. This disaster led to an increased
interest in the ecology of rodents and their impact on native
species as well as ways to control or eradicate them along
with other introduced predators (Towns & Broome, 2003).
Polynesians arrived in New Zealand bringing with
them the Pacific rat or kiore. The rats, along with kuri or
native dog (Canis familiaris), were brought for food and
clothing and led to the first wave of extinctions in New
Zealand (Tennyson & Martinson, 2006). In 1770, James
Cook mentions vermin in his journals and this may refer to
Norway rat (Innes, 2005). House mice had arrived in New
Zealand by 1830 (Ruscoe & Murphy, 2005). Ship rats were
introduced with early European settlers between 1860 and
1890 and had both cumulative and additional impacts to
the rodent species that were already present (King, 2005;
Tennyson & Martinson, 2006).
Demobilised World War II pilots in New Zealand
began an industry applying fertiliser and grass seed to hill
country and established the skills to fly accurate parallel
swath patterns. The spread of fertiliser and seed initially
used fixed wing aircraft as outlined by Alexander & Tullett
(1967), but the skills were later transferred to the use of
helicopters.
ERADICATION TOOLS AND ADVANCES
Early application of aircraft in New Zealand
agriculture
The skill and experience of the pilots is a crucial
component of any aerial baiting operation. In addition to
having experience with all the systems that are to be used
Invasive species have caused ecological problems
around the world since humans started exploring but it was
in New Zealand, where biodiversity loss was obviously
due to introduced predators (Tennyson & Martinson,
2006), that organisations began to consider ways to
minimise these impacts. It was not until the mid-1990s that
technology advanced to a stage where this human induced
disaster could be offset on any significant scale (Towns, et
al., 2013).
This paper outlines the key events that led to the
development of a rodent eradication tool used around the
world today and discusses the role played in this process by
the New Zealand agricultural aviation industry.
RECOGNITION OF THE DAMAGE RODENTS
COULD DO TO NEW ZEALAND WILDLIFE
The ship rat invasion of Taukihepa (Big South Cape
Island; 93.9 km²) in the early 1960s and the extinction of
Fig. 1 Auster aircraft loading rabbit bait, MacKenzie Basin
1951.
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
114
up to meet the challenge, pp. 114–119. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Garden, et al.: History of aerial rodenticide, NZ
in the operation (e.g. helicopter, bucket, GPS etc.), they
are often required to fly under adverse conditions such
as during poor weather, across islands with challenging
topography and frequently a high risk of bird strikes.
Pilots are expected to fly accurate lines in spite of these
challenges whilst also monitoring the bait flow out of the
bucket. It is highly desirable that the pilots are involved in
the planning for an eradication as they can identify both
risks and opportunities associated with the bait application.
The establishment of the Department of Conservation
The establishment of the New Zealand Government’s
Department of Conservation (DOC) out of the Wildlife
Service, Forest Service and Department of Lands and
Survey brought the various government agencies charged
with protecting biodiversity under one management regime
and allowed better focus on prioritising ‘endangered
species’ programmes, including predator removal. The
Department of Conservation was able to provide the
financial and political support necessary to carry out this
work. This was especially so with the larger projects such
as Campbell Island (113.3 km²) in the New Zealand subAntarctic. Current operations now follow the international
trend of joint venture or partnership operations with
Non-Government Organisations (NGOs) and private
conservation trusts.
Fig. 3 Purpose built eradication bucket 2001.
IMPROVEMENTS IN TECHNOLOGY
Development of toxins
Development of bait spreading equipment
On the mainland, compressed grain bait (pellets)
suitable for dispersal through a mechanised spreader
bucket (Fig. 2) were also laced with 1080 and phosphorus
to target brushtailed possums (Trichosurus vulpecula) (Bill
Simmons pers. comm.). Prior to this, aerial bait application
had been predominantly diced carrot or grain.
Various New Zealand agricultural helicopter companies
had been developing underslung cargo hook-mounted
spreader buckets for the application of fertiliser and seed.
By 1980, these spreader buckets had been modified to
spread toxin-laced chopped carrot and cereal-based pellets
for the control of rabbits and possums (Peter Garden,
unpublished data).
The development of the second-generation blood
anticoagulant toxicant brodifacoum in England in the mid1970s provided a toxicant suitable for large-scale rodent
eradication (Dubock & Kaudeinen, 1978). The delayed
action of the anticoagulant toxicants meant that rodents
would consume a lethal dose of toxicant before showing
any symptoms, thus eliminating the risk of bait avoidance.
Brodifacoum also has the ability to kill a rodent with a single
feed, compared to the first-generation anti coagulants that
required multiple feeds over several days. Brodifacoum is
currently registered in over 40 countries in the form of over
100 separate registrations covering different formulations
or product forms (Kaudeinen & Rampaud, 1986).
Purpose-built bait-spreading buckets have continued to
be developed (Fig. 3), and these now allow for a consistent
swath width and density of bait application on a large
scale. Buckets have been repeatedly refined to provide a
wider bait swath and, most importantly, the addition of an
internal deflector to direct bait just out one side minimising
any bait that may go into the marine environment as well
as being able to treat cliffs. Additional improvements
including linking the bait flow to the flight track recording
system are currently being developed.
Development of guidance and data recording
equipment
Various methods to assist pilots in following straight
lines have been tried. One of these, the Decca Navigation
System, was used on forestry spraying operations as
early as 1980 and used in a possum control operation on
Rangitoto Island in 1990. Another method trialled was
using reciprocal compass headings at the end of each run.
This required the pilot to make calculations using compass
variation, deviation and cross wind headings.
Fig. 2 Compressed cereal bait impregnated with
brodifacoum.
The United States military developed a constellation of
global orbiting satellites in the late 1970s to provide very
accurate navigation information. The Global Positioning
System (GPS) relies on highly accurate time and position
information transmitted by these satellites to receivers on
the ground or in aircraft. The receivers use triangulation
to compute three-dimensional position, direction and
speed of travel information. To preserve security of this
information, deliberate errors were factored in and the
corrections for these errors were only available to those
with security clearance to use them. This error factor was
115
Island invasives: scaling up to meet the challenge. Ch 1B Rodents: Review
known as ‘selective availability’. The civilian world was
keen to access this information and several companies
developed simple navigation devices that could be used
for guidance within the expected error range. The error
range was not consistent but was never much more than
a few hundred meters, which was acceptable to support
other navigational equipment. However, to be an effective
guidance tool for aerial application this error could be no
more than one or two metres. In 1993, attempts were made
to use GPS for guiding bait spread onto Cuvier Island, but
a suitable satellite triangulation system at that time was not
available (D.R. Towns, pers. comm.).
In 1995, an American avionics manufacturer, Trimble
Navigation, set up a facility in Christchurch New Zealand
with the specific purpose of developing systems for use
in aerial agricultural application that could meet the very
stringent accuracy requirements of that industry. The
system required the use of a ‘base station’ that recorded
satellite signals transmitted over time and calculated the
errors. The corrected information was then transmitted to
the aircraft by radio telemetry.
By 2000, the US military had switched off the
‘selective availability’ function so the use of base stations
was no longer necessary. More recently, a New Zealand
based company, TracMap Ltd™, has developed a system
designed specifically for aerial application – for the
distribution of both agricultural products and bait (Fig. 4).
The first island eradication where GPS guidance
equipment was successfully used was on Tiritiri Matangi
(1.7 km²) in 1993 (Veitch, 2002d).
ERADICATION HISTORY
Early aerial application of toxicants
Rabbits (Oryctolagus cuniculus) were introduced for
sport and as supplementary food for settlers in the 1830s
(King 2005). However, the animals soon developed into
plague proportions, particularly in the drier inland areas
where they contributed to significant land erosion (King,
2005). Systems were developed for the aerial application of
toxicants to control rabbits using fixed wing aircraft (Fig. 1).
This was predominantly using either carrot pieces or grain
laced with the toxin 1080 (sodium monofluoroacetate).
The first recorded island rat eradication in New Zealand
was the removal of Norway rats by hand baiting from
Maria Island (1 ha), Noises Islands, in 1960 (Towns &
Broome, 2003). as the first in a series of unintended rodent
eradications when control had been the expected outcome.
Fig. 4 TracMap™ GPS guidance equipment fitted to South
Georgia Heritage Trust aircraft,2015.
116
The first use of bait stations was by Ian McFadden on
Rurima Island (0.045 km²) in 1983, using maize laced with
the anticoagulant bromadialone and the same product was
used successfully on Korapuki Island (0.18 km²) in 1986.
Both campaigns were against Pacific rats (and rabbits on
Korapuki, McFadden & Towns, 1991). Between 1986
and 1988, commercially available Talon™ (brodifacoum)
wax blocks in bait stations were used to eradicate Norway
rats from Hawea (9 ha) and Breaksea (1.70 km²) islands
in Fiordland (Thomas & Taylor, 2002). While this type
of technique has been used on islands as large as 31
km² Langara Island, Canada (Taylor, et al., 2000), the
usefulness of this method is limited by topography of
the target island and logistical difficulties associated with
ensuring complete coverage of the island.
Early use of aircraft targeting rodents on islands
In 1986, Moutohora Island (1.43 km²) in the Bay of
Plenty was the first island in New Zealand to be treated
using aerially distributed toxic bait (Talon™ 20P, active
ingredient brodifacoum) to target rabbits using a fertiliser
spreading bucket. As an unplanned side effect, Norway rats
were also removed as part of this operation (Jansen, 1993).
The first attempt at aerially distributing rodenticide
targeting rats in New Zealand occurred on the Mokohinau
Islands (0.73 km²) in the Hauraki Gulf in 1990 (Towns &
Broome, 2003). This operation was carried out using a
‘monsoon’ firefighting bucket to spread Talon™ 20P and
resulted in the removal of Pacific rats. However, it was
identified that the bait spread was concentrated along a
narrow swath, due to the bucket not having a spinner to
spread the bait out, and hand spreading was required to fill
in the gaps (McFadden & Greene, 1994).
Between 1991 and 1993 a partnership was developed
between DOC and ICI Crop Care, to improve the durability
of Talon™ 20P (brodifacoum) and to license the product
for aerial spread against rodents. An efficient means of
spreading the baits also needed to be developed. By 1993,
Ian McFadden of DOC and Tony Monk of Heletranz
had developed a bait bucket with spinner, for use against
rodents on offshore islands. The bucket was used to spread
Talon™ 20P to target Pacific rats on Cuvier Island (1.81
km²) in 1993 (Towns & Stephens, 1997).
Increasing the scale
The first large scale aerial application operation
specifically targeting rodents (Norway and Pacific rats)
was carried out on 19.65 km² Kapiti Island (Fig. 5) off
the south-west side of the North Island, New Zealand
(Miskelly & Empson, 1999). The operation succeeded in
removing both species. This island was four times larger
than any previously attempted (Broome, 2009).
Fig. 5 Mechanical loading of bait for Kapiti Island, 1996.
Garden, et al.: History of aerial rodenticide, NZ
Fig. 6 Hand loading bait on Codfish/Whenua Hou, 1997.
Pacific rats were eradicated from Putauhina Island
(1.41 km²) and Raratoka Island (0.88 km²) off southern
Stewart Island in 1997 in the lead-up to rodent eradication
on Whenua Hou (Codfish Is; 13.96 km²). (McClelland,
2002) Although these islands had significant conservation
values in their own right, the removal of rats was largely to
establish procedures and issues for the treatment of Whenua
Hou in order to provide a predator free environment to
establish a kakapo breeding base (Merton, et al 2006)
In August 1998, two applications of brodifacoum-laced
compressed cereal bait were aerially applied to 13.96 km²
Whenua Hou (Fig. 6) to remove Pacific rats (McClelland,
2011). The Kapiti project used two applications and this
has become the standard methodology for aerial bait
applications for eradicating rats on islands worldwide, with
modifications as required for each island.
Tuhua/Mayor Island (12.83 km²) in the Bay of Plenty,
New Zealand was successfully treated for the removal of
Norway rats and Pacific rats in 2000, largely to test the
methods required against rats and cats on the much larger
and more remote Raoul Island in the Kermadecs (Williams
& Jones, 2003).
Campbell Island followed on from the success of the
Kapiti and Codfish/Whena Hou eradication programmes.
DOC embarked on a very ambitious plan to eradicate
Norway rats from this 113.31 km2 island, 700 kilometres
south of mainland New Zealand. The logistics of this
project far exceeded anything that had been contemplated
previously and required a rethink on how such operations
could be streamlined to make them logistically and
Fig. 7 Spreading bait on cliffs, Campbell Island, 2001.
Fig. 8 Bait spreading on Mokonui Island, off Stewart Island,
2006.
financially feasible. The resulting operational plan called
for a single application of just 50% of the standard bait
rate. This was a substantial risk but the GPS navigation
and spreader bucket technology and experienced pilots
gave planners confidence in being able to achieve complete
coverage. A 600 ha trial involving the aerial application of
non-toxic bait with a biomarker was carried out to test the
proposed methodology before the full operation (Fig. 7)
was started. (McClelland, 2011). Over the period 2000 to
2008, more than a dozen islands around the New Zealand
coastline were treated including: Raoul (29.38 km²) in the
Kermadecs (Ambrose, 2006; Little Barrier (30.83 km²) in
the Hauraki Gulf (Griffiths, et al., 2019); Bench (1.21 km²)
and Pearl (5.12 km²) off Stewart Island (Brent Beaven pers.
comm.); Coal (11 km²) Preservation Inlet, (Brown, 2013);
Pomona (2.62 km²) and Rona Islands (0.6 km²) (Shaw &
Torr, 2011). Notable during this period was the Rakiura Titi
Islands restoration project (McClelland, et al., 2011) which
included Mokonui (0.86 km²) (Fig. 8) and Taukihepa/
Big South Cape (9.39 km²) islands. Managing non-target
risks, multi-species eradications and reinvasion issues are
all now part of the planning process and this culminated
in the Rangitoto/Motutapu project 34.81 km² in 2009 that
targeted seven species of introduced mammals including
the four species of rodent (M. musculus, R. rattus, R.
norvegicus, R. exulans). (Griffiths, et al., 2015).
Mice removal from 20.02 km² Antipodes Island 850
km south-east of Bluff (New Zealand) occurred in winter,
2016 (Horn & Hawkins, 2017) (Fig. 9). Success has been
confirmed.
Fig. 9 Mouse eradication operations Antipodes Island,
2016.
117
Island invasives: scaling up to meet the challenge. Ch 1B Rodents: Review
INTERNATIONAL PROJECTS
Exporting the technology
Because of the concern for the critically endangered
Seychelles magpie robin (Copsychus sechellarum), an
operation to carry out the eradication of Norway rats from
Denis (1.43 km²), Frigate (2.19 km²) and Curieuse (2.86
km²) Islands in the Seychelles was completed in June and
July 2000 (Merton, et al., 2002).
The same basic technique, usually using New Zealandmade spreader buckets and often with experienced New
Zealand pilots, has been and is used to eradicate rodents
on islands worldwide. Methods are modified for each
island with alterations made to sowing density, number
of drops, timing between drops etc., To date rodents
have been eradicated from more than 300 islands using
this technique, making it the most widely used and most
successful technique for rodent eradications compared
to bait stations, hand broadcast or traps. (Howald, et al.,
2007). Whereas there are still some situations where the
other techniques are the most suitable option, e.g. on
islands where it is not practical to use aerial eradication
methods it has allowed islands that could never previously
have been considered for eradication programmes to be
treated successfully. The largest island worked on to date
is 3900 km² (1070 km² treated) South Georgia Island in the
sub-Antarctic (Black, et al., 2013), which had Norway rats
and an isolated population of mice treated in three phases
over a five-year period from 2011 to 2015. Other successful
international eradications using this methodology include
Macquarie (128 km²) where rabbits, ship rats and mice
were eradicated in 2012 (Parks and Wildlife Service, 2014)
and Rat Island/Hawadax (10 km²) in the Aleutians where
Norway rats were eradicated in 2008 (Buckelew, et al.,
2011).
Aerial distribution of bait has now been successfully
used for the eradication of rodents in more than ten
countries including Australia, USA, Canada, Mexico,
Japan, Italy and several smaller Pacific Island nations.
CONCLUSION
The aerial dispersal of rodenticide has been a ‘game
changer’ allowing large and geographically challenging
islands and tracts of land to be treated quickly and efficiently.
The advent of GPS guidance and recording equipment and
purpose-built distribution systems (spreader buckets) has
given project managers confidence that a lethal dose of
toxic bait can be delivered into each home range of the
target species, maximising the chances of eradication.
Many organisations and islands around the world have
benefited from the developments carried out in New Zealand
since the availability of second-generation anticoagulant
toxicants. Now NGOs and Government departments in
all corners of the globe are using this information to carry
out their own projects. These in turn are now providing
feedback to advance the knowledge base needed to carry
out ever more complex and challenging projects.
While aerial application of toxic bait has been a
major advancement in habitat restoration, ground based
techniques – bait stations and hand broadcast – are still used
where relevant. These methods tend to be used on smaller,
more accessible islands as well as around dwellings on
inhabited islands during aerial operations. However, the
ability to treat large areas in a short space of time and the
lower overall cost per hectare of treatment make aerial
application a valuable tool in the continuing fight against
invasive predators. The scale of islands that may be treated
in the future is limited only by the supporting logistics,
funding and political support.
118
The fact that much of the aerial application expertise
resides in New Zealand has more to do with the incremental
development of systems, procedures and technology that
has occurred here over the past 30 years. As the baiting
pilot has the final control over the success of any project,
it is vital that they have complete commitment to that end.
Project managers should involve the likely application
pilot(s) at an early stage to ensure this commitment.
Many challenges still exist, especially in tropical and
subtropical regions where success rates have been lower,
and there is room for continued development of equipment
and systems, but the use of this method of distribution
of rodenticide will continue into the foreseeable future.
An increasing number of inhabited islands is now being
treated and this brings a new series of challenges for
project managers. Numerous issues that do not need to be
considered on uninhabited islands come into play, making
these operations considerably more complex.
ACKNOWLEDGEMENTS
We would like to thank David Towns, James Russell,
Dick Veitch, Richard Griffiths, Keith Hawkins and Don
Sanders for helping to fill in the gaps. Photo credits John
King (New Zealand Aviation News), Colin Miskelly, Pete
McClelland, Pete Tyree and Stephen Horn.
REFERENCES
Ambrose, M. (2006). Operational Report for Cat, Kiore and Norway
Rat Eradication on Raoul Island in the Kermadec Group. Unpublished
Report, Department of Conservation, New Zealand.
Alexander, G. and Tullett, J.S. (1967). The Super Men. Wellington: A.H.
& A.W. Reed.
Bell, E.A.; Bell, B.D. and Merton D.V. (2016). ‘The legacy of Big South
Cape: rat irruption to rat eradication’. New Zealand Journal of Ecology
40(2): 212–218.
Black, A., Poncet, S., Wolfaardt, A., Peters, D., Hart, T., Wolfaardt, L-A.,
Tasker, M. and Rexer-Huber, K. (2013). Rodent Eradication on South
Georgia – Preparation and Evaluation: A Summary Report of Activities
During the 2011/2012 Field Season. United Kingdom: Overseas
Territory Environment Programme.
Broome, K.G. (2009) ‘Beyond Kapiti – A decade of invasive rodent
eradications from New Zealand islands’. Biodiversity 10: 14–24.
Brown, D. (2013). Coal Island (Te Puka-Hereka) Restoration Plan.
Unpublished Report, Department of Conservation, Southland
Conservancy, New Zealand.
Buckelew, S., Byrd, V., Howald, G., MacLean, S. and Sheppard, J.
(2011). ‘Preliminary ecosystem response following invasive Norway
rat eradication on Rat Island, Aleutian Islands, Alaska’. In: C.R. Veitch,
M.N. Clout and D.R. Towns (eds.) Island invasives: eradication and
management, pp. 275–279. Occassional Paper SSC no. 42. Gland,
Switzerland: IUCN and Auckland, New Zealand: CBB.
Cash, W. and Gaze, P. (2000). ‘Restoration of Motuara Island – Queen
Charlotte Sound’. Ecological Management 8: 31–36.
Chappell, R. (2008). Operational Plan for Eradication of Ship Rats (Rattus
rattus), Kiore (Rattus exulans), House Mouse (Mus musculus) and
Rabbit (Oryctolagus cuniculus cuniculus) from Ohinau and Motutapere
Islands. Department of Conservation, New Zealand. DOCDM-314307.
Corson, P. and Hawkins, K. (2016). The Eradication of Kiore, Ship
Rats and Cats from Ahuahu - Great Mercury Island. Department of
Conservation internal report DOCDM-1477863.
Dubock, A.C. and Kaukeinen, D.E. (1978). ‘Brodifacoum (TalonTM
Rodenticide), a Novel Concept’. In: W.E. Howard and R.E. Marsh
(eds.). Proceedings of the Eight Vertebrate Pest Conference, p. 127.
Lincoln: University of Nebraska.
Griffiths, R., Buchanan, F., Broome, K., Neilson, J., Brown, D. and
Weakley, M. (2015). ‘Successful eradication of invasive vertebrates on
Rangitoto and Motutapu islands, New Zealand’. Biological Invasions
17: 1355–1369.
Garden, et al.: History of aerial rodenticide, NZ
Griffiths, R., Bell, E.A., Campbell, D.J., Cassey, P., Ewen, J.G., Green,
C.J., Joyce, L., Rayner, M.J., Toy, R., Towns, D.R., Wade, L., Walle, R.
and Veitch, C.R. (2019) ‘Costs and benefits for biodiversity following
rat and cat eradication on Te Hauturu-o-Toi/Little Barrier Island’. In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.)
Island Invasives: Scaling Up to Meet the Challenge, pp. 558–567.
Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Golding, C. (2010). ‘House mouse (Mus musculus) eradication by aerial
bait application on Adele, Tonga and Fisherman Islands, Abel Tasman
National Park, New Zealand’. Conservation Evidence 7: 62–88.
Gutsell, M. (2011). Indian Island Rodent Eradication. Department of
Conservation internal report DOC-752616.
Hawkins, K. (2011). Taranga (Hen) Restoration Project Restore, Taranga
Island kiore eradication programme final report May to June 2011.
Department of Conservation internal report DOC-5455561.
Howald, G., Donlan, C.J., Galván, J.P., Russell, J.C., Parkes,
J., Samaniego, A., Wang, Y., Veitch, D., Genovesi, P., Pascal,
M., Saunders, A. and Tershy, B. (2007). ‘Invasive rodent eradication on
islands’. Conservation Biology 21: 1258–1268.
Horn, S.R. and Hawkins, K. (2017). Project Report, Antipodes Island
Mouse Eradication. Department of Conservation internal report DOC3000055.
Innes, J. (2005). ‘Norway Rat’. In: C.M. King (ed.) The Handbook of New
Zealand Mammals, pp.174–187. Melbourne: Oxford University Press.
Jansen, W.P. (1993). ‘Eradication of Norway rats and rabbits from
Moutohora (Whale) Island, Bay of Plenty’. Ecological Management 1:
10–15.
King, C.M. (2005). The Handbook of New Zealand Mammals. Melbourne:
Oxford University Press.
Kaudeinen, D.E. and Rampaud, M.A. (1986). ‘Review of Brodifacoum
Efficacy in the US and Worldwide’. In: P. Salmon, R.E. Marsh and D.E.
Beadle (eds.) Proceedings of the Twelfth Vertebrate Pest Conference, p.
17. California: University of California, Davis.
Lee, M. (2016) Motu Kaikoura Sanctuary Rat Management Programme
– Progress Report (No.3). Unpublished report. Motu Kaikoura Trust.
McClelland, P.J. (2002). ‘Eradication of Pacific rats from Whenua Hou’.
In: C.R. Veitch and M.N. Clout (eds.) Turning the tide: the eradication
of invasive species, pp. 181–189. Occasional Paper SSC no. 28. IUCN
SSC Invasive Species Specialist Group. IUCN, Gland, Switzerland and
Cambridge, UK.
McClelland, P.J. (2011). ‘Campbell Island, Pushing the Boundaries’.
In: C.R. Veitch, M.N. Clout and D.R. Towns (eds.) Island invasives:
eradication and management, pp. 204–207. Occassional Paper SSC no.
42. Gland, Switzerland: IUCN and Auckland, New Zealand: CBB.
McClelland, P.J., Coote, R., Trow, M., Hutchins, P., Nevins, H.M.,
Adams, J., Newman, J. and Moller, H. (2011). ‘The Rakiura Tītī
Islands restoration project’: In: C.R. Veitch, M.N. Clout and D.R.
Towns (eds.) Island invasives: eradication and management, pp. 451–
454. Occassional Paper SSC no. 42. Gland, Switzerland: IUCN and
Auckland, New Zealand: CBB.
McFadden, I. and Towns, D.R. (1991). ‘Eradication campaigns against
kiore (Rattus exulans) on Rurima Rocks and Korapuki Island,
northern New Zealand’. Science and Research Internal Report No. 97.
Wellington: Department of Conservation.
McFadden, I. and Greene, T. (1994). ‘Using brodifacoum to eradicate
kiore (Rattus exulans) from Burgess Island and the Knights Group of the
Mokohinau Islands’. Science and Research Series No. 70. Wellington:
Department of Conservation.
MacKay, J.W.B., Murphy, E.C., Anderson, S.H., Russell, J.C., Hauber,
M.E., Wilson, D.J. and Clout, M.N. (2011). ‘A successful mouse
eradication explained by site-specific population data’. In: C.R. Veitch,
M.N. Clout and D.R. Towns (eds.) Island invasives: eradication and
management, pp. 198–203. Occassional Paper SSC no. 42. Gland,
Switzerland: IUCN and Auckland, New Zealand: CBB.
Merton, D.V., Climo, G., Laboudallon, V., Robert, S. and Mander, C.
(2002). ‘Alien mammal eradication and quarantine on inhabited islands
in the Seychelles’. In: C.R. Veitch and M.N. Clout (eds.) Turning the
tide: the eradication of invasive species, pp. 85–91. Occasional Paper
SSC no. 28. IUCN SSC Invasive Species Specialist Group. IUCN,
Gland, Switzerland and Cambridge, UK.
Merton, D.V., Powlesland, R.G. and Cockrem, J.F. (2006). ‘A parrot
apart: the natural history of the kakapo (Strigops habroptilus), and the
context of its conservation management’. Notornis 53(1): 3–26.
Owen, K. (1997). ‘Transformation of an Inland Island (Mokoia Island)
Now Home to Endangered Species’. In: T. Johnson and P. Boyes (eds.)
New Zealand Conservation Management Group Annual Conference
Proceedings, pp. 74–87. Rotorua: DOC Research & Development
Series 284.
Parks and Wildlife Service (2014). Evaluation Report: Macquarie Island
Pest Eradication Project, August 2014. Hobart, Tasmania: Department
of Primary Industries, Parks, Water and Environment.
Project Island Song (2009). ‘Building a Wildlife Sanctuary’. <www.
projectislandsong.co.nz>.
Ruscoe, W.A. and Murphy, E.C. (2005). ‘House Mouse’. In: C.M.
King (ed.) The Handbook of New Zealand Mammals, pp. 204–221.
Melbourne: Oxford University Press.
Shaw, V. and Torr, N. (2011). ‘Eradicating mammal pests from Pomona
and Rona Islands in Lake Manapouri, New Zealand: A focus on rodents’.
In: C.R. Veitch, M.N. Clout and D.R. Towns (eds.) Island invasives:
eradication and management, pp. 356–360. Occassional Paper SSC no.
42. Gland, Switzerland: IUCN and Auckland, New Zealand: CBB.
Taylor, R., Kaiser, G.W. and Drever, M.C. (2000). ‘Eradication of
Norway rats for recovery of seabird habitat on Langara Island, British
Columbia’. Restoration Ecology 8: 151–160.
Tennyson A. and Martinson P. (2006). Extinct Birds of New Zealand.
Wellington: Te Papa Press.
Thomas, B.W. and Taylor, R.H. (2002) ‘A History of Ground-based
Rodent Eradication Techniques Developed in New Zealand, 1959–
1993’. In: C.R. Veitch and M.N. Clout (eds.) Turning the tide: the
eradication of invasive species, pp. 301–310. Occasional Paper SSC
no. 28. IUCN SSC Invasive Species Specialist Group. IUCN, Gland,
Switzerland and Cambridge, UK.
Torr, N. (2002). ‘Eradication of rabbits and mice from Subantarctic
Enderby and Rose Islands’. In: C.R. Veitch and M.N. Clout (eds.)
Turning the tide: the eradication of invasive species, pp. 319–336.
Occasional Paper SSC no. 28. IUCN SSC Invasive Species Specialist
Group. IUCN, Gland, Switzerland and Cambridge, UK.
Towns, D.R. and Stephens, T. (1997). ‘Island management and
commercial sponsorship: The Mercury Islands experience’. Science
and Research Series No. 103. Wellington: Department of Conservation.
Towns, D.R. and Broome, K.G. (2003). ‘From small Maria to massive
Campbell: Forty years of rat eradications from New Zealand islands’.
New Zealand Journal of Zoology 30: 377–398.
Towns, D.R. and Parrish, R (2003). ‘Restoration of the Principal Marotere
Islands’. Department of Conservation, New Zealand; Resource
Management Unit, Ngatiwai Trust Board.
Towns, D.R., West, C.J. and Broome, K.G. (2013). ‘Purposes, outcomes
and challenges of eradicating invasive mammals from New Zealand
islands: an historical perspective’. Wildlife Research 40: 94–107.
Veitch, C.R. (2002a). ‘Eradication of Norway rats (Rattus norvegicus)
and house mouse (Mus musculus) from Browns Island (Motukorea),
Hauraki Gulf, New Zealand’. In: C.R. Veitch and M.N. Clout (eds.)
Turning the tide: the eradication of invasive species, pp. 350–352.
Occasional Paper SSC no. 28. IUCN SSC Invasive Species Specialist
Group. IUCN, Gland, Switzerland and Cambridge, UK.
Veitch, C.R. (2002b). ‘Eradication of Norway rats (Rattus norvegicus)
and house mouse (Mus musculus) from Motuihe Island, New Zealand’.
In: C.R. Veitch and M.N. Clout (eds.) Turning the tide: the eradication
of invasive species, pp. 353–356. Occasional Paper SSC no. 28. IUCN
SSC Invasive Species Specialist Group. IUCN, Gland, Switzerland and
Cambridge, UK.
Veitch, C.R. (2002c). ‘Eradication of Pacific rats (Rattus exulans) from
Fanal Island, New Zealand’. In: C.R. Veitch and M.N. Clout (eds.)
Turning the tide: the eradication of invasive species, pp. 357–359.
Occasional Paper SSC no. 28. IUCN SSC Invasive Species Specialist
Group. IUCN, Gland, Switzerland and Cambridge, UK.
Veitch, C.R. (2002d). ‘Eradication of Pacific Rats (Rattus exulans) from
Tiritiri Matangi Island, Hauraki Gulf, New Zealand’. In: C.R. Veitch and
M.N. Clout (eds.) Turning the tide: the eradication of invasive species,
pp. 360–364. Occasional Paper SSC no. 28. IUCN SSC Invasive Species
Specialist Group. IUCN, Gland, Switzerland and Cambridge, UK.
Williams, D. and Jones, A. (2003). Operational Report for Cat, Kiore,
Norway Rat Eradication on Mayor Island (Tuhua). Unpublished
Report, Department of Conservation, New Zealand.
Miskelly, R.A. and Empson, C.M. (1999). ‘The risks, costs and benefits of
using brodifacoum to eradicate rats from Kapiti Island, New Zealand’.
New Zealand Journal of Ecology 23: 241–254.
119
R. Griffiths, D. Brown, B. Tershy, W.C. Pitt, R.J. Cuthbert, A. Wegmann, B. Keitt, S. Cranwell and G. Howald
Griffiths, R.; D. Brown, B. Tershy, W.C. Pitt, R.J. Cuthbert, A. Wegmann, B. Keitt, S. Cranwell and G. Howald.
Successes and failures of rat eradications on tropical islands: a comparative review of eight recent projects
Successes and failures of rat eradications on tropical islands: a
comparative review of eight recent projects
R. Griffiths1, D. Brown2, B. Tershy3, W.C. Pitt4, R.J. Cuthbert5, A. Wegmann1, B. Keitt1, S. Cranwell6 and G. Howald7
Island Conservation, Santa Cruz, CA 95060, USA. <rgriffiths@islandconservation.org>, 2Picton 7281, New Zealand.
Coastal Conservation Action Lab, Center for Ocean Health, Santa Cruz, CA 95060, USA. 4Smithsonian Conservation
Biology Institute, Front Royal, VA 22630, USA. 5Royal Society for the Protection of Birds, Bedfordshire, SG19 2DL,
United Kingdom. World Land Trust, Halesworth, IP19 8AB, UK. 6Birdlife International, Pacific Secretariat, Suva, Fiji.
7
Island Conservation, Kelowna, British Columbia, V6B 2M9, Canada.
1
3
Abstract Rat eradication is a highly effective tool for conserving biodiversity, but one that requires considerable planning
effort, a high level of precision during implementation and carries no guarantee of success. Overall, rates of success are
generally high but lower for tropical islands where most biodiversity is at risk. We completed a qualitative comparative
review on four successful and four unsuccessful tropical rat eradication projects to better understand the factors influencing
the success of tropical rat eradications and shed light on how the risk of future failures can be minimised. Observations
of juvenile rats surviving more than four weeks after bait application on two islands validate the previously considered
theoretical risk that unweaned rats can remain isolated from exposure to rodent bait for a period. Juvenile rats emerging
after bait was no longer readily available may have been the cause of some or all the project failures. The elevated
availability of natural resources (primarily fruiting or seeding plants) generated by rainfall prior to project implementation
(documented for three of the unsuccessful projects) may also have contributed to project failure by reducing the likelihood
that all rats would consume sufficient rodent bait or compounding other factors such as rodent breeding. Our analysis
highlights that rat eradication can be achieved on tropical islands but suggests that events that cannot be predicted with
certainty in some tropical regions can act individually or in concert to reduce the likelihood of project success. We
recommend research to determine the relative importance of these factors in the fate of future tropical projects and suggest
that existing practices be re-evaluated for tropical island rodent eradications.
Keywords: best practice, conservation, invasive, restoration, rodent
INTRODUCTION
Marine islands house an estimated 15–20% of
terrestrial biodiversity and are home to 61% of IUCN
Extinct species and 37% of IUCN Critically Endangered
species (B. Tershey unpubl. data). Invasive species have
been the most frequent cause of extinctions on islands and
the second leading cause of Critical Endangerment (B.
Tershey unpubl. data). Commensal rats (Rattus spp.) are
considered the most damaging group of invasive species
on islands because of their near global distribution and the
frequency with which they cause extinctions, extirpations
and ecosystem-level impacts (Towns, et al., 2006; Howald,
et al., 2007; Kurle, et al., 2008). Rats can be eradicated
from islands (Keitt, et al., 2011) resulting in significant
species and ecosystem recovery (Bellingham, et al., 2010).
Thus, rat eradication is a powerful tool with which to
prevent extinctions.
Although this tool has been widely deployed, with
more than 500 successful rat eradications to date (DIISE,
2017), most rat eradications have been on small, mid
to high latitude islands (Howald, et al., 2007) where
endemic species diversity is lower. If rat eradication is to
realise its full potential to prevent extinctions, then future
eradications need to be more frequently conducted where
endemic species diversity is high: on larger tropical islands
(Kier, et al., 2009). However, while rat eradication is being
successfully conducted on increasingly large, high latitude
islands, with a failure rate of less than 3% (Russell &
Holmes, 2015), success on both large and small tropical
islands has been more elusive, with a failure rate of 10%
and very little understanding as to the underlying causes of
failure (Holmes, et al., 2015; Keitt, et al., 2015).
In an attempt to better understand the mechanisms
responsible for eradication failure on tropical islands and
improve the rate of success of future projects, a global
review of rodent eradication practice on tropical islands
was instigated (Russell & Holmes, 2015). In support of
the review, Holmes, et al. (2015) performed a statistical
analysis on as many rat eradication attempts as possible to
determine correlative factors that might pinpoint important
influences on tropical rat eradication success. However,
rat eradication projects are complex and multifaceted
(Cromarty, et al., 2002) and, like complex projects within
other disciplines, it can be challenging to determine the
reason(s) for project failure. To reduce the risk that the
broad-brush approach utilised by Holmes, et al. (2015)
overlooked important and influential factors, we completed
a second review, this time using a qualitative framework on
a subset of the projects assessed by Holmes, et al. (2015).
Qualitative comparative reviews are used extensively
in the social and behavioural sciences (e.g. Ragin, 1989;
George & Bennett, 2005; Bennett & Elman, 2006), but also
in other fields such as software engineering (Abrahamsson,
et al., 2003), human resource management (e.g. Allen, et al.,
1997), and political science (e.g. Bennett & Elman, 2006).
A qualitative comparative review offers the opportunity
to compare projects and their nuances in detail, which
superficially, statistical analyses cannot do, but also allows
for the possibility for making generalisations if they exist
(Ragin, 1989). This approach, which we believe has greater
utility in conservation biology, offered a complementary
mechanism for verifying or dispelling the importance of
factors identified as significant or insignificant in Holmes,
et al. (2015).
We examined in depth, reported data from eight
well-planned and sufficiently resourced tropical rat
eradication attempts, balanced among four successful and
four unsuccessful projects, to better understand: 1) the
variability in factors influencing tropical rat eradication
projects irrespective of outcome, 2) the factors that
consistently differentiate successful from failed tropical
rat eradication attempts for projects where full reported
data are available, 3) what steps can be taken to improve
eradication reporting and minimise the risk of failure for
future tropical rat eradications.
In:120
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 120–130. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Griffiths, et al.: Success & failure, rats on tropical islands
METHODS
Island eradication study sites
For the purposes of this study we focused on rat
eradication projects that used the method shown to have
the greatest chance of success, but which faced all of the
challenges associated with tropical islands and described
by Keitt, et al. (2015). We did not consider geographical
location to be important if these conditions were met.
Projects that met the following criteria were selected for
our analysis:
● Rodent bait was applied by helicopter, guided by
GPS. Projects that used the aerial application of bait
were the focus for our study because this method
has the best record of success in both temperate and
tropical climates (Howald, et al., 2007).
● The project was undertaken on a tropical island or
islands. Although Henderson lies just south of the
tropic of Capricorn at a latitude of 24°21’S, we
considered this island to be tropical in the context
of rodent eradication due to the island’s temperature
range, vegetation and absence of pronounced
seasonality (Spencer, 1995; Brooke, et al., 1996).
● The project was undertaken on an island or islands
with a Precipitation Coefficient of Variance (CV)
of mean monthly rainfall of less than 50% (Fig.
1). We focused our analysis not on particularly
wet or dry islands, but on islands where rainfall
and ecosystem productivity were more difficult to
predict. We excluded projects completed on arid or
semi-arid islands such as along the Pacific Coast
of Mexico or North-western Australia because, for
rodent eradication, these islands share the seasonality
associated with temperate islands i.e. an eradication
operation can be undertaken when natural food
resources are scarce and breeding, within the rat
population, is less likely. The island of Banco
Chinchorro, Mexico was excluded from our analysis
because it had a rainfall CV greater than 50%.
Nevertheless, Banco Chinchorro is another well
documented project and could have been a useful
addition to our comparative review.
● The project was undertaken on an island or islands
with land crabs. The presence of land crabs was
identified as a significant influence on project success
in Holmes, et al. (2015).
● Projects where reinvasion could be dismissed as an
unlikely cause of failure. Projects were only included
if reinvasion had been ruled out through comparative
DNA analysis or were undertaken on uninhabited
islands that were rarely visited and extremely remote.
This excluded islands such as Denis and Curieuse in
the Seychelles (Merton, et al., 2002) and the Aleipata
Islands in Samoa (Butler, et al., 2011).
● Sufficiently detailed information was available
to allow the project to be reviewed within the
framework recommended by Keitt, et al. (2015).
Of the 17 discrete projects completed on tropical
islands that applied rodent bait containing a second
generation anticoagulant by helicopter, eight were selected
for analysis. Six were completed on islands located in the
tropical Pacific; Henderson (part of the United Kingdom
Overseas Territory of Pitcairn), Wake (an unincorporated
territory of the United States north of the Marshall Islands),
Palmyra (an unincorporated territory of the United States
in the Northern Line Islands), Enderbury and Birnie (part
of the Phoenix Islands Group of the Republic of Kiribati)
and the Ringgolds (part of Fiji). Two projects were
located outside of the Pacific Region; Desecheo (Puerto
Rico Archipelago) located in the Caribbean and Frégate
(Seychelles) in the Indian Ocean.
Island size varied from 49 to 4,310 ha (Table 1) and all
islands experienced relatively similar temperature ranges
and annual rainfall (Table 1). Except for the Wake project
that targeted Pacific rat (R. exulans) and Asian house rat
(R. tanezumi), the eradication operations targeted the
removal of just one species. R. exulans was targeted in
four operations, ship rat (R. rattus) in two and Norway rat
(R. norvegicus) in one (Table 1). Holmes, et al. (Holmes,
et al., 2015) found no significant difference in eradication
success between rat species for projects that applied bait
aerially. Four of the islands were inhabited; Wake, Frégate,
Palmyra and the Ringgolds (Table 1).
Determining success and failure
In line with best practice guidelines produced by the
New Zealand Department of Conservation (Broome,
et al., 2011), we considered an eradication project to be
successful where the absence of rats was determined after
a minimum of two breeding seasons (at least one year)
after the completion of the operation, as rat populations
may remain low and undetected for shorter periods. Rats
were first reported as being present five months after the
operation on Wake Island; eight months after on Henderson
Island; 13 months after on Desecheo; and two years after
on Enderbury. At the time of writing 14, six, four and
three years have passed for the Frégate, Ringgolds, Birnie
and Palmyra projects, respectively, and all four islands
remain rat free. A failed attempt to eradicate rats from
Palmyra Atoll in 2001 was hampered by both technical
and implementation constraints and was not evaluated
(USFWS, 2011).
Identifying potential factors that influenced success
and failure.
Fig. 1 Monthly Precipitation Coefficient of Variance (CV)
for tropical islands where rodent eradications have been
attempted using rodent bait containing a 2nd generation
anticoagulant applied by helicopter.
While there are other alternate or contributing
hypotheses (Table 2; Holmes, et al., 2015), the most
proximate reason for the reduced rate of success for tropical
rodent eradications is likely to be that not all rats consumed
a lethal dose of brodifacoum, the rodenticide used in most
rat eradications (Howald, et al., 2007) either because they
did not have access to sufficient bait or because they did
not consume bait that was available (Holmes, et al., 2015).
We used the framework outlined in Keitt, et al. (2015) to
review the four unsuccessful projects. To determine if some
individuals within the rat population could not eat a lethal
dose of bait, we reviewed operational design, operational
procedures, GIS maps of bait coverage, baiting density,
121
122
Grass, low
growing scrub,
and Casuarina
forest.
R. exulans and
R. tanezumi
100–200
6.4
Coral atoll
0
R. rattus
213
Small mountainous
island of volcanic
origin
Grass and dry
tropical Bursera
forest
24–32
40
750–1,039
116
0
R. exulans
Grass and
low growing
scrub
6.7
Coral atoll
24–30
40
750–1,300
608
3°8’S
Failed
Failed
18°23’N
Enderbury
Desecheo
0
R. exulans
Grass and
low growing
scrub
4
Coral atoll
24–30
38
750–1,301
49
3°35’S
Succeeded
Birnie
12
R. rattus
Pisonia and
Cocos forest
1.8
Coral atoll
24–27
19
4,422
235 (98)
5°53’S
Succeeded
Palmyra
The Wake project successfully removed Rattus tanezumi but not R. exulans from Wake Atoll and successfully removed both species from part of the atoll, Peale Island (95 ha).
a
0
R. exulans
Tangled, scrub
and scrub-forest
Principal vegetation types
Target species for
eradication
Permanent human
population
33.5
Raised coral atoll
Maximum elevation (m)
Principal landform
49
10
24–28
1,780
1,623
24–30
696 (602)
4,310
Temperature Range (°C)
19°18’N
24°21’S
Latitude
Area (area of largest subunit of land) (ha)
Mean annual precipitation
(mm)
Coefficient of variance for
monthly precipitation (%)
Failed
Failed
a
Wake
Project outcome
Henderson
Table 1 Characteristics of the islands where the eight projects were undertaken.
2
R. exulans
Low scrub and
Cocos forest
12
Coral atoll
22–28
39
2,467
266 (147)
16°30’S
Succeeded
Ringgolds
20–30
R. norvegicus
125
Granite island
with two coastal
plateaux
Modified forest
24–32
49
2,182
219
4°35’S
Frégate
Succeeded
Island invasives: scaling up to meet the challenge. Ch 1B Rodents: Review
Griffiths, et al.: Success & failure, rats on tropical islands
Table 2 Hypotheses to explain increased failure of rat eradications on tropical islands.
Proximate Underlying cause
Possible response to increase success rates
cause
Some individuals within the island’s rat population could not eat a lethal dose of bait
Land crabs or other species
Higher bait application rates
consume bait
Additional bait applications
Bait at a time when competitors are at lower density or less active
Rats have small home ranges
Higher bait application rates
Flexible scheduling to apply bait when food supply low
Bait decomposes rapidly
More preservatives in bait
Additional bait applications
Lactating females or young in nest Bait available longer (more bait, additional applications)
when bait available
Flexible scheduling to drop bait when breeding is reduced or nonexistent.
Rats don’t leave human dwellings Comprehensively bait entire island including within commensal
areas
Some individuals within the island’s rat population would not consume a lethal dose of bait
Bait biodegrades rapidly
More wax or preservatives in bait
Additional bait applications
Abundant natural food
Multiple bait formulations
Bait available longer (more bait, additional applications)
Flexible scheduling to drop bait when food supply low
Individual foraging preferences
Multiple bait formulations
Bait available longer (more bait, additional applications)
Lactating females very neophobic Bait available longer (more bait, additional applications)
Different dietary preferences of
Multiple bait formulations
lactating females
Bait available longer (more bait, additional applications)
Flexible scheduling to drop bait when food supply low
Poor quality planning and implementation
Lack of capability
More training & collaboration
Appointment of experienced staff
Adequate resourcing
Peer review during the planning process
Lax regulatory requirements
Plan & implement using internationally recognised standards
Insufficient resourcing
Source more funding
Increase collaboration
Higher rate of reinvasion
Warm water allows increased
Select more isolated islands
swimming distances
Human use characteristics
Better biosecurity
Incorporate human use into island selection criteria
bait availability over time, timing between applications,
and any operational difficulties noted. The statistical
approach of Holmes, et al. (2015) could not address all of
these issues because of the scarcity of well documented
projects such as those we investigated. We also assessed
bait toxicity and the chance that rats were resistant or
tolerant to anticoagulants. Insufficient information was
available to evaluate the impact of any spatial variation in
land crab density across each of the islands.
To evaluate if some individuals within the rat
population would not eat a lethal dose of bait, we looked
at the operational design, the bait type, data from trials
completed, the environmental conditions present at the
time of the eradication and any observations made during
implementation. Evidence for and against each factor
was evaluated and used to form an opinion on its relative
importance to the project’s outcome. Evidence for the
existence of a similar or different set of conditions for the
successful projects was used to inform this analysis.
Not all projects monitored bait availability over
time and for those projects that did, different methods
were used, making it difficult to compare how long bait
remained available to rats after its application. To compare
between projects we used both the minimum period of time
that bait was available in all plots or transects sampled
and, where data were available, the lower limit of 99%
CI of the T-Statistic for bait availability four days after
its application as recommended by Pott, et al. (2015). For
those islands where no monitoring was undertaken we used
anecdotal reports to provide an estimate of the minimum
period of bait availability.
123
Island invasives: scaling up to meet the challenge. Ch 1B Rodents: Review
Comparison among projects
We undertook a qualitative comparative review
because the number of projects that formed the basis of our
assessment was small, there was inconsistency between
projects in the data collected and the methods by which
data were obtained. A qualitative comparative review
allows for generalisations to be made among cases and
we considered it the best option for this study. Akin to
Abrahamsson, et al. (2003), we cross-examined all projects
to identify factors common to successful or unsuccessful
projects. To inform this cross examination we drew from
Holmes, et al. (2015) and our cumulative experience to
identify a set of environmental variables and components
of operational design we considered to be important to
the success of rat eradication operations. These variables
are listed in Tables 1–3. Information on each project was
obtained from documentation prepared prior to and after
project implementation and from personal communications
with project team members.
RESULTS
Identifying causes of operational failure
Some individuals within the island’s rat population could
not eat a lethal dose of bait
The design of each of the four unsuccessful eradications,
encompassing aerial application, overlapping aerial bait
swaths, application rates comparatively higher than those
applied in temperate regions and a minimum of two
applications (Table 4), should have ensured comprehensive
coverage of the islands with rodent bait. During the first
bait application on Desecheo, some technical difficulties
resulted in several small areas of the island (the largest
being ~0.8 ha in size) receiving bait at less than the planned
application rate. These issues were remedied for the second
application when a more even spread of bait was achieved
and, between both applications, comprehensive coverage
of the island was achieved. Similarly, with the exception
of areas deliberately excluded from bait application such
as the sealed runway on Wake, we could not discern any
biologically significant gaps in bait distribution from a
review of the GIS data accumulated for any of the four
unsuccessful projects. A biological gap was defined for our
analysis as a gap greater than 0.015 ha in area. This was the
smallest home range size reported in the literature for any
of the four rat species targeted (Wirtz, 1972; King, 1990;
Shiels, 2010; Low, et al. 2013).
On this basis we conclude that the operational strategy
employed on Henderson, Desecheo and Enderbury likely
ensured that all foraging rats encountered rodent bait.
Although not identified from GIS maps of bait spread, it
was more difficult to reach the same conclusion for Wake
because of the more complex operational strategy (multiple
methods of bait application) employed there (Griffiths, et
al., 2014). The existence of interspecific competition, not
a factor for the other islands, also likely limited access
to bait for some individual rats. However, the successful
eradication of R. tanezumi, formerly widespread across
the atoll (Griffiths, et al., 2014), demonstrated that broad
coverage across all habitats was achieved.
All four projects had factored bait consumption by
non-target species such as land crabs into operational
decisions on application rates (Table 4). However, bait
disappeared more rapidly than anticipated from some
transects monitored on Wake and Desecheo (Brown, et
al., 2013; Brown & Tershy, 2013) (Table 4). Bait persisted
in all transects monitored on Henderson until close to
the end of the 30-day monitoring period (Brooke, et al.,
2011). However, as described by Pott, et al. (2015), a
124
different monitoring method was used and, because of
the inaccessible nature of the island, monitoring was
confined to a small part of the island. No monitoring of
bait availability was undertaken on Enderbury but ad hoc
observations suggest that rodent bait was broadly available
for at least the first five days after its initial application
(Pierce & Kerr, 2013).
Rat pups yet to emerge from the nest may not have
had immediate access to bait. Evidence of rat breeding
activity was documented on all four islands at the time of
implementation (Brooke, et al., 2011; Brown, et al., 2013;
Brown & Tershy, 2013; Pierce & Kerr, 2013). A rat of
indeterminate age was sighted and captured on Desecheo,
23 days after the first bait application. On Wake, a juvenile
R. exulans was found inside a bait station 18 days after
bait was first applied and a second juvenile R. exulans was
caught alive at the base of a coconut (Cocos nucifera) palm
after 47 days. A low body weight and large head relative
to body size indicated the latter individual had suffered
from malnutrition likely because of having been weaned
prematurely. As evidenced by liver assay, it had been
exposed to brodifacoum (Griffiths, et al., 2014). No live rats
were seen by project team members monitoring Henderson
rails (Porzana atra) at the north-east end of Henderson
beyond five days after the initial bait application, despite
being on the island for more than three months after the
operation. However, two very small, freshly dead, likely
juvenile, rats were discovered 11 and 14 days after bait was
applied suggesting these animals had survived for 10–13
days after the initial bait application.
Operational procedures were in some instances modified
during project implementation due to environmental
and physical factors encountered during the operation
and/or the detection of a small number of rats after bait
application. Lack of accurate geographical data led to an
underestimate of island size for Henderson during project
planning. As a consequence, the application rate for the
second application across the island’s plateau had to be
reduced from 7 kg/ha to 6 kg/ha (Torr & Brown, 2012).
Methods for applying bait to vegetated intertidal habitats
were modified during implementation on Wake (Griffiths,
et al., 2014). Bait stations were also deployed and bait
was hand spread at several sites on Wake to target rats
detected within five months of bait application, although
such efforts were eventually abandoned after increasing
numbers of rats sighted confirmed the eradication had been
unsuccessful for R. exulans (Griffiths, et al., 2014). We do
not consider the operational changes made for these three
projects to have reduced the availability of bait to rats.
No significant changes to the operational strategy were
reported for the Enderbury project and bait application,
as described by team members, followed the prescription
outlined within the project’s operational plan.
Based on the evidence available, we conclude that some
individuals within the island’s rat populations could not eat
a lethal dose of bait. Unweaned rats present at the time of
bait application did not have immediate access to bait and,
as evidenced by individuals surviving for so long after bait
application on Wake, this is also likely for some breeding
female rats. However, we cannot conclude that this factor
was the only cause of failure for the four failed projects.
Bait toxicity
Assays of samples of the rodent bait applied on
Henderson (mean brodifacoum concentrations of 16.4 ppm),
Wake (28.3 ppm) and Desecheo (29.3 ppm) confirmed that
bait toxicity was within normal tolerances (Brown, et al.,
2013; Brown & Tershy, 2013; RSPB, unpublished data).
Inadequate bait toxicity is unlikely to have been a factor
on Enderbury because the bait used there was produced
Griffiths, et al.: Success & failure, rats on tropical islands
at the same time as the bait used for the successful Birnie
operation. Mortality associated with the operation and a
rapid decline in rat numbers was also observed at all sites.
All three bait types used are produced via an industrial
production process with quality assurance checks in place
to ensure appropriate rodenticide concentrations prior
to shipping and all have been used successfully on both
temperate and tropical islands. Based on the evidence
available we conclude that inadequate bait toxicity was
not a factor in the failure of the four unsuccessful projects
reviewed.
Resistance
There were no indications to suggest rats on Henderson,
Wake, Enderbury and Desecheo were resistant or tolerant
to anticoagulants. Rats on Henderson, Enderbury and
Desecheo had no prior exposure to anticoagulants so
there was no selection pressure for pharmacodynamic
resistance involving mutations in the Vkorc1 gene. For
Henderson, subsequent testing of rats from the surviving
population confirmed the lack of any genetic basis for
resistance to brodifacoum (RSPB, unpubl. data). Although
anticoagulants were used on Wake prior to the eradication
(Mosher, et al., 2008) available evidence, as discussed
in Griffiths, et al. (2014), did not support resistance as a
factor in the project’s outcome. Most importantly, although
increased tolerance to brodifacoum has been documented
for some rat populations, ‘practical’ resistance, as defined
by Buckle & Prescott (2012), that might have caused the
Wake project to fail, has never been encountered, even at
sites where anticoagulants have been used repeatedly for
long periods of time (Lund, 1984; Bailey, et al., 2005).
It is unknown if any plant species present on Henderson,
Wake, Desecheo and Enderbury contained elevated levels
of vitamin K, but dietary-based resistance is not considered
a major mechanism of resistance elsewhere (Buckle &
Prescott, 2012). Based on the lack of evidence for resistance
or increased tolerance to anticoagulants we conclude that
this mechanism was not a factor in the recorded failures.
Some individuals within the island’s rat population would
not consume a lethal dose of bait
All four of the unsuccessful projects used proven bait
types (Table 4) that have achieved rat eradication on other
tropical islands. In addition, palatability of two of the bait
types was proven by bait exposure trials undertaken on
Henderson and Desecheo that showed, through use of a
biomarker, 100% acceptance by trapped rats (Swinnerton
& McKown, 2009; Brooke, et al., 2010). On Wake,
concerns about behavioural resistance were generated
after some rats in a two-choice laboratory trial undertaken
on the island (Mosher, et al., 2008) were documented
not eating rodent bait. Three R. exulans also avoided
exposure during an in situ biomarker trial (Wegmann, et
al., 2009). However, as outlined by Griffiths et al. (2014),
the successful elimination of R. tanezumi from the atoll,
the complete removal of R. exulans from a discrete part
of the atoll (Peale Island), and the marked reduction of R.
exulans for a period of time, are not consistent with a bait
shy rat population. No pre-eradication trials to assess bait
palatability were undertaken on Enderbury.
Some evidence for neophobia or rats preferring
alternative foods over rodent bait was seen at the time of
bait application for Enderbury and Wake. On the first night
after the initial application of bait on Enderbury, rats were
observed walking past rodent bait, despite it being readily
available, to forage on the flowers and fruits of Tribulus
cistoides on the island (Pierce & Kerr, 2013). Observations
of rats foraging on natural foods in the presence of bait
were also made on Wake (Griffiths, et al., 2014). However,
it is unknown if such observations are unusual or should be
considered the norm for rodent eradications, because of a
lack of information.
Relative to previous site visits, signs of elevated
resource availability were observed on Henderson and
Enderbury islands (Cuthbert, 2012; Pierce & Kerr, 2013)
at the time of project implementation. Rainfall leading up
to the operations is presumed to have led to this increase
(Cuthbert, 2012; Pierce & Kerr, 2013). On Henderson,
three plant species, Cyclophyllum barbatum, Myrsine
hosakae and Eugenia reinwardtiana were observed with
more fruit than seen in previous years and the presence of
a large number of recently fledged fruit doves (Ptilinopus
insularis) indicated that a large fruiting event had occurred
shortly prior to the operation (Cuthbert, 2012). On
Enderbury, 10 of the 11 common plant species present were
recorded as either flowering or fruiting at the time of the
operation including the four dominant plants T. cistoides,
Portulaca lutea, Boerhavia albiflora and Sida fallax. Higher
than average rainfall prior to the unsuccessful Desecheo
eradication (as evidenced by mainland weather records)
may have also generated increased food availability
there (Brown & Tershy, 2013). It is unknown if resources
on Wake were elevated at the time of the operation, but
abundant seed observed on Casuarina trees growing across
the island at the time of the operation and high numbers of
rats observed at the time of the operation correspond with
this possibility.
Based on available evidence we cannot reach a definite
conclusion on the role of this factor in the outcome
observed in the four unsuccessful projects. However, the
elevated availability of alternative resources may have
compounded other factors such as rat breeding to influence
project outcome.
Comparison among all eight projects
We could not separate unsuccessful projects from
successful projects based on geographic location, habitat
or standard climatic variables (Table 1). However, three
of the unsuccessful projects were undertaken on islands
significantly larger than those that were successful. Rats
were also successfully removed from the smaller of the
two disconnected land masses that comprise the Wake
Atoll complex (Griffiths, et al., 2014). Commensal
issues associated with the presence of a resident human
population, a known risk factor for rodent eradications
(Oppel, et al., 2011), were a significant component of the
Wake project but were also present, albeit on a smaller
scale, on three of the islands where rats were successfully
removed suggesting these issues were not insurmountable.
Similarly, more parallels than differences were evident
between successful and unsuccessful projects for the
environmental variables identified by Holmes, et al. (2015)
and ourselves as important to eradication success (Table
3). Elevated rainfall preceding the eradication operation
differentiated three of the unsuccessful projects, Desecheo,
Henderson and Enderbury. However, abundant natural
food resources, as observed on Henderson, Enderbury,
Desecheo and Wake at the time of project implementation,
were also observed on Palmyra, the Ringgolds and Frégate
where rats were successfully removed. Fruiting Pandanus
tectorius, coconut and nesting sooty terns (Onychoprion
fuscatus) on Palmyra, Terminalia littoralis fruit and
coconut on the Ringgolds and coconut, multiple fruiting
tree species, breeding seabirds, kitchen refuse, cultivated
crops and food for livestock on Frégate all offered
plentiful resources to rats. However, the level of natural
food availability during project implementation relative
to other times of the year for these islands is unknown.
An abundance of natural resources was not documented
125
Island invasives: scaling up to meet the challenge. Ch 1B Rodents: Review
on Birnie, where rats were successfully removed. Little
flowering or fruiting by the four common plant species
that are present was noted on this island at the time of the
implementation (Pierce & Kerr, 2013).
Land crabs were an influential factor on all eight
islands. Bait availability data provided some indication
of their relative impact on each of the operations but,
in the absence of crab survey data for each island, an
independent assessment of relative crab population density
among islands was not possible. Such data would have
provided a clearer picture of the relative impact of land
crabs on project success. Anecdotal observations suggest
that rat numbers were high on all eight islands at the
time of project implementation, but relative population
densities were once again unknown. Reproduction was not
investigated on the Ringgolds, but evidence indicates that
rats were breeding at the time of the eradication at the other
sites. On Palmyra, where rats were successfully removed,
a juvenile rat was sighted and captured 28 days after the
initial bait application within the island’s commensal area
where bait stations were being maintained. This individual
was near death and an assay of its liver confirmed
exposure to brodifacoum. Like the second of the two
juveniles discovered on Wake after bait application, this
rat also appeared malnourished. It is possible, based on
observations of elevated rainfall and increased resource
availability, that the intensity of rat breeding was higher
on Henderson, Enderbury and possibly Desecheo than
on the islands where rats were successfully removed but
in the absence of data this cannot be confirmed. Two of
the successful projects targeted rat populations that had
previously been exposed to anticoagulants (Table 3). Rats
on Palmyra, where anticoagulants had been used previously,
were thought to be tolerant to brodifacoum because some
individuals survived for longer than anticipated during a
toxicity trial (Howald, et al., 2004), yet this project was
successful.
Details for each of the eight eradication operations
are presented in Table 4. All projects used a helicopter
and bait spreading bucket as the principal method for bait
application, utilised proven rodent bait types and applied
bait with a similar swath overlap. The main difference
between operations was in the amount of bait applied,
which ranged between 10 and 84 kg/ha for the first
application and between 6 and 79 kg/ha for the second.
Difference in application rate was largely a function of
decisions made by respective project teams based on an
assessment of relative bait competition by land crabs for
each island. While this difference was evident, there was no
clear relationship between application rate and success or
failure for the eight projects (Table 4). Relative to the three
unsuccessful projects where monitoring of bait availability
was undertaken, bait on Palmyra also disappeared rapidly
but remained at higher densities beyond the seven-day
observation period in coconut canopy (Berentsen, et
al., 2013), a preferred habitat for rats (Wegmann, 2008).
Bait persisted in all plots monitored on Frégate for 10
days after its application and bait availability would have
been extended by the third application (Merton, et al.,
2002) but this was not monitored. No monitoring of bait
was undertaken on Birnie or the Ringgolds, but bait was
reported to be widely available on both islands for the six
days between the first and second applications of bait.
As with two of the failed projects, operational procedures
were also modified during project implementation for
two successful projects. For instance, an unplanned third
application of bait was completed following the sighting of
Henderson
Wake
Desecheo
Enderbury
Birnie
Palmyra
Ringgolds
Frégate
Table 3 Environmental variables present at the time of project implementation that could have influenced the project’s
outcome.
Project outcome
Failed
Failed
Failed
Failed
Succeeded
Succeeded
Succeeded
Succeeded
Hermit crabs present
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Other land crab species present
No
No
Yes
No
No
Yes
Yes
Yes
Ant species present
Yes
Yes
Yes
No
No
Yes
Yes
Yes
Permanent human population
present
Rat population had been
previously exposed to
anticoagulants
Higher than anticipated rainfall
preceded operation
Observations of high natural food
availability immediately prior to
or during project implementation
Seabirds nesting at time of
implementation
Rat population breeding at time of
project implementation
No
Yes
No
No
No
Yes
Yes
Yes
No
Yes
No
No
No
Yes
No
Yes
Yes
No
Yes
Yesa
Unknown
No
No
No
Yes
Yes
Yes
Yes
No
Yes
Yes
Yes
Yes
No
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Unknown Yes
Inferred from observations of flowering and fruiting during project implementation.
a
126
Griffiths, et al.: Success & failure, rats on tropical islands
DISCUSSION
a surviving rat on Frégate Island (P. Garden, pers. comm.).
On Palmyra, bait was hand broadcast across a 10 ha area
on Cooper Island following the discovery of the juvenile
rat mentioned above (Wegmann, et al., 2012). No changes
to the operational strategy were reported for the Ringgolds
and Birnie projects and, as with the Enderbury project, bait
application proceeded according to plan.
Reasons for project failure
Based on the robust design of the eradication operations
reviewed and GIS maps of bait coverage, we conclude
that bait was made available to all rats actively foraging
at the time of the operation for the Henderson, Enderbury
and Desecheo projects. We cannot be as confident of
this for Wake, despite one rat species being successfully
eradicated, because the more complex operational strategy
employed there coupled with competitive exclusion may
have led to functional gaps in bait availability (Griffiths, et
al., 2014). Notwithstanding the greater risk on Wake, some
individuals within the rat population were not actively
foraging at the time of bait application on all four islands
From our qualitative comparative analysis, we could
not reach a conclusion on the role of geographic, habitat,
climatic and environmental variables or operational
parameters on the relative outcome of the eight projects
reviewed. The two variables that best differentiated
unsuccessful from successful projects were elevated
rainfall preceding the operation and island size.
Failed
Failed Failed
Failed
Succeeded
Succeeded
Succeeded Succeeded
Bait typea
20R
25W
25D
20R
20R
25W
20R
20R
10/6c
18/9
19/10
22/17
25/25
84/79
16/11
14/9/12
17.4
27.7
29
38.4
50
165
27
35
50/25
50/50
50/50
50/25
50/25
50/50
50/50
50/50/50
~270 m2 25 m2
25 m2
NA
NA
2.49 m2
NA
10 m2
25+
3
2
6d
6d
1e
10d
10f
20+
5
1
Unknown Unknown
1d
Unknown
5e
5
9
10
5
6
10
5/24
1.93
6.33
0.25
Unknown Unknown
19.16
Unknown
-3.32
No
Yes
No
No
Yes
No
Yes
Application rate 1st
/2nd/3rd bait applications
(kg/ha)b
Mean total bait
application rate (kg/ha)
Percentage swath
overlap per application
Area of plot/transect
used to sample bait
availability
Number of days that bait
remained available in all
sampled plots/transects
after 1st application
Number of days that bait
remained available in all
sampled plots/transects
after 2nd application
Number of days between
applications
Lower 99% CI of the
T-statistic for bait
available four days after
the 1st application
(kg/ha)
Areas excluded from
aerial bait application
6
No
Frégate
Palmyra
Project outcome
Wake
Birnie
Ringgolds
Enderbury
Desecheo
Henderson
Table 4 Key elements of operational design for the eight projects.
a
Bait pellet types listed are 20R – Pestoff 20R rodent bait produced by Animal Control Products, Wanganui, New Zealand; 25W – Brodifacoum25W Conservation manufactured by Bell Laboratories, Wisconsin, USA; 25D – Brodifacoum-25D Conservation manufactured by Bell Laboratories,
Wisconsin, USA.
b
Areas subject to hand broadcast were applied at the same rates as for aerial application.
c
Rates listed here were used across the island’s plateau which amounted to 95% of the island’s area. Higher bait application rates were applied in the
vicinity of the island’s beaches where hermit crabs were most numerous.
d
No monitoring of bait availability was undertaken and figures are inferred from ad hoc observations. The project team left the islands after the
number of days listed.
e
The figure reported is for terrestrial plots: bait persisted longer in coconut palm canopy.
f
No monitoring was undertaken after the 3rd application which would have extended the number of days that bait was available.
127
Island invasives: scaling up to meet the challenge. Ch 1B Rodents: Review
where rats survived. Rats were breeding on Henderson,
Wake, Desecheo and Enderbury at the time of project
implementation and evidence suggests that brodifacoum is
not passed on in sufficient amounts via lactation to cause
mortality (Milne, et al., 2001; Gabriel, et al., 2012). Pups
in the nest at the time of bait application were therefore
effectively isolated for the period they were dependent on
the lactating female.
Such a scenario has been previously considered by
eradication practitioners as a theoretical possibility (e.g.
Broome, et al., 2011), but the discovery of juvenile rats on
both Palmyra and Wake after bait application validates it as
a very real concern for tropical island rodent eradications,
where breeding cycles cannot be predicted with certainty.
Weaning times reported for R. exulans (Wirtz, 1972; Tobin,
1994), R. rattus (Cowan, 1981; Yom-Tov, 1985) and R.
norvegicus (King, 1990) range from 21 to 28 days, much
longer than the period over which bait is typically available
for tropical rat eradication projects including a number of
the projects reviewed here.
It has generally been accepted that breeding females,
like other individuals within a rat population, would access
and ingest a lethal dose of bait and die within a few days of
bait application. However, there are reasons to be sceptical
that this will always occur. Home ranges for female rats
(e.g. R. rattus) can be significantly smaller than those of
males (Pryde, et al., 2005) and, as has been documented
for house mice (Mus musculus) (Krebs, et al., 1995),
lactating female rats may have constricted foraging ranges.
Changes in dietary requirements by rats can also occur
during lactation (Leshner, et al., 1972) potentially affecting
bait palatability. The maximum period of time documented
for mortality following the ingestion of a lethal dose
of brodifacoum is 21 days, from a trial conducted with
captive R. rattus on Palmyra (Howald, et al., 2004). Any of
these traits could increase the chance of juveniles emerging
after bait is no longer readily available on an island and,
with natural food abundant on many tropical islands, these
individuals have an enhanced probability of survival.
The fact that bait remained available in all transects
monitored on Henderson for more than 25 days challenges
the premise of juvenile survival as a potential cause of
failure for this project. However, as described by Pott, et al.
(2015), a different method of monitoring bait availability
was used for this project and monitoring was confined to
one small corner of the island (Brooke, et al., 2011) so
comparison with other projects is difficult. It is also possible
that bait disappeared more rapidly in unmonitored parts of
the island. Bait was applied at a lower rate on Henderson
than in the other projects reviewed and this, coupled
with the island’s complicated ‘makatea’ or uplifted coral
substrate, may have reduced the rate at which breeding
female rats encountered bait.
Rats were confirmed as breeding during project
implementation on Birnie, Palmyra and Frégate where
rats were successfully removed. Why did these projects
succeed? Some explanations can be tendered but, without
additional evidence, cannot be verified. For example, the
high bait application rate used on Palmyra likely ensured
that breeding female rats rapidly encountered bait plus
bait in the coconut palm canopy, a known nesting habitat
for female rats, was accessible for a longer period. On
Frégate, a third bait application extended the period of bait
availability out beyond 24 days and less competition by
hermit crabs and lower rat densities on Birnie may have
increased bait availability there. It is also plausible that
in the absence of the supplementary interventions made
on Palmyra and Frégate, these projects could also have
failed. Insufficient information is available to form similar
conclusions for the Ringgolds project.
128
We were able to rule out inadequate bait toxicity and
resistance as factors for the survival of rats on Henderson,
Enderbury and Desecheo and the persistence of R. exulans
on Wake. Neither has been documented for any of the 490
attempted higher latitude rat eradications and we know of
no viable hypothesis that would predict a greater incidence
of resistance in rats or insufficient bait toxicity for tropical
rat eradication projects. For the unsuccessful projects
we reviewed we reject bait toxicity as a factor based on:
factory test results demonstrating that the bait used on
Henderson, Wake and Desecheo contained a sufficient
concentration of brodifacoum; the marked reduction in rat
numbers on all three islands; and the fact that R. tanezumi
was successfully removed from Wake. The bait applied on
Enderbury was produced as part of the same consignment
as that was used successfully to remove rats from Birnie.
Similarly, we found no evidence to support anticoagulant
resistance as a factor in the unsuccessful outcome seen
on Henderson, Wake, Desecheo and Enderbury. Rat
populations on Henderson, Enderbury and Desecheo had
no prior exposure to anticoagulants and the successful
eradication of R. tanezumi from Wake, the removal of
R. exulans from part of the atoll, and the reduction of R.
exulans to undetectable levels elsewhere is at odds with
the levels of survivorship reported for rodent populations
in which practical resistance has been documented (e.g.
Drummond & Rennison, 1973; Greaves, et al., 1982).
Most importantly, ‘practical’ resistance to brodifacoum
that might have caused the failure of these projects, has
never been encountered, even at sites where anticoagulants
have been used repeatedly for long periods of time (Buckle
& Prescott, 2012). Increased tolerance to brodifacoum
has been detected in some locations (Buckle & Prescott,
2012) and may have been present on the three islands
where anticoagulants had been used previously. However,
rats were successfully removed from two of these islands
including Palmyra where a bait toxicity trial had suggested
the possibility of anticoagulant tolerance.
Conflicting evidence meant we could not rule out the
possibility that some rats avoided rodent bait in preference
for natural foods. Certainly, for all four unsuccessful
projects, natural food was readily available to rats at the time
of project implementation. Observations of rats foraging on
natural foods after bait application on Enderbury and Wake
lend weight to this hypothesis. However, this may simply
have been a function of neophobia, as described by Barnett
(1956), and not necessarily active bait avoidance. We are
unaware of similar observations from other projects, but
this is likely a result of insufficient observational effort.
The discovery of recently weaned juvenile rats on Palmyra
and Wake, more than four weeks after bait application,
suggests that some individuals, in this case lactating female
rats, may have avoided bait for a period. Rats detected on
Desecheo and Fregate after bait application also point to
this possibility. Set against this evidence is the fact that
natural food was also available on the islands where rats
were successfully removed, and signs of malnutrition
and early weaning of the juveniles found on Palmyra
and Wake suggest that the females producing these pups
died because they consumed bait. A necropsy verified bait
consumption for the Desecheo rat and the Frégate project
was ultimately successful, confirming all individuals there
were eventually exposed. The successful removal of the
more dominant rat species on Wake also perhaps points
to bait availability rather than bait palatability as the more
important influence.
In summary, it is unknown if the elevated availability
of natural resources on Henderson, Enderbury, Wake and
Desecheo led to bait avoidance, but the possibility cannot
be discounted. Increased natural food availability may
have also compounded other factors influencing project
Griffiths, et al.: Success & failure, rats on tropical islands
success such as the intensity of rat breeding. Given the
unpredictability of resource availability within many
tropical island ecosystems this will need to be an important
consideration for future rat eradication projects.
on whom we relied upon for information include the
Royal Society for the Protection of Birds, U.S. Air Force,
U.S. Fish and Wildlife Service, The Nature Conservancy,
EcoOceania Pty Ltd, Island Conservation, the New Zealand
Department of Conservation and Peter Garden.
Comparative analysis
We are also extremely grateful to the organisations
and individuals who supported the tropical island rodent
eradication workshop held in Auckland, New Zealand in
2013. The views expressed in this paper reflect the wider
experience and knowledge of workshop attendees. The
workshop was funded in part by the National Fish and
Wildlife Foundation and the David and Lucile Packard
Foundation. We are indebted to James Russell and Nick
Holmes for constructive criticism on the structure and
content of the manuscript.
We could not separate unsuccessful projects from
successful projects based on habitat or standard climatic
variables. However, three of the unsuccessful projects were
undertaken on islands significantly larger than those that
were successful and both rat species present on Wake were
removed from Peale Island, one of the two land units that
make up the Wake Atoll complex. This is consistent with
the trend identified by Holmes, et al. (2015) of an increasing
failure rate for larger islands. It is therefore possible that the
outcomes observed on Henderson, Wake and Enderbury
were simply a consequence of biogeographic theory.
Larger populations on the bigger islands increased the
chance that some individuals would avoid bait or that some
breeding females would survive for long enough to wean
juveniles when bait was no longer readily available. No
threshold for island size has yet been identified for rodent
eradications undertaken using the methodology reviewed
in this paper. However, the threshold may be smaller for
tropical islands because of increased availability of natural
resources, higher rat population densities and the likelihood
that a proportion of the population will be breeding during
project implementation.
Rainfall is closely linked to ecosystem productivity
on tropical islands (Murphy & Lugo, 1986) and elevated
rainfall levels preceding the eradication were associated
with three of the unsuccessful projects reviewed.
Variability in rainfall was also found by Holmes, et
al. (2015) to be correlated with failure for tropical rat
eradications. However, as discussed above, we could not
fully resolve whether rainfall contributed to an increased
risk of failure for these projects because palatability of
rodent bait was reduced in the presence of increased natural
food availability or greater reproductive activity within
the targeted rat populations led to juveniles surviving the
eradication attempt.
In summary, although our review of eight tropical
rodent eradications could not discern the relative
importance of bait availability or bait palatability in the
outcome of the four unsuccessful projects, it suggests that
both are important to consider in the planning of future
rodent eradications on tropical islands. In the absence of a
more palatable bait type, we recommend greater emphasis
is placed on operational design for future tropical island
rodent eradications. As recommended by Keitt, et al. (2015),
projects should aim to ensure that bait is readily available
within all rat territories for a period of time that allows all
individuals within the population to encounter bait. Even
though the projects we reviewed were well documented,
our analysis was limited by a lack of consistency in data
collection. Until more is known about the mechanisms
that promote survival during a rat eradication attempt,
future monitoring of eradication projects undertaken on
tropical islands should aim to document as many of the
variables discussed in this paper as possible to determine
the relative importance of these factors in the project’s fate.
Standardisation of monitoring protocols, as promoted by
Keitt, et al. (2015) and Pott, et al. (2015), should also be
instigated.
ACKNOWLEDGEMENTS
We would like to thank the organisations who supported,
advised or implemented the projects we reviewed and
who were willing to share the information on which this
publication was based. The organisations and individuals
All authors were involved in drafting the manuscript or
revising it critically for important intellectual content, and
all authors approved the final version to be published. RG
had full access to all the data used in the study and takes
responsibility for the accuracy of its tabulation. RG, AW,
WP, RC, DB, SC supplied information on which the paper
is based.
REFERENCES
Abrahamsson, P., Warsta, J., Siponen, M.T. and Ronkainen, J. (2003).
‘New Directions on Agile Methods: A Comparative Analysis’.
In: Proceedings of the 25th International Conference on Software
Engineering, pp. 244–254. Washington, DC, USA: IEEE Computer
Society.
Allen, T.D., Poteet, M.L. and Burroughs, S.M. (1997). ‘The mentor’s
perspective: A qualitative inquiry and future research agenda’. Journal
of Vocational Behavior 51(1): 70–89.
Bailey, C., Fisher, P. and Eason, C.T. (2005). ‘Assessing Anticoagulant
Resistance in Rats and Coagulation Effects in Birds Using Smallvolume Blood Samples’. Science for Conservation 249. Wellington,
New Zealand: Department of Conservation.
Barnett, S.A. (1956). ‘Behaviour components in the feeding of wild and
laboratory rats’. Behaviour 9: 24–43.
Bellingham, P.J., Towns, D.R., Cameron, E.K., Davis, J.J., Wardle, D.A.,
Wilmshurst, J.M. and Mulder, C.P.H. (2010). ‘New Zealand island
restoration: Seabirds, predators, and the importance of history’. New
Zealand Journal of Ecology 34(1): 115–136.
Bennett, A. and Elman, C. (2006). ‘Qualitative research: Recent
developments in case study methods’. Annual Review of Political
Science 9: 455–476.
Berentsen, A.R., Pitt, W.C., Eisemann, J.D. and Engeman, R.M. (2013).
‘Longevity of rodenticide bait pellets in a tropical environment
following a rat eradication program’. Environmental Science and
Pollution Research 21: 2283-2288.
Brooke, M.d.L., Jones, P.J., Vickery, J.A. and Waldren, S. (1996).
‘Seasonal patterns of leaf growth and loss, flowering and fruiting on a
sub-tropical central Pacific island.’. Biotropica 28: 164–179.
Brooke, M., Cuthbert, R., Henricson, A., Torr, N., Warren, P. and
O’Keefe, S. (2010). Towards Rat Eradication on Henderson Island:
Fieldwork Report, August–September 2009. Unpublished Report.
Sandy, Bedfordshire, UK: Royal Society for the Protection of Birds.
Brooke, M.d.L., Harrison, G. and Cuthbert, R.J. (2011). The Capture,
Husbandry and Captive Breeding of Henderson Rails During the
Henderson Island Rat Eradication and the Impact of the Project on
Native Landbirds. Unpublished Report. Sandy, Bedfordshire, UK:
Royal Society for the Protection of Birds.
Broome, K.G., Brown, D., Cox, A., Cromarty, P., McClelland, P., Golding,
C., Griffiths, R. and Bell, P. (2011). Current Agreed Best Practice
for Rodent Eradication – Aerial Broadcasting Poison Bait (Version
2.2). Unpublished Report. Wellington, New Zealand: Department of
Conservation.
Brown, D., Pitt, W.C. and Tershy, B.R. (2013). A Review of the Wake
Island Rat Eradication Project. Unpublished Report. Honolulu, Hawaii:
US Fish and Wildlife Service.
Brown, D. and Tershy, B.R. (2013). A Review of the Desecheo Island Rat
Eradication Project. Unpublished Report. Cabo Rojo, Puerto Rico: US
Fish and Wildlife Service.
Buckle, A. and Prescott, C. (2012). The Current Status of Anticoagulant
Resistance in Rats and Mice in the UK. Unpublished Report. Reading,
Berkshire, UK: Rodenticide Resistance Action Group of the United
Kingdom, Vertebrate Pests Unit, University of Reading.
129
Island invasives: scaling up to meet the challenge. Ch 1B Rodents: Review
Butler, D., Tye, A., Wylie, M. and Tipama’a, F. (2011). ‘Eradicating
Pacific rats (Rattus exulans) from Nu’utele and Nu’ulua Islands,
Samoa–Some of the Challenges of Operating in the Tropical Pacific’.
In: C.R. Veitch, M.N. Clout and D.R. Towns (eds.) Island invasives:
eradication and management, pp. 407–412. Occassional Paper SSC no.
42. Gland, Switzerland: IUCN and Auckland, New Zealand: CBB.
Cowan, P. (1981). ‘Early growth and development of roof rats, Rattus
rattus L.’. Mammalia 45(2): 239–250.
Cromarty, P.L., Broome, K.G., Cox, A., Empson, R.A., Hutchinson,
W.M. and McFadden, I. (2002). ‘Eradication planning for invasive
alien animal species on islands – the approach developed by the New
Zealand Department of Conservation’. In: C.R. Veitch and M.N. Clout
(eds.) Turning the tide: the eradication of invasive species, pp. 85–91.
Occasional Paper SSC no. 28. IUCN SSC Invasive Species Specialist
Group. IUCN, Gland, Switzerland and Cambridge, UK.
Cuthbert, R. (2012). Notes on Flowering and Evidence of Breeding
Birds on Henderson July–August 2011. Unpublished Report. Sandy,
Bedfordshire, UK: Royal Society for the Protection of Birds.
DIISE. (2017). ‘The Database of Island Invasive Species Eradications,
Developed by Island Conservation, Coastal Conservation Action
Laboratory UCSC, IUCN SSC Invasive Species Specialist Group.
University of Auckland and Landcare Research, New Zealand.’ <http://
diise.islandconservation.org>. Accessed 30 September 2017.
Drummond, D. and Rennison, B.D. (1973). ‘The detection of rodent
resistance to anticoagulants’. Bulletin of The World Health Organization
48(2): 239–242.
Gabriel, M.W., Woods, L.W., Poppenga, R., Sweitzer, R.A., Thompson,
C., Matthews, S.M., Higley, J.M., Keller, S.M., Purcell, K., Barrett, R.H.,
Wengert, G.M., Sacks, B.N. and Clifford, D.L. (2012). ‘Anticoagulant
rodenticides on our public and community lands: Spatial distribution
of exposure and poisoning of a rare forest carnivore.’. PLOS ONE 7:
e40163.
George, A.L. and Bennett, A. (2005). Case Studies and Theory
Development in the Social Sciences. Cambridge: Cambridge University
Press.
Greaves, J., Shepherd, D. and Gill, J. (1982). ‘An investigation of
difenacoum resistance in Norway rat populations in Hampshire’. Annals
of Applied Biology 100(3): 581–587.
Griffiths, R., Brown, D., Tershy, B., Pitt, W., Wegmann, A., Hanson, C.,
Moran, M., Rex, K., White, S., Flint, B., Keitt, B., Holmes, N., Howald,
G. and Torr, N. (2014). ‘The Wake Island Rodent Eradication – Part
Success, Part Failure, but Wholly Instructive’. In: R.M. Timm (ed.)
Proceedings of the 26th Vertebrate Pest Conference, pp. 101–111. Davis,
CA: University of California.
Holmes, N.D., Griffiths, R., Pott, M., Alifano, A., Will, D., Wegmann, A.S.
and Russell, J.C. (2015). ‘Factors associated with rodent eradication
failure’. Biological Conservation 185: 8–16.
Howald, G., Samaniego, A., Buckelew, S., McClelland, P., Keitt, B.,
Wegmann, A., Pitt, W.C., Vice, D.S., Campbell, E., Swift, K. and
Barclay, S. (2004). Palmyra Atoll Rat Eradication Assessment Trip
Report, August 2004. Unpublished Report. Santa Cruz, California,
USA: Island Conservation.
Howald, G., Donlan, C.J., Galván, J.P., Russell, J., Parkes, J., Samaniego,
A., Wang, Y., Veitch, D., Genovesi, P., Pascal, M., Saunders, A. and
Tershy, B. (2007). ‘Invasive rodent eradication on islands’. Conservation
Biology 21(5): 1258–1268.
Keitt, B., Campbell, K., Saunders, A., Clout, M., Wang, Y., Heinz,
R., Newton, K. and Tershy, B. (2011). ‘The global islands invasive
vertebrate eradication database: a tool to improve and facilitate
restoration of island ecosystems’. In: C.R. Veitch, M.N. Clout and
D.R. Towns (eds.) Island invasives: eradication and management, pp.
74–77. Occassional Paper SSC no. 42. Gland, Switzerland: IUCN and
Auckland, New Zealand: CBB.
Keitt, B., Griffiths, R., Boudjelas, S., Broome, K., Cranwell, S., Millett,
J., Pitt, W. and Samaniego-Herrera, A. (2015). ‘Best practice guidelines
for rat eradication on tropical islands’. Biological Conservation 185:
17–26.
Kier, G., Kreft, H., Lee, T.M., Jetz, W., Ibisch, P.L., Nowicki, C., Mutke,
J. and Barthlott, W. (2009). ‘A global assessment of endemism and
species richness across island and mainland regions’. Proceedings of
the National Academy of Sciences 106(23): 9322–9327.
King, C.M. (ed.) (1990). The Handbook of New Zealand Mammals.
Oxford: Oxford University Press.
Krebs, C.J., Kenney, A.J. and Singleton, G.R. (1995). ‘Movements of
feral house mice in agricultural landscapes’. Australian Journal of
Zoology 43(3): 293–302.
Kurle, C.M., Croll, D.A. and Tershy, B.R. (2008). ‘Introduced rats
indirectly change marine rocky intertidal communities from algae- to
invertebrate-dominated’. Proceedings of the National Academy of
Sciences 105(10): 3800–3804.
Leshner, A.I., Siegel, H.I. and Collier, G. (1972). ‘Dietary self-selection
by pregnant and lactating rats’. Physiology & Behavior 8(1): 151–154.
130
Lund, M. (1984). ‘Resistance to the Second-generation Anticoagulant
Rodenticides’. In: D.O. Clark (ed.) Proceedings of the 11th Vertebrate
Pest Conference, pp. 89–94. Davis, CA: University of California.
Merton, D., Climo, G., Laboudallon, V., Robert, S. and Mander, C. (2002).
‘Alien mammal eradication and quarantine on inhabited islands in the
Seychelles’. In: C.R. Veitch and M.N. Clout (eds.) Turning the tide:
the eradication of invasive species. pp. 182–198. Occasional Paper SSC
no. 28. IUCN SSC Invasive Species Specialist Group. IUCN, Gland,
Switzerland and Cambridge, UK.
Milne, L., Wright, G., Eason, C. and O’Connor, C. (2001). ‘Human
Exposure to Poisons Through Contaminated Food’. In: Vertebrate Pest
Committee (ed.) Proceedings of the 12th Australasian Vertebrate Pest
Conference, pp. 141–145. Melbourne, Australia: Department of Natural
Resources and Environment.
Mosher, S., Hebshi, A., Swift, K., Dunlevy, P., Vice, D., Wegmann,
A., Jacobs, B., Thomas, B. and Gilardi, J. (2008). Rat Eradication
Feasibility Study: 29 September–27 October 2007. Unpublished Report.
Hickam Air Force Base: Deptartment of Defense.
Murphy, P.G. and Lugo, A.E. (1986). ‘Ecology of tropical dry forest’.
Annual Review of Ecology and Systematics 17(1): 67–88.
Oppel, S., Beaven, B.M., Bolton, M., Vickery, J. and Bodey, T.W. (2011).
‘Eradication of invasive mammals on islands inhabited by humans and
domestic animals’. Conservation Biology 25(2): 232–240.
Pierce, R.J. and Kerr, V. (2013). Preliminary Report on a Biota Survey
of the Phoenix Islands Atolls, Kiribati, May 2013. Unpublished report.
Speewah, QLD, Australia: Eco Oceania Pty.
Pott, M., Wegmann, A., Griffiths, R., Samaniego-Herrera, A., Cuthbert,
R., Brooke, M., Pitt, W., Berentsen, A., Holmes, N., Howald, G. and
Russell, J.C. (2015). ‘Stacking the deck: Assessing bait availability
before rodent eradications to inform bait application rates’. Biological
Conservation 185: 27–35.
Pryde, M., Dilks, P. and Fraser, I. (2005). ‘The home range of ship rats
(Rattus rattus) in beech forest in the Eglinton Valley, Fiordland, New
Zealand: A pilot study’. New Zealand Journal of Zoology 32(3): 139–
142.
Ragin, C.C. (1989). The Comparative Method: Moving Beyond
Qualitative and Quantitative Strategies. Berkeley, CA: University of
California Press.
Russell, J.C. and Holmes, N.D. (2015). ‘Tropical island conservation: Rat
eradication for species recovery’. Biological Conservation 185: 1–7.
Shiels, A.B. (2010). ‘Ecology and Impacts of Introduced Rodents (Rattus
spp. and Mus musculus) in the Hawaiian Islands’. PhD thesis. Honolulu,
HI: The University of Hawaii.
Spencer, T. (1995). ‘The Pitcairn Islands, South Pacific Ocean: Plate
tectonic and climatic contexts’. Biological Journal of the Linnean
Society 56: 13–42.
Swinnerton, K. and McKown, M. (2009). Scoping Trip to Desecheo
Island, Puerto Rico, to Evaluate Feasibility of Rodent Eradication,
and Implement Baseline Biological Surveys: February 12–26, 2009.
Unpublished Report. Santa Cruz, California, USA: Island Conservation.
Tobin, M.E. (1994). ‘Polynesian Rats’. In: S.E. Hygnstrom, R.M. Timm
and G.E. Larson (eds.) Prevention and Control of Wildlife Damage, pp.
121–124. Lincoln, NE: University of Nebraska.
Torr, N. and Brown, D. (2012). Henderson Island Restoration Project
Post-Operational Report. Unpublished Report. Sandy, Bedfordshire,
UK: Royal Society for the Protection of Birds.
Towns, D., Atkinson, I. and Daugherty, C. (2006). ‘Have the harmful
effects of introduced rats on islands been exaggerated?’. Biological
Invasions 8: 863–891.
USFWS. (2011). Palmyra Atoll National Wildlife Refuge Rat Eradication
Project Final Environmental Impact Statement. Honolulu, HI: US Fish
& Wildlife Service.
Wegmann, A. (2008). ‘Home Range and Habitat Use Observations of
Introduced Rats (Rattus rattus) and Native Land Crabs (Coenobitidae
and Gecarcinidae) at Palmyra Atoll’. PhD thesis. Honolulu, HI:
University of Hawaii.
Wegmann, A., Helm, J., Torr, N., Golding, C. and Howald, G. (2009).
Wake Atoll Rat Eradcation Strategic Plan: Biomarker Study Trip Report
(27 February to 27 March, 2009). Unpublished Report. Kelowna,
Canada: Island Conservation.
Wegmann, A., Flint, E., White, S., Fox, M., Howald, G., McClelland, P.,
Alifano, A. and Griffiths, R. (2012). ‘Pushing the Envelope in Paradise:
A Novel Approach to Rat Eradication at Palmyra Atoll’. In: R.M. Timm
(ed.) Proceedings of the 25th Vertebrate Pest Conference, pp. 48–53.
Davis, CA: University of California.
Wirtz, W.O.I. (1972). ‘Population ecology of the Polynesian rat, Rattus
exulans, on Kure Atoll, Hawaii’. Pacific Science 26: 433–461.
Yom-Tov, Y. (1985). ‘The reproductive rates of Australian rodents’.
Oecologia 66(2): 250–255.
S. Horn, T. Greene and G. Elliott
Horn, S.; T. Greene and G. Elliott. Eradication of mice from Antipodes Island, New Zealand
Eradication of mice from Antipodes Island, New Zealand
S. Horn, T. Greene and G. Elliott
Department of Conservation, P.O. Box 743, Invercargill, New Zealand. <shorn@doc.govt.nz>
Abstract In winter 2016, the New Zealand Department of Conservation (DOC) eradicated mice (Mus musculus) from
the Antipodes Islands located at 49°S 178°E, 760 km south-east of New Zealand’s South Island. Mice were the only
mammalian pest species present. They have extensively impacted the abundance and survival of invertebrates, with
likely secondary impacts on endemic terrestrial birds and nesting seabird fauna. Public-private partnerships with DOC
instigated the project and provided essential financial support. Baseline scientific data for operational planning and
outcome monitoring were collected by a research expedition in July 2013 and project planning began in 2014. At the
time of writing, this is the largest eradication of mice undertaken where mice are the sole mammalian pest species.
Logistical challenges were complicated by a broad range of regulatory obligations. The expedition-style project used a
ship to deliver a team and equipment to Antipodes Island where they established camp and remained until the completion
of baiting. Bait spread was completed incrementally as weather allowed, comprehensively covering the islands in two
separate treatments between 18 June 2016 and 12 July 2016. The last sign of mice was detected 20 days after the first
application of bait and the eradication of mice was confirmed by monitoring in late summer 2018. Public engagement was
achieved with regular operational updates across multiple platforms and positive media coverage. Non-toxic bait trials
accurately predicted some by-kill of pipit (Anthus novaeseelandiae steindachneri) but did not anticipate poisoning of
some Antipodes parakeet (Cyanoramphus unicolor) and Reischek’s parakeet (Cyanoramphus hochstetteri). Known pestfree islands were not baited, providing refuge for land birds to mitigate the risk. Fledging success of Antipodean albatross
(Diomedea antipodensis antipodensis) chicks was not impacted by the operation and those species that were affected had
recovered by summer 2018.
Keywords: brodifacoum, house mouse, Million Dollar Mouse, Mus musculus, non-target impacts, subantarctic
INTRODUCTION
The Antipodes Islands group (2,100 ha) is in New
Zealand’s Subantarctic Islands region and was gazetted
as a Nature Reserve in 1978 and a World Heritage site
in 1998. The group comprises six islands and one islet
located in the Southern Ocean, at 49°41’S, 178°48’E, 760
km from New Zealand’s South Island (Fig. 1). The islands
are uninhabited and administered by New Zealand’s
Department of Conservation (DOC). House mouse (Mus
musculus) was the only mammalian pest species present
and known only on the main island, Antipodes Island
(2,012 ha).
The Antipodes were discovered in 1800 and sealers
arrived by 1804 (Taylor, 2006). A small shelter (castaway
depot) was built in 1886 to support shipwreck survivors.
It was resupplied periodically until 1927 (Taylor, 2006).
Mice were first recorded on Antipodes Island in 1907 but
probably arrived earlier (McIntosh, 2001) with sealers or
as the result of a foreign shipwreck (Spirit of the Dawn) in
1896 (Taylor, 2006). DNA studies of the mouse population
identified a mtDNA haplotype also found in Spain but not
elsewhere in New Zealand (Searle, et al., 2009.).
Mice were abundant; their density has been recorded
as high as 147/ha in the coastal zone (Russell, 2012). They
have had a significant detrimental impact on the endemic,
rare and threatened animal species. Invertebrates have been
severely depleted. Mice are responsible for the general
absence of large beetles and the extirpation of at least two
taxa: Loxomerus n.sp. and Tormissus guanicola (Marris,
2000); and several large ground dwelling species are
severely restricted in distribution (Marris, 2000; Russell,
2012). Mice also compete with the four endemic land birds
and have suppressed at least two species of burrowing
seabirds: black-bellied storm petrels (Fregetta tropica) and
subantarctic little shearwater (Puffinus elegans) (Imber, et
al., 2005).
The aim of the project was to eradicate mice from the
archipelago to halt the degradation of biodiversity and
allow native species to recover and flourish. Eradicating
mice would also protect potentially vulnerable species,
for example the nationally critical Antipodean albatross
(Diomedea antipodensis antipodensis), from potential
attacks as recorded on Gough Island and Marion Island
(Davies, et al., 2015; Dilley, et al., 2016).
Fig. 1 Map of the Antipodes Island group.
The site has good ongoing biosecurity integrity. The
islands are remote and isolated, landing requires a permit
and the coastline is generally inaccessible, with no harbour.
In 2012, DOC partnered with the Morgan Foundation to
initiate the project. The Morgan Foundation fronted a
highly publicised fundraising campaign “Million Dollar
Mouse” (MDM), and matched public donations dollar
for dollar. Additional funding came from DOC and other
partners, WWF New Zealand and Island Conservation.
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 131–137. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
131
Island invasives: scaling up to meet the challenge. Ch 1B Rodents: Review
MATERIALS AND METHODS
Planning
DOC planned and managed the operation from its
Murihiku Office in Invercargill. Planning started in
February 2014, with the employment of a full-time project
manager, and took two and half years with a core team of
two increasing to four in the last six months. A much larger
DOC team supported pre-departure preparations. The
Department’s Island Eradication Advisory Group (IEAG)
was engaged from the start, providing technical oversight.
Eradication design was based on agreed best practice
(Broome, et al., 2019). DOC’s Animal Pest Framework
and elements of DOC’s Project Management Framework
(PMF) provided the tools to manage the project.
Procurement
Helicopter and shipping services were sourced using
government processes. In early 2016, DOC contracted the
services of the M.V. Norfolk Guardian, a coastal freighter
flagged in Kingdom of Tonga and a yacht, S.V. Evohe, to
supplement passenger transport.
An experienced eradication pilot was engaged as a
consultant to progress planning while a helicopter supplier
was being sought. Following consultation with potential
suppliers, a temporary hangar (16 m × 12 m × 5. 6 m high)
and a large wooden platform (29 m × 13.8 m) incorporating
a helipad were added to the planned infrastructure to help
protect helicopters and other sensitive equipment from
the elements. The hangar was fastened to the wooden
platform and the whole structure anchored with 38 t of
water ballast positioned around the base of the hangar
frames in palletised 1,000 l cage tanks (Intermediate Bulk
Containers). The anchoring system was designed for easy
installation and extraction and to withstand winds of up to
190 km/hr.
A specialist company “Island Aerial Solutions Ltd”
(IASL) was contracted to supply helicopter services and
a helicopter engineer. Three helicopters were taken to the
island, two AS350 Squirrels (1 × B2 and 1 × FX2) and one
Robinson R44. The R44 provided contingency for marine
search and rescue, enabling baiting to continue using one
AS350 if the other became inoperative.
Preparations
The hangar construction was trialled in a large
warehouse prior to departure. The International Chamber
of Shipping Guide to Helicopter/Ship Operations (2008)
was used in the development of protocols for managing
shipborne helicopter operations. Ship preparations included
establishing a helipad and upgrading emergency response
capabilities onboard. Two months before departure,
interaction trials allowed pilots to practice shipborne
helicopter operations and familiarise the ship’s crew.
Two methods were also trialled for loading helicopters
onto the ship and baiting systems were tested during the
same period. Bucket calibration was done by sowing nontoxic bait across a line of marked quadrants (5 m × 10 m)
extending 65 m perpendicularly from each side of a flight
line over tarmac. Baits were counted in every quadrat to
determine “usable swath width” – the distance to which
bait is reliably spread at or above the desired rate.
An experienced operational team was selected, with
additional skills and experience including engineering and
mechanical repairs, a recovery doctor with extensive patient
extraction and remote emergency medicine experience,
biodiversity monitoring, bait bucket mechanics, technical
eradication knowledge, remote construction, digger
driving and rigging and receiving external helicopter loads.
132
Biosecurity was a significant part of preparations,
and actions were coordinated with a biosecurity plan. A
dedicated DOC team quarantined equipment and supplies
arriving from all around New Zealand. Quarantined items
were generally wrapped in plastic or sealed in plywood
boxes (pods). Pest detection and prevention devices,
including inked tracking cards in tunnels, insect traps, and
rodent bait stations, were in place at the ports of departure
and facilities where equipment and supplies were stored.
The cargo ship’s holds were fumigated for insects.
Transport vessels required a certified clean hull to travel
to the island. A dive inspection of the Norfolk Guardian
discovered biofouling on its hull and the invasive organism
Mediterranean fan worm (Sabella spallanzanii) in the
seachests. A hull clean and treatments of the seachests were
completed and inspected before each voyage to the island.
Animal Control Products (ACP now trading as Orillion)
based in Whanganui, New Zealand, produced 65.5 t of
Pestoff® 20R Rodent Bait containing 20 ppm brodifacoum
between 21 April 2016 and 3 May 2016. ACP analysed
samples from each 500 kg batch of bait, measuring toxicity
using Liquid Chromatography Mass Spectrometry with a
detection limit of 1×10-5% (0.1ppm). The agreed acceptable
range was 16 ppm to 24 ppm brodifacoum by weight.
The bait was packed in four-walled paper bags each
containing 25 kg of bait and transported and stored on
Antipodes Island in large plywood boxes (pods) portable by
forklift and helicopter. The maximum safe load capability
of the helicopters determined the size of the pods (each
contained 28 bags of bait and weighed a total of 805 kg).
The weatherproof pods included a large plastic liner to
protect bait against water ingress.
On 23 May 2016, the Evohe departed Dunedin, New
Zealand, for Antipodes Island with 12 of the project
team onboard. On 25 May 2016, the Norfolk Guardian
departed Timaru, New Zealand with seven project team
members, three helicopters, bait in 94 pods, 30 t of jet
fuel and 20 t of sundry equipment and supplies. Two 1.6
tonne diggers were taken to the island to prepare a level
site for the helicopter hangar. A satellite dish was installed,
providing a fast internet connection. The Evohe remained
at the island while the cargo ship was present, transferring
personnel between ship and shore, and ready to respond in
case of an incident over water during helicopter unloading
of the ship.
Poison baiting
Bait uptake trials were conducted on Antipodes Island
in winter 2013 to assess the palatability of the proposed
bait to mice and the potential risks to non-target species.
The trial used a non-toxic version of Pestoff® 20R Rodent
Bait with the biotracer pyranine added. Baits were spread
by hand over 6 ha at 16 kg/ha. Subsequently, mice were
captured in a grid of Longworth live capture traps and land
birds were captured with hand nets. Captured individuals
were inspected for signs of bait consumption using a UV
light. Observations of birds interacting with baits were
also recorded. Bird faeces were collected opportunistically
along a transect and inspected under UV light. Faecal
samples were assigned to a species by visual inspection
or by DNA analysis for a subset of samples that tested
positive for pyranine (Elliott, et al., 2015).
A boundary flight recorded the treatment area as 2,114
ha before baiting commenced. The boundary was flown
again more tightly before treatment two, recording the area
as 2,075 ha. An advisory team (technical advisor, chief
pilot and assistant project manager) assisted the project
manager with finalising the load site location and layout,
and daily assessment of conditions for baiting. AS350
helicopters, directed by Tracmap GPS systems, spread
Horn, et al.: Antipodes mouse eradication
65.5 t of 2 g Pestoff® 20R Rodent Bait from underslung
bait buckets to complete two comprehensive treatments.
The nominal application rate was 16 kg/ha for treatment
one and 8 kg/ha for treatment two. A minimum interval of
14 days between treatments was preferred, to increase the
likelihood of bait availability for emergent young if mice
were breeding. Parallel flight lines were set at 45 m apart
for a usable swath width of 90 m, giving 50% overlap of
baiting swaths to minimise the risk of gaps. During each
treatment, additional bait was applied to the coastline, steep
slopes (50° to 70°), cliffs (greater than 70°) and other areas
of concern to the pilots or identified by geospatial analysis
as having potentially insufficient coverage. An observer in
the back of the helicopter monitored distribution of bait
on cliff baiting flights, which were undertaken at about 40
metre vertical increments.
Bait was made available inside storage containers
and the interior and sub floor spaces of buildings by hand
spreading or placing baits in bait stations. A bait station
comprised a numbered shallow clear petri dish with ten
Pestoff® 20R Rodent Bait pellets. These were placed in
each compartment or room of a structure and checked
daily. A total of 72 bait stations were placed in structures
on 18 June. Baits were thrown by hand to achieve coverage
of approximately four bait pellets per square metre under
the hut and Castaway Depot and in the open wastewater
drain. Toilet pits were checked daily and a handful of
baits were scattered down each pit as required to maintain
availability to mice. Holes were drilled in the floor of the
helipad and hangar to access the subfloor space, and baits
dropped through. Mouse activity was monitored around the
accommodation area using inked tracking cards secured in
tunnels (tracking tunnels) and baited with Pestoff® 20R
Rodent Baits; and three trail cameras focused on bait
stations under the hut and Castaway Depot. Approximately
4 kg of bait was used for structure baiting.
West Windward Island (7.0 ha) and East Windward
Island (8.5 ha) were not baited during the first treatment
as it was unknown if mice were present. These islands
were monitored for mice between treatments using ten
inked tracking cards baited with peanut butter and placed
in tunnels (tracking tunnels) for 12 nights. Bollons Island
(52.6 ha) was believed to be mouse-free prior to the
operation but six tracking tunnels were installed between
bait treatments for 12 nights and baited with peanut butter
to provide further confidence in its status.
Monitoring to determine if mice had been eradicated
occurred in late summer 2018, approximately 18 months
after the baiting operation. By this time, a surviving mouse
population should have recovered to detectable levels. Late
summer was chosen as any breeding would have peaked
and juveniles would have been present. Monitoring for
mice was undertaken using 280 inked tracking cards in
tunnels baited with peanut butter and distributed along
28 transect lines. Each transect comprised 10 tracking
tunnels spaced 200 m apart. The transects were distributed
extensively across Antipodes Island. They were placed
in all habitat types, particularly in areas where mice had
previously been in high abundance (e.g. near penguin
colonies) and adjacent to inaccessible terrain. Tracking
cards were checked and replaced approximately every
five days for a period of three weeks. Supplementing this,
two rodent detection dogs and their handlers searched the
island for mice between 21 February and 15 March 2018.
The dogs searched in accessible areas across the plateau
and southern coast.
Non-target species
A non-target species technical advisory group
recommended a strategy for managing risks to native
species that did not include captive management but
relied on natural populations outside of the treatment area.
This strategy became part of the application to DOC, as
administrators of the site, for consent to spread bait. Three
of the four endemic land bird taxa were considered at risk
from either primary or secondary poisoning. Bollons Island
(52.6 ha) and Archway Island (6.2 ha) were excluded from
the treatment area during planning because evidence from
historic studies of invertebrates (Marris, 2000; McIntosh,
2001; Russell, 2012) and limited monitoring for mice on
Bollons Island in 2014 (B. Rance pers. comm. 2014) gave
sufficient confidence that mice were not present. These
islands provided a natural refuge of 58.8 ha, 1.5 km north
of Antipodes Island, where species would not be exposed
to bait.
Baseline monitoring of endemic land bird taxa was
conducted on Antipodes Island between 2013 and 2016
including immediately prior to bait application in winter
2016. Post-eradication monitoring occurred in the weeks
after bait application in July 2016, and in the summers of
2017 and 2018, to record any population impacts of the
operation. Distance sampling (Buckland, et al., 2001)
was used to estimate the density and abundance of the
endemic Antipodes parakeet (Cyanoramphus unicolor),
Reischek’s parakeet (Cyanoramphus hochstetteri), and
the endemic subspecies of the New Zealand pipit (Anthus
novaeseelandiae steindachneri). The perpendicular
distance to individuals or groups of birds was measured
from transect lines of variable length to the nearest metre
using a laser range-finder. Transects were distributed
throughout the island and repeated as often as practicable.
The aim was a sample of 60 to 80 encounters of each
species for robust modelling of the detection probability
and resultant population density. The technique relies on
sightings of birds, so sampling was generally avoided when
the weather was wet and cold as birds are less conspicuous.
The computer software ‘Distance 6.2’ (Thomas, et al.,
2010) was used to analyse the data and compute population
estimates. As the number of detections recorded was low
for many of the survey periods, data were pooled and a
global detection function was computed, from which survey
specific estimates of density were calculated (Buckland
et al. 2001). Visual comparison of point estimates and
their 95% confidence intervals were reinforced using
a comparison of Poisson rates (poisson.test; R
Core Team, 2013) for three paired pre- and post-toxin
application survey dates and departures from a hypothesis
of no change in density tested.
Antipodes
snipe
(Coenocorypha
aucklandica
meinertzhagenae) were monitored by recording the number
of snipe seen per hour by observers traversing the island on
foot, to give an encounter rate. The change in encounter
rate between years was assessed using a generalised linear
model with negative binomial errors.
To determine if the breeding success of Antipodean
albatross was impacted by the operation, the fledging
success of Antipodean albatross chicks within 50 m of
the load site was recorded in summer 2017 by visiting the
nests prior to chicks fledging. The results were compared
with fledging success of chicks, alive at the time of bait
application, in two study areas on Antipodes Island.
No formal searching for potentially poisoned animals
was done but carcasses found opportunistically were
examined. The gut cavity was opened and inspected for
haemorrhaging and or the presence of green bait in the
stomach or intestines indicating poisoning by brodifacoum.
Liver samples were collected from the carcasses of
pipits and snipe and stored frozen. Samples were sent to
Landcare Research and analysed using High Performance
Liquid Chromatography with a detection limit of 1×10-6%
(0.001ppm).
133
Island invasives: scaling up to meet the challenge. Ch 1B Rodents: Review
Project communication
Public engagement was measured by recording the
number of media articles about the project (on television,
radio, print) and visits to the project’s website www.
milliondollarmouse.org.nz) and Facebook page (www.
facebook.com/milliondollarmouse) during the operational
phase.
RESULTS
The baiting operation was implemented and completed
in winter 2016. Insufficient resourcing in the first year
of planning and competition with other organisational
priorities put pressure on the project team and risked
delaying implementation. The development of project
knowledge and a wealth of experience enabled quality
advice from DOC’s IEAG. Their strong support maintained
focus on objectives and influenced the prioritisation of
resources in the preparation phase. Procuring helicopter
services and a cargo ship were the crux of logistics planning
but proved difficult due to a small pool of suitable suppliers
and complex processes. Over a year and a half was spent
investigating options and developing trust with potential
suppliers to prove the viability of the work and find capable
operators who were willing to commit.
Calibration of bait buckets gave a usable swath width
of 90 m for standard buckets (360° spread) and 40 m
for the deflector bucket (180° spread). Pre-departure
trials identified important improvements in systems and
componentry including changes to the pneumatic feed from
helicopters to the bait bucket, replacement of incorrectly
sized bracing elements on the hangar and refinement of
the system for its construction. Trials identified that lifting
helicopters by the rotor head was the best technique to
manoeuvre them in and out of the ship’s hold.
The toxicity of all 131 batches of bait supplied met the
contract standards. The average toxicity was 19.8 ppm of
brodifacoum and the range was 16.5 ppm to 23.9 ppm ±
7%. The operational team arrived at Antipodes Island on
27 May 2016. It took approximately 90 minutes to extract
each helicopter from the ship’s hold and ready them for
flying. Ship unloading was completed with 250 loads flown
ashore over 12 days with suitable weather for helicopter
operations occurring periodically on five of those days.
Helicopter long-line operations to unload and load the ship
were challenging and required precision from the pilots
and a strong communicator on the deck of the ship to
inform the pilot of the position of the hook and help direct
the work. The construction team of six people established
the field camp, completed complex site preparations and
safely installed temporary infrastructure within 11 days
before departing with the transport vessels on 7 June 2016.
An emergency response exercise was conducted on 8 June
to practice helicopter recovery of a person from the water
with a rescue scoop net and a rescuer in a human sling on
a long-line.
Readiness for baiting was achieved by 9 June 2016
but poor weather delayed baiting until 18 June 2016 when
a brief respite in conditions allowed baiting of a small
area (54 ha). This gave the opportunity for an initial test
of personnel, loading systems and equipment ahead of
better weather windows. The baited area incorporated the
field camp and load site, enabling structure baiting to be
completed to make bait available early in the programme
around the accommodation area where there was the
highest risk of alternative food sources for mice. Aerial
baiting continued incrementally as the weather allowed
until coverage was complete. Suitable weather windows
for baiting operations were generally short, and conditions
were changeable and generally windy. The longest
continuous period of bait application achieved was 3.5
hours. Each day’s baiting built on previous work using a
“rolling front” approach, with the aim of minimising the
area needing rebaiting if work was interrupted for too long.
Treatment one was completed on 29 June 2016 with
bait application occurring on 18, 21, 22, 27, 28 and 29
June. The interruption after baiting on 22 June was greater
than three days, so the last two bait swaths sown that day
were sown again on 27 June with 50% overlap. A total
application of 45.6 t of bait was applied during treatment
one at an average rate of 21.6 kg/ha. No mouse sign was
detected on either of the Windward islands so neither were
baited, increasing the area where land birds would not be
exposed to bait to 75.3 ha.
Treatment two commenced on 8 July, continued to 10
July and was completed on 12 July 2016. A total of 19.9
t was spread at an average application rate of 9.6 kg/ha.
The average sowing rate for both treatments combined was
31.2 kg/ha, including application of all the contingency
bait. Contingency bait was additional bait (20% of the
planned total) taken to mitigate the risk of loss or damage
during transport and storage, or of the treatment area being
larger than expected. The rate of bait spreading averaged
1.79 t/hr for the first treatment and 0.93 t/hr for the second,
giving an overall average of 1.44 t/hr. The interval between
treatments was at least 16 days for 97% of the area, and
between ten and twelve days for the remainder. Few
technical issues with bait spread were encountered and
none limited operations.
Rainfall data were collected daily, and some form
of precipitation fell most days. A total of 7.9 mm fell in
the 48 hours following application of 15.6 t of bait on 22
June in treatment one. Bait degradation was not formally
monitored. However, visual inspection showed baits were
weathered but generally intact at the start of treatment two,
20 days after application.
Analysis of GPS flight records for aerial bait spread
showed that comprehensive bait coverage was achieved
with no apparent gaps. The total maximum amount of bait
taken from all bait stations set up for structure baiting was
240 g of the 4 kg available. Most of the bait take occurred
in the first three nights and 73% of consumption occurred
by night six. Imagery from a trail camera showed mice
picking up and carrying away the 2 g bait pellets. Two
mice were last recorded taking bait on 7 July, 20 days after
application. Dissection of a mouse trapped nearby on the
same day showed the stomach and intestines were green
and full of bait.
Table 1 Incidental dead bird finds on Antipodes Island following bait application.
Species
Antipodes parakeet Cyanoramphus unicolor
Reischek’s parakeet Cyanoramphus hochstetteri
Pipit Anthus novaeseelandiae steindachneri
Snipe Coenocorypha aucklandica meinertzhagenae
Mallard duck Anas platyrhynchos
134
Autopsy
1 poisoned
1 poisoned
3 poisoned
2 no sign
1 poisoned
Brodifacoum (μg/g) ± 6%
Unknown
Unknown
0.028; 0.034; 0.01
0.015; 0.031
Unknown
Horn, et al.: Antipodes mouse eradication
Table 2 Comparison of Poisson rates at two time points pre- and post-application of toxin on Reischek’s parakeet,
Antipodes parakeet and Antipodes pipit. Rate ratios, their 95% CI’s and tests of departure from a hypothesis of
no change in density between surveys are reported. Rate ratios <1.0 indicate population decline and those >1.0
population increase between surveys.
Comparison of Poisson rates between surveys
Pre-drop 2016 & PostPre-drop 2016 & Jan/Feb
Post-drop 2016 & Jan/Feb
drop 2016
2018
2018
Reischek’s parakeet
Antipodes parakeet
Antipodes pipit
**
0.17 (0.13–0.23)**
0.57 (0.36–0.95)*
0 (0.05–0.10)**
0.85 (0.63–1.17)#
2.91 (1.81–4.88) **
1.38 (1.08–1.76)*
4.97 (3.92–6.30)**
5.09 (3.81–6.77)**
19.44 (13.84–27.94)**
P <0.001; * P <0.05; # not significant
No mouse sign was detected from 7,170 tracking tunnel
nights and searching with dogs during mouse monitoring
in summer 2018. The search effort and the evidence were
reviewed by DOC’s Island Eradication Advisory Group
and the eradication of mice from Antipodes Island was
declared successful in March 2018.
Non-target species impacts
Bait trials in 2013 demonstrated 100% uptake of the
bait by mice and suggested a risk of primary poisoning for
pipits but not for parakeets or snipe (Elliott, et al., 2015).
During the eradication operation itself, eight dead birds
of five species were found incidentally and all had been
poisoned (Table 1). The associated search effort was at
least 103 hours of extensive field work for monitoring land
birds. Additionally, staff walked an 800 m route between
Reef Point and the load site (Fig 1) almost daily for the
six weeks between initial bait application in the area and
departure. During the operation, some pipits were observed
occasionally pecking at baits and some baits were found
to have been chewed by parakeets, but most baits were
untouched.
that a significant number of pipits and parakeets probably
succumbed to brodifacoum poisoning immediately
following the application of bait. However, the populations
of pipits and both parakeet species were able to persist and
have increased greatly each year, recovering to densities
that are similar to or higher than pre-eradication estimates
by summer 2018 (Table 2; Figs 2, 3 and 4). Pipits have
responded particularly strongly with very large year on
year increases in density estimates since 2016. Anecdotal
observations in summer 2018 were consistent with the
reported increase. On most occasions when monitoring
team members sat down in the field, pipits would
immediately appear and walked around and on them,
finding food items such as caterpillars within minutes (F.
Cox, pers. comm. 2018).
Despite the use of a global detection function, low
numbers of observations led to large confidence intervals
about density estimates derived from distance sampling
(Figs 2, 3 and 4). Prior to 2016, only the sampling of
Reischek’s parakeets in October 2014 (61 encounters)
reached the desired sample size of 60 to 80 encounters.
In 2016, pre-baiting sampling for Antipodes parakeets
(22 encounters) and post-baiting sampling for pipits (40
encounters) failed to reach this target. Overall, more
sampling was done immediately post-baiting in 2016
(329 encounters) than before (186 encounters) due to
time constraints. Poor weather also often constrained the
method. The results (Table 2; Figs 2, 3, and 4) suggest
Fig. 3 Distance sampling results for Antipodes Island
parakeets, Antipodes Island.
Fig. 2 Distance sampling results for Reischek’s parakeets,
Antipodes Island.
Fig. 4 Distance sampling results for pipits, Antipodes
Island.
135
Island invasives: scaling up to meet the challenge. Ch 1B Rodents: Review
Table 3 Results of snipe encounter rate surveys recorded on Antipodes Island between 2013
and 2017.
Year
Person hours
2013
2014
2015
2016
2017
2018
a
b
341
206.75
140.5
178
224
783
Snipe seen
Snipe seen per
hour
38
26
17
6
8
132
0.1079
0.1322
0.1279
0.0330
0.0345
0.1640
123
97
26
105
475
p
0.4938
0.9267
0.0085**a
0.9373
0.0001***b
Note significant difference in encounter rate between 2015 and 2016 prior to the eradication operation.
Note significant difference in encounter rate between 2018 and 2017.
Snipe have been monitored each summer between
2013 and 2018. Snipe were more abundant in 2018 than
ever before, but there has been considerable inter-annual
variation in snipe abundance and the difference between
2018 and all the other years is not significant (Table 3).
The between-year change in snipe abundance is probably
more informative. Significant changes in snipe abundance
occurred in 2015–2016 (a decline) and 2017–2018 (an
increase). The large decline (72%) in the snipe encounter
rate between 2015 and 2016 occurred before the mouse
eradication so was not a result of the poison operation.
The reason for this is unknown. There was a small, nonsignificant increase in the snipe encounter rate between
2016 and 2017 (Table 3), suggesting little or no by-kill of
snipe during the mouse eradication. In contrast a dramatic
increase (475%) occurred in snipe encounters between
2017 and 2018.
Helicopter activity did not have a detrimental effect
on nearby Antipodean albatross chicks. All seven chicks
within 50 m of the bait loading site were alive at the
completion of operations and six out of the seven of them
(86%) fledged successfully in early 2017, comparable with
90% outside the load site.
Scientists visiting Antipodes Island in summer 2017
and summer 2018 also noted a greater abundance of moths
and the endemic fly (Xenocalliphora antipodea) than
before the eradication of mice, observing them on flowers
of the native groundsel (Senecio radiolatus) and Macquarie
Island cabbage (Stilbocarpa polaris). This endemic fly was
also abundant inside the Antipodes Hut for the first summer
in over 20 years of visitation. A gathering of hundreds of
large noctuid moths, suspected to be Graphania ustistriga,
was also observed for the first time in 2018 despite 10
previous month-long summer visits to Antipodes Island
between 1996 and 2017 (K. Walker, pers. comm. 2018).
Large caterpillars, suspected to be larvae of the same
noctuid moth species were regularly seen and observed
being preyed on by pipits (K. Walker, pers. comm. 2018).
Project communication
Media coverage of the operation included seven primetime television news stories and several radio interviews,
print and online stories. Social media engagement peaked
in June 2016 with 23,906 views of the MDM website and
71,967 on the MDM Facebook page. DOC social media
also peaked at 77,710 views for the month. Outreach was
amplified through the communications networks of project
partners, the Morgan Foundation, WWF-New Zealand and
Island Conservation.
136
Change between
years (%)
DISCUSSION
A robust plan was formulated and delivered despite
initial difficulties sourcing shipping and helicopter services.
Complex projects require good resourcing in the planning
phase and organisational prioritisation with significant scale
up in resourcing for the preparation phase. Key factors for
the delivery of the project were a) quality technical advice,
b) single point accountability for overseeing the work
and a team approach during preparations and operational
phases, c) use of experienced personnel in key roles, d) a
proven bait product, e) dependable and tested equipment, f)
extensive contingency planning, g) a partnership approach
with suppliers and e) the financial and moral support of
private and public partners.
The brevity and inconsistency of weather opportunities
in this environment showed the importance of being
prepared and effectively using every opportunity to
complete baiting. Additional skills and operational
experience improved team performance and selfsufficiency. Equipment could generally be maintained on
site and situational decision-making benefitted from the
advice of senior team members. High speed internet access
and video production capabilities enabled the team to
communicate the project directly and engage an audience.
Pilots’ long-lining capabilities for ship operations could be
considered a separate skill from baiting and, if necessary,
pilots with specific skills should be engaged for the task.
Similarly, coastal baiting with the deflector bucket requires
specific attention and experience.
Non-target impacts
Monitoring evidence suggests the adverse effects of the
operation on land birds were short lived. These impacts
are expected to be outweighed by the long-term benefits to
native species from the permanent removal of competition
with mice. The risk to non-target species was effectively
limited by relying on natural populations on Bollons and
Archway Islands where they weren’t exposed to bait. Prior
to the mouse eradication, both parakeets and the pipit had
rarely been observed making flights of more than 100 m
on Antipodes Island, so while they are capable of crossing
the 1.5 km strait between Bollons Island and the main
Antipodes Island, it must have been a rare event. The risk
of parakeets and pipits, resident on Bollons and Archway
Islands, being killed by poison when they commuted across
the strait was judged low. This reasoning eliminated the
need to catch and maintain a captive population. During
the bait uptake trial neither parakeet species was detected
eating bait, yet both species were killed by the poison.
Parakeets may have become habituated to the bait during
the operation because of the longer exposure (more than 35
days) and changing palatability of baits as they weathered
relative to the non-toxic trial (14 days). The large
Horn, et al.: Antipodes mouse eradication
variability in population density estimates derived from
distance sampling were largely driven by the relatively
low encounter rates for all three species monitored using
this method and should be treated as indicative only. More
data would have improved the robustness of the results as
would an improved sampling design to account for only
recently discovered shifts in winter distribution for both
parakeet and pipits. This, however, is difficult to achieve
for such a remote and expensive site to visit and for one that
frequently experiences less than ideal survey conditions in
generally time-constrained survey periods.
It is unlikely that recruitment alone could account for
the apparent rapid recovery of pipits and parakeets by
summer 2017 (Figs 2, 3, and 4), suggesting the distance
sampling results overestimated the losses and/or recovery.
For both parakeet species, the large increases in population
density, relative to post-baiting lows, were observed before
most chicks had fledged (G. Elliott pers. comm. 2017).
Pipits are unlikely to have raised more than one clutch by
January 2017 which doesn’t account for the nearly 500%
increase in the population density estimate in summer 2017
since their post-baiting low. The similarly large increase in
the estimated density of pipits between 2017 and 2018 (Fig.
4) is more likely to be real considering the observations of
field staff.
The very large increases in the encounter rate of
snipe and the density estimate of pipits in summer 2018
are presumed to be the result of large increases in the
abundance of invertebrates following the eradication of
mice and the resultant increases in reproductive output and
survival.
Effective distance sampling for pipits within dense
coastal vegetation, a habitat favoured by pipits in winter,
was problematic. The short time-frames available during
the operation for monitoring immediately before and after
baiting meant distance sampling occurred in variable
conditions and with variable effort across different habitat
types, which may have exaggerated the estimated population
declines following bait application. The extraordinarily
large estimate of pipit population density pre-baiting in
2016 (Fig. 4) is possibly biased by proportionally greater
sampling effort of abandoned penguin colonies (where
pipits and parakeets are now known to congregate in
winter) relative to that within the island interior (and where
most of the 2013 counts were done). This reinforces the
uncertainty of results.
The seasonal timing of distance sampling for land birds
before and after baiting was also inconsistent (Figs 2, 3
and 4). The observed changes in seasonal distribution of
these species therefore makes the use of a global detection
function (which assumes constant detectability across
surveys) problematic and dilutes direct comparability
of the density results. Changes in detectability caused
by movements to and from the coast may be biasing the
results and at least partly account for the relatively low
population density estimates so soon after the bait spread.
It is recommended that results from surveys done at the
same time be pooled if sufficient data are available.
The eradication of mice from Antipodes Island is
a huge achievement for conservation in New Zealand.
Hundreds of years of ecological devastation by mice has
been halted and indigenous wildlife has started to recover.
The importance of the result is reflected by the national and
international protection of the site, recognising its special
natural heritage values. The result provides momentum to
New Zealand’s Predator Free 2050 initiative and is a step
closer to the vision of a New Zealand Subantarctic Islands
region free of mammalian pests. Of the five island groups
in the region, only Auckland Island now has mammalian
pests: pigs (Sus scrofa), cats (Felis catus) and mice (Mus
musculus). Over time it is expected that the invertebrate
fauna on Antipodes Island will recover to reflect the
abundance and species diversity recorded on Bollons
Island and Archway Island, where no mice were present. It
is hoped that species of larger-bodied ground invertebrates
(for example tenebrionids), reduced to low abundance, will
recover and others which became extinct on Antipodes
Island through predation by mice (for example the
unidentified weta and Loxomerus sp.), can be successfully
reintroduced from the offshore islands where they may
survive. The population densities of land bird species are
expected to further increase and stabilise with the recovery
of food sources and lack of competition with mice. Absent
burrowing seabirds, for example black-bellied storm
petrel, are also expected to recommence breeding on
Antipodes Island. Further monitoring for land birds will
occur opportunistically on an annual basis in conjunction
with albatross research. Broader outcome monitoring will
be repeated in approximately five to ten years’ time and
will include a repeat of invertebrate sampling, sampling
of the seabird species breeding on Antipodes Island and
measurement of change in vegetation monitoring plots.
ACKNOWLEDGEMENTS
The project drew on the support of hundreds of people.
Special thanks to the DOC team, the operational team,
public supporters, project partners, suppliers, DOC’s
Island Eradication Advisory Group, Kath Walker and
Helene Thygesen.
REFERENCES
Broome, K., Brown, D., Brown, K., Murphy, E., Birmingham, C.,
Golding, C., Corson, P. and Cox, A. (2019). ‘House mice on islands:
management and lessons from New Zealand’. In: C.R. Veitch, M.N.
Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) Island invasives:
scaling up to meet the challenge, pp. 100–107. Occasional Paper SSC
no. 62. Gland, Switzerland: IUCN.
Buckland, S.T., Anderson, D.R., Burnham, K.P., Laake, J.L., Borchers,
D.L. and Thomas, L. (2001). Introduction to Distance Sampling:
Estimating Abundance of Biological Populations. UK: Oxford
University Press.
Davies, D., Dilley, B.J., Bond, A.L., Cuthbert, R.J. and Ryan, P.G. (2015).
‘Trends and tactics of mouse predation on Tristan albatross Diomedea
dabbenena chicks at Gough Island, South Atlantic Ocean’. Avian
Conservation and Ecology 10(1): 5.
Dilley, B., Schoombie, S., Schoombie, J. and Ryan, P. (2016). ‘Scalping’
of albatross fledglings by introduced mice spreads rapidly at Marion
Island’. Antarctic Science 28(2):73–80.
Elliott, G.P., Greene, T.C., Nathan, H.W. and Russell, J.C. (2015).
‘Winter bait uptake trials and related field work on Antipodes Island
in preparation for mouse (Mus musculus) eradication’. DOC Research
and Development Series 345. Wellington, New Zealand: Department of
Conservation.
Imber, M.J., Bell, B.D. and Bell, E.A. (2005). ‘Antipodes Island birds in
autumn 2001’. Nortornis 52(3): 125–132.
International Chamber of Shipping (2008). Guide to Helicopter/ship
Operations; Fourth Edition. 12 Carthusian Street, London: Marisec
Publications.
Marris, J.W.M. (2000). ‘The beetle (Coleoptera) fauna of the Antipodes
Islands, with comments on the impact of mice and an annotated checklist
of the insect and arachnid fauna’. Journal of the Royal Society of New
Zealand, 30(2): 169–195.
McIntosh, A.R. (2001). ‘The Impact of Mice on the Antipodes Islands’.
In: P. McClelland (ed.) Antipodes Island Expedition, October–
November 1995, pp. 52–57. Invercargill, New Zealand: Department of
Conservation.
R Core Team (2013). R: A Language and Environment for Statistical
Computing. Vienna, Austria: R Foundation for Statistical Computing.
Russell, J.C. (2012). ‘Spatio-temporal patterns of introduced mice and
invertebrates on Antipodes Island’. Polar Biology 35: 1187–1195.
Searle, J.B., Jamieson, P.M., Gunduz, I., Stevens, M.I., Jones, E.P.,
Gemmill, C.E.C. and King, C.M. (2009). ‘The diverse origins of New
Zealand house mice’. Proceedings of the Royal Society Bulletin 276:
209–217.
Taylor, R. (2006). Straight through from London. The Antipodes and
Bounty Islands, New Zealand. Christchurch, New Zealand: Heritage
Expeditions.
Thomas, L., Buckland, S.T., Rexstad, E.A., Laake, J.L., Strindberg, S.,
Hedley, S.L., Bishop, J.R.B., Marques, T.A. and Burnham, K.P. 2010.
‘Distance software: design and analysis of distance sampling surveys
for estimating population size’. Journal of Applied Ecology 47: 5–14.
137
C.E. Main, E. Bell, K. Floyd, J. Tayton, J. Ibbotson, W. Whittington, P.R. Taylor, R. Reid, K. Varnham, T. Churchyard, L. Bambini, A. Douse, T. Nicolson and G. Campbell
Main, C.E.; E. Bell, K. Floyd, J. Tayton, J. Ibbotson, W. Whittington, P.R. Taylor, R. Reid, K. Varnham, T. Churchyard, L. Bambini, A. Douse, T. Nicolson and G. Campbell.
Scaling down (cliffs) to meet the challenge: the Shiants black rat eradication
Scaling down (cliffs) to meet the challenge: the Shiants’
black rat eradication
C.E. Main1, E. Bell2, K. Floyd2, J. Tayton2, J. Ibbotson2, W. Whittington2, P.R. Taylor3, R. Reid1, K. Varnham4,
T. Churchyard5, L. Bambini5, A. Douse6, T. Nicolson7 and G. Campbell1
Royal Society for the Protection of Birds, North Scotland Regional Office, Etive House, Beechwood Park, Inverness,
IV2 3BW, UK. <charlottemain@gmail.com>. 2Wildlife Management International Limited, P.O. Box 607, Blenheim
7240, New Zealand. 3Open Seas, c/o J&H Mitchell, 51 Atholl Rd, Pitlochry, Scotland, PH16 5BU. 4Royal Society for
the Protection of Birds, RSPB Headquarters, The Lodge, Sandy, Bedfordshire, SG19 2DL, UK. 5Royal Society for the
Protection of Birds, RSPB Scotland Headquarters, Edinburgh EH12 9DH. 6Scottish Natural Heritage, Great Glen
House, Leachkin Road, Inverness, IV3 8NW, UK. 7Sissinghurst Castle, Cranbrook, Kent, TN17 2AB, UK.
1
Abstract A successful ground-based eradication of black rats (Rattus rattus) was undertaken on the remote, uninhabited
Shiant Isles of north-west Scotland over winter (14 October–28 March) 2015–16. The rat eradication was carried out as
part of the Shiants Seabird Recovery Project, which aims to secure long-term breeding habitat for protected seabirds and
to attract European storm petrels and Manx shearwaters to nest on the Shiants. Throughout the eradication operation,
teams were stationed on two of the three main Shiant islands (Eilean an Tighe, Eilean Mhuire), with access to the
third (Garbh Eilean) via a boulder causeway from Eilean an Tighe. Bait (Contrac® blocks containing the anticoagulant
bromadiolone 0.005% w/w), was deployed in a grid of 1,183 bait stations covering all areas of the islands and sea stacks.
Bait stations were set 50 m apart, with intervals reduced to 25 m in coastal areas of predicted high rat density. Difficult
areas were accessed by boat and cliffs of ~120 m in height were accessed by abseiling down ropes made safe using either
bolted anchors or ground stakes. The team of staff and volunteers worked through difficult conditions, deploying bait and
monitoring intensively for any surviving rats using a combination of tools. The islands were declared rat free in March
2018. This ambitious and challenging project has greatly enhanced UK capacity in rodent eradications for the purposes
of conservation.
Keywords: biosecurity, conservation priority, eradicate, incursion, invasive alien vertebrate, island restoration, reinvasion
INTRODUCTION
The Shiant Isles is a group of small, uninhabited islands
that lie in the Minch (57.9° N, 6.4° W), ca. 6 km east of
the island of Lewis and Harris, north-west Scotland. Of the
Shiants’ three main islands, the largest two: Garbh Eilean
(GE, 88 ha) and Eilean an Taighe (ET, 54 ha) are connected
by a boulder causeway, and ~500 m to the east of GE lies
Eilean Mhuire (EM, 31 ha). A chain of sea stacks, the
Galtachan, lie to the west of GE (Fig. 1).
Archaeological
evidence
documents
previous
inhabitation of these islands by humans dating back perhaps
to the Iron Age (Foster, 2000) but since the 18th century the
Shiants have remained uninhabited, and the last remaining
building (the ‘bothy’ on ET, close to the boulder causeway)
Fig. 1 Location of the islands in the wider area of northwest Scotland.
is occupied only during visits by the islands’ owners, or by
tourists.
The Shiants consist mainly of dolerite sills, formed by
intrusion of igneous rock between overlying sedimentary
rock strata. These sills were then exposed to reveal
impressive, columnar structures that now rise steeply
to a height of ~ 150 m at their highest point on GE and
have been eroded to form extensive boulder scree areas,
particularly on the east side of GE (Walker, 1930). The
smallest of the three main islands, EM, has cliffs rising to
around 80 m, and more conglomerate substrate than ET
and GE (Walker, 1930; Gibb & Henderson, 1996).
Habitats present on the islands range from blanket
bog and wet heath across the interior of GE and ET, to
fertile, species rich grasslands along the coasts of GE and
ET and across ME. The maritime environment has a strong
influence on the composition of the islands’ vegetation
and soils have been enriched by guano from centuries of
seabird occupation and from past human cultivation. The
three main islands have all been historically grazed by
sheep (Ovis aries) (counts of sheep performed year-round
gave estimates of 50 to 80 per island). A colony of grey
seals (Halichoerus grypus) breeds on the islands, and both
common seals (Phoca vitulina) and otters (Lutra lutra) are
frequent visitors. Other than the sheep and an introduced
population of black rats (Rattus rattus) there are no other
known resident populations of terrestrial mammals.
The remoteness of the Shiants, their large amount of
suitable habitat and proximity to feeding grounds makes
the islands ideal breeding sites for various seabirds.
Their importance is internationally recognised through
designation as a Site of Special Scientific Interest (SSSI;
site code 8575) and as a Special Protection Area (SPA;
EU code UK9001041) for breeding populations of puffins
(Fratercula arctica) (approximately 10% of the UK
breeding population, Mitchell, et al., 2004), razorbills (Alca
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
138
up to meet the challenge, pp. 138–146. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Main, et al.: Rat eradication on the Shiants
torda), common guillemots (Uria aalgae), European shags
(Phalacrocorax aristotelis), black-legged kittiwakes (Rissa
tridactyla), and northern fulmar (Fulmarus glacialis), and
wintering barnacle geese (Branta leucopsis). The seabird
assemblage also includes great skua (Stercorarius skua),
black guillemots (Cepphus grylle), herring gulls (Larus
argentatus), common gulls (L. canus), great black-backed
gulls (L. marinus) and lesser black-backed gulls (L.
fuscus). White-tailed eagles (Haliaeetus albicilla) returned
to breed on the islands in 2014 after an absence of over
100 years, following the re-introduction of the species to
Scotland (Love, 1983). Other seabirds such as European
storm-petrel (Hydrobates pelagicus) and Manx shearwater
(Puffinus puffinus) have not been recorded as breeding at
the Shiants, despite the large amount of suitable habitat
for these birds. Both of these species are of international
conservation concern. At the last major census, European
storm petrels breeding on the isles of the UK and Ireland
were estimated to number 125,000 pairs, representing
3–11% of the global population. In the same census Manx
shearwaters were estimated at 332,000 pairs breeding in
the UK and Ireland, with the majority found on the islands
of Rum, north-west Scotland (120,000 pairs), Skomer,
Wales (102,000 pairs) and Skokholm, Wales (46,000
pairs, Mitchell, et al., 2004). A further survey of Manx
shearwaters on Rum has also estimated 70,000 breeding
pairs at the Rum colony (Murray & Shewry, 2002).
Rats (Rattus spp.) are among the highest risk invasive
species, having had devastating effects on native wildlife
on island groups such as New Zealand (Towns, et al., 2006)
and worldwide through predation, and both competition
for and modification of habitat (Jones, et al., 2008). Rats
have been recorded on more than 80% of the world’s island
groups (Atkinson, 1985), but their successful removal from
islands ranging in size from less than 1 ha to 12,875 ha has
been pioneered in New Zealand and is being applied across
the globe. In the UK, rats have been successfully eradicated
from islands ranging in size from just one hectare (e.g.
Inchgarvie, Firth of Forth, Scotland) to 1,300 ha (Canna
& Sanday, Scotland) (Ratcliffe, et al., 2009; Thomas,
et al., 2017a; Bell, 2019). Of the successful UK island
rat eradications, all were of brown (Norway) rat (Rattus
norvegicus) except in the case of Lundy, which included
populations of both brown rat and black rat (Thomas, et al.,
2017a; Bell, 2019). The removal of rats is essential where
predation either limits productivity or threatens to lead to
the complete loss of important seabird colonies.
Black rats were introduced to the Shiant Isles
(accidentally, it is assumed) by humans, either through
stock movements by previous island inhabitants or by
shipwreck (e.g. Haswell Smith, 2004), though no evidence
has established how the rats arrived. The rats are thought to
have had negative impacts on the seabirds at these islands
as follows. Diet analysis at the Shiants has indicated that
rats consumed a range of material of marine origin (Stapp,
2002) as well as vegetation and invertebrates present at
the Shiants (Stapp, 2002; Bell, 2013). The stable isotope
ratios of carbon and nitrogen, extracted from rat tissues of
individuals caught at seabird colonies were closer to those
from tissues of seabird origin than those of rats caught
from areas away from seabird colonies (Stapp, 2002).
This indicated that in the seabird breeding season, coastal
colonies of rats were likely to have fed upon on seabird
eggs and chicks.
Following a detailed assessment of UK islands
with invasive, non-native species the Shiant Isles were
identified as being a priority site for rat eradication because
of their abundance of potential petrel and storm-petrel
breeding habitat (Ratcliffe, et al., 2009). A successful rat
eradication at these islands would additionally benefit the
existing colonies of protected seabirds. Since the islands
lie approximately 6 km offshore and are uninhabited by
humans, the risk of natural invasion by brown rats from the
nearest islands of the Outer Hebrides is considered to be
low. A feasibility study commissioned by the Royal Society
for the Protection of Birds (RSPB), and undertaken by
Wildlife Management International Ltd (WMIL) in April
2012, found that eradication of the black rat population at
the Shiants was feasible (Bell, 2013).
Subsequently, the Shiant Isles Seabird Recovery
Project (SSRP) was established as a four-year partnership
between the islands’ owners (the Nicolson family), RSPB
and Scottish Natural Heritage (SNH). The four core aims
of the project were: i) to eradicate the invasive black rat
population; ii) actively encourage petrels (European storm
petrel and Manx shearwater) to nest at the islands; iii)
audit island biosecurity at UK SPAs and iv) increase UK
capacity in island restoration. Funding for the project was
provided by the EU LIFE fund (LIFE13/NAT/UK/000209
LIFE+ SHIANTS), SNH, and RSPB.
The eradication component of the SSRP was undertaken
over the period 2015–2016. An open tender process was
used to invite operators to bid for a contract to undertake
eradication work at the Shiants. This resulted in the
selection of WMIL to carry out the eradication operations.
The eradication set up, methods and technical operations
will be reported on here.
METHODS
Pre- and post-eradication monitoring
Monitoring of the two main islands’ (ET and GE)
existing seabirds, land birds, vegetation and invertebrates
was carried out for one year before the eradication and
for the subsequent three years post-eradication. The aims
of this ecosystem monitoring were to detect changes, if
possible, and hence assess the benefits of the eradication.
Full methodology and results for this will be presented
elsewhere. A population census of all seabirds, carried out
by RSPB and SNH, was undertaken at the Shiants during
June 2015, as part of SNH’s programme of Common
Standards Monitoring of protected areas (SSSIs and SPAs)
(Taylor, et al., 2018). A pre-eradication assessment site
visit was undertaken during July 2015 to finalise plans,
logistics, and health and safety requirements.
Permits and authorisations
A Habitats Regulations Appraisal (HRA) was carried
out by SNH to assess the likelihood of any adverse impact
of the rat eradication on the qualifying features of the SPA.
This required a full Appropriate Assessment under the
Conservation (Natural Habitats, &c.) Regulations 1994 (as
amended). In addition, a full assessment of the Operations
Requiring Consent (ORC) was also undertaken for the
Shiant Isles SSSI. Justification of the chosen rodenticide
(bromadiolone)
formulation,
estimated
quantity
needed, and method of application was presented in the
Appropriate Assessment and ORC application, detailing
how the operation would be undertaken across all islands
and sea stacks of the Shiants. A licence under the Wildlife
& Countryside Act (1981) was granted to cover possible
disturbance to breeding golden eagles (Aquila chrysaetos)
and white-tailed eagles which are specially protected by
Schedule 1 of the Act. Planning permission was obtained
from the Comhairle na Eilean Siar (Western Isles Council)
to allow the temporary installation of portable cabins on
the island to store rodenticide bait and provide shelter
for winter eradication teams. For the installation of two
temporary moorings, a five-year marine license (issued by
Marine Scotland) was granted for which an annual fee was
paid to the Crown Estate. Assessments of archaeological
139
Island invasives: scaling up to meet the challenge. Ch 1B Rodents: Review
sensitivity were carried out in person by experts from the
Comhairle nan Eilean Siar and also by RSPB. Maps of
archaeologically sensitive sites were used to ensure that
these features were not disturbed by placement of cabins,
bait stations or by the passage of workers around the
islands during the eradication.
A detailed health and safety plan was written in
collaboration by RSPB and WMIL. This outlined living
and working protocols and the establishment of emergency
procedures. As part of health and safety requirements,
the islands were zoned to indicate areas considered too
dangerous to access without the use of support ropes. Rope
access was hence deemed necessary to place bait and to
check bait stations on steep vegetated slopes or ground
ending abruptly at steep cliffs.
Contracts with two local boat operators (Sea Harris
Ltd and Engebret Ltd) were established in order to provide
boat access to and around the Shiants through the winter
operations. An invitation to quote was issued, with the
subsequent selection of contractors based on project needs,
cost, and suitability of boat service provision.
Rodent anticoagulant resistance tests
An assessment of potential resistance of the rats on
the Shiants to bromadiolone rodenticide was carried out
by Reading University (Vertebrate Pests Unit), using
protocols developed to extract and sequence DNA for
the identification of anticoagulant resistance mutations
in brown rats (Pelz, et al., 2005; Prescott, et al., 2010).
A similar protocol developed specifically for black rat
rodenticide resistance testing was not available. However,
the approach used represented the best option available
because of the lack of rodenticide resistance work that had
been undertaken on black rats at the time of the eradication.
Rat DNA samples were collected by project personnel from
ET in July 2015. Snap traps placed inside tunnels were set
overnight and baited with peanut butter. These were visited
early the next morning and any rat specimens caught were
collected and dissected. A portion of the tail was placed in
100% ethanol for subsequent rodenticide resistance testing.
Morphometric measurements (body length, tail length,
hind foot length, ear length) were recorded. Stomach
contents, sex and reproductive status were also assessed
for all of these trapped rats (Bell & Boyle, 2015). All DNA
samples were archived for reference in case of resistance
or reinvasion by rats at the islands.
Equipment preparation
Off-island preparation of equipment included the
construction of 0.75 m long bait stations (Fig. 2) from
lengths of 10 cm diameter plastic drain coil. Help was
sought from local community volunteers from the Isle of
Harris and construction of approximately 700 bait stations
was carried out over two days at the Harris Volunteer Centre
in Tarbert. The remaining bait stations were constructed by
project personnel off and on the Shiants. Bait stations and
other equipment were airlifted to the Shiants over two days
as part of the set-up phase of the eradication.
less accessible areas such as the Galtachan sea stacks, and
a large rock to the east of EM.
Rope access training was undertaken by seven WMIL
and RSPB personnel. Bolts were set in rock at the top of
twelve rope access routes (eleven on ET and one on EM). A
further eight routes (one on ET, two on EM and five on GE)
were accessed using ropes secured by anchors manually set
up using a series of three lashed metal stakes.
Non-target mitigation
Measures to prevent secondary poisoning of eagles
were provided by the establishment of diversionary
feeding protocols. Dead rabbits (Oryctolagus cuniculus,
collected by manual trapping at a nearby site on the Isle
of Harris) were attached to two tables located on GE, with
motion activated cameras set up to monitor activity. The
diversionary feeding was made available to the eagles from
October 30, 2015 until March 17, 2016. However, no fresh
food was attached to the table after 12 November 2015
because bait take had reduced to such low levels that the
risk of secondary poisoning was deemed negligible. It was
also noted that eagles were only intermittently present at
the islands, and there was no evidence, e.g. from motion
activated cameras, to suggest that diversionary food was
utilised by any eagle. Wire clips were fitted to all bait
stations after a raven was seen to open one and access the
bait – no further instances of non-target vertebrate species
accessing bait were observed.
Bait quantity
An estimate of the quantity of rodenticide needed
for the eradication was calculated as follows during the
planning phase. An application rate was assumed of 0.28
kg rodenticide per bait station (i.e. 10 blocks per station
in 684 bait stations on a 50 × 50 m grid; 1.12 kg/ha) for
the first four weeks, then 0.14 kg per bait station (0.56 kg/
ha) for the subsequent four weeks and 0.056 kg per bait
station (0.224 kg/ha) for the remainder of the operation,
with consumed bait replenished at each check. Each
application, or “round” of bait station checks was expected
to take one day using a team of 10 people. It was expected
to require at least 30 complete rounds (with replacement
of bait) of each station to ensure the eradication of all
the rats. At this rate, up to four tonnes of bromadiolone
(LD50, oral ingestion 1.125 mg/kg, Meehan, 1978) were
estimated to be required to cover the combined island
area (171 ha) over approximately five months. Note,
that although the stated LD50 for bromadiolone as given
by the manufacturer is 0.525 mg/kg (Bell Laboratories
Material Safety Data Sheet) this is based on laboratorybred brown rats. Wild populations of black rats may be
more tolerant to bromadiolone (Sridhara & Krishnamurthy,
1992). Individual and sex-specific variations in toxicity of
bromadiolone to black rats have also been reported (Garg
Access to challenging terrain
Camps were established on ET and EM for the winter
teams. Portable cabins were installed for safe storage of
rodenticide baits and shelter for winter eradication teams.
The flat-packed cabins were airlifted to the islands by
helicopter in October 2015 and constructed on-site. The
existing bothy on ET was re-roofed during the summer of
2015 and was also used as a base camp during the winter
operation. Two moorings were installed close to ET, to
improve safety for boat access. Boats were used to land on
140
Fig. 2 Bait station shown open, with rodenticide blocks
wired in place.
Main, et al.: Rat eradication on the Shiants
wired into bait stations (one block alongside the Contrac®
blocks) during January and February 2016 (depending on
the island) (Tables 1 and 3). This provided an alternative
bait for rats not consuming the Contrac® blocks.
Monitoring
Fig. 3 Map of the bait station grid.
& Singla, 2014). Therefore, a higher LD50 was assumed
here (Meehan, 1978), to account for these differences.
The actual application rates, bait station grid and
number of days to complete a full bait station check during
the eradication were different to those calculated in the
planning phase due to operational requirements, predicted
high rat density areas (i.e. coastal areas) and adaptation to
both bait consumption and rat behaviour.
Bait application
During baiting operations, personnel regularly
searched for carcasses, including any dead or dying rats
present at the surface. Systematic monitoring for surviving
rats commenced at the islands on November 28, 2015
and continued for 14 weeks, in tandem with baiting.
Monitoring stations were set up at every bait station and
at intervals half way between bait stations. Monitoring
tools employed included: non-toxic flavoured paraffin wax
blocks (chocolate, peanut butter, peanut essence, aniseed);
soap; tracking tunnels; snap traps and motion-activated
cameras. After 14 weeks, intensive monitoring was reduced
and permanent monitoring stations were established at key
locations (Fig. 4, Fig. 5) on the three main islands where
early detection of any surviving black rats, or an invasion
of brown rats, would be likely. These comprised 44
commercially available Protecta™ boxes and ten wooden
rodent motels baited with non-toxic chocolate wax blocks
(Fig. 4). Monitoring stations were checked in winter of
October 2016, January 2017, March 2017 and November
2017, with replacement of old blocks each time. Regular
checks were also carried out in the summer (April–August
2016–2018) during island monitoring as part of the Shiants
Seabird Recovery Project.
Island-wide baiting grids with a total of 1,183 bait
stations (Fig. 2) were established across all islands
and sea stacks (347 on ET, 594 on GE, 207 on EM, 28
on Galtachan, four on Arch Island and three on Seann
Chaisteal) during October 2015. Bait stations were spaced
at 50 m intervals across the islands, and 25 m along coasts
and through areas of boulder scree e.g. the large area of
boulders known as Carnach Mhòr on GE (Fig. 3). This grid
spacing has become the current UK best practice protocol
for rat eradications (Thomas, et al., 2017b). Rodenticide
bait and rat monitoring tools were placed on the Galtachan
sea stacks on 17 October 2015.
Following the placement of rodenticide and monitoring
tools at the Galtachan sea stacks, subsequent checks on 4
November 2015 and 27 February 2016 revealed no rat sign
or bait take. The sea stacks were assumed to be rat free
and were not treated with further rodenticide during the
eradication phase.
A team of between two and four people was stationed
on ET throughout the operation with a break of one week
over the Christmas period. A second team of between
two and three people was stationed on EM on a week on/
week off schedule. Rotation of people within these teams
occurred on a weekly basis, where weather permitted
safe access for boat transfers. Baiting commenced at the
islands on 4th November (Table 1). Cereal-based wax
blocks (28 g Contrac® All-weather Blox™ (Cas No.
28772-56-7, EU 528/2012), containing the anticoagulant
rodenticide bromadiolone at 0.005% w/w) were initially
placed loose within bait stations. Regular checks were then
made. Stations that were accessible on foot were checked
between 16 and 23 times with an average of 6–8 days
between checks (Tables 1 and 2). Rope access routes were
checked between 3–13 times with an average of 10–15 days
between checks (Tables 3 and 4). Frequency of checks at
rope access routes was limited by the availability of trained
staff on each island and the length of time required to check
relatively few bait stations on the routes. The rodenticide
blocks were wired in to bait stations starting from the
seventh round of baiting on GE, round eight on EM and
round nine on ET, with all bait wired in by 8 January
2016 (Tables 1 and 3). Records of bait application at each
station were kept in waterproof notebooks and transferred
to an electronic database each night. Bait was replaced if
damaged by weather or slugs, or at the sign of rat incisor
marks. An alternative bait, a soft block (100 g Romax® Rat
CP™ (Cas No. 5836-29-3, UK UK-2016-1003), containing
the anticoagulant coumatetralyl at 0.0375% w/w), was
RESULTS
A database of all baiting, monitoring, and other
activities was maintained throughout the eradication.
Observations of potential non-target species, carcasses, and
other relevant information were documented throughout
the operation.
Bait consumption
Consumption of bait was higher around the coasts of all
islands and was the highest around areas of known seabird
colonies (Fig. 6). Rats consumed approximately 270 kg (or
9666 blocks) of Contrac® bait in total, mainly during the
phase when blocks were loose in bait stations and available
for rats to remove whole and cache (Fig. 7). Consumption
of Romax® blocks by rats was zero.
An estimate of the numbers of rats present at the time
of the eradication was made as follows. Assumptions were
made that: each block removed by rats was consumed in its
entirety by a single rat; all rats consumed between three and
24 times the lethal dose of bromadiolone (where a lethal
dose is delivered by consuming 9.5 g of bait corresponding
to approximately one third of a block, assuming an LD50 of
1.125 mg/kg for black rats). Hence, it is estimated that there
were between 1,208 and 9,666 rats present on the Shiants
at the start of the eradication (assuming between one and
eight blocks were taken by each rat before death). The
mean (± SE) of 4.6 ± 0.1 blocks consumed per bait station
overall leads to the estimation that there were 2,099 ± 97
rats on the Shiants. This is lower than previous estimates
for black rat on the Shiants; 22–85 rats/ha (3,762–14,535
rats) in 1998 (Key, et al., 1998) and 14–27 rats/ha (2,394–
4,617) in 2012 (Bell, 2013). Methods used and timings of
each of these population estimates vary.
141
Island invasives: scaling up to meet the challenge. Ch 1B Rodents: Review
Table 1 Bait application for stations accessible by foot – bait check, dates for each bait check, quantities deployed and
schedule. All rodenticide used was Contrac® All-weather BloxTM unless indicated.
Baiting No
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Eilean an Tighe
Garbh Eilean
Eilean Mhuire
4/11/15–5/11/15
(8 blocks)
9/11/15
(8 blocks)
16/11/15–17/11/15
(8 blocks)
20/11/15
(8 blocks)
22/11/15
(8 blocks)
28/11/15–29/11/15
(8 blocks)
3/12/15–4/12/15
(8 blocks)
5/12/15–6/12/15
(2 blocks wired in place)
8/12/15–10/12/15
(2 blocks wired in place)
17/12/15–6/1/16
(2 blocks wired in place)
12/1/16
(2 blocks wired in place)
18/1/16
(2 blocks wired in place)
24/1/16
(2 blocks wired in place)
30/1/16–2/2/16
(1 block wired in place)*
6/2/16
(1 block wired in place)*
11/2/16
(1 block wired in place)*
14/2/16–5/2/16
(1 block wired in place)*
21/2/16
(1 block wired in place)*
26/2/16–27/2/16
(1 block wired in place)
4/3/16
(1 block wired in place)
8/3/16
(1 block wired in place)
12/3/16
(1 block wired in place)
22/3/16–23/3/16
(1 block wired in place)
6/11/15–7/11/15
(8 blocks)
11/11/15–15/11/15
(8 blocks)
18/11/15–19/11/15
(8 blocks)
22/11/15–27/11/15
(8 blocks)
30/11/15–2/12/15
(8 blocks)
10/12/15–15/12/15
(8 blocks)
1/1/16–4/1/16
(2 blocks wired in place)
8/1/16–11/1/16
(2 blocks wired in place)
14/1/16–16/1/16
(2 blocks wired in place)
19/1/16–21/1/16
(2 blocks wired in place)
28/1/16–5/2/16
(1 block wired in place)*
7/2/16–9/2/16
(1 block wired in place)*
12/2/16–14/2/16
(1 block wired in place)*
17/2/16–21/2/16
(1 block wired in place)*
22/2/16–23/2/16
(1 block wired in place)*
25/2/16–26/2/16
(1 block wired in place)
29/2/16–1/3/16
(1 block wired in place)
5/3/16–7/3/16
(1 block wired in place)
9/3/16–10/3/16
(1 block wired in place)
16/3/16–20/3/16
(1 block wired in place)
7/11/15
(8 blocks)
9/11/15
(8 blocks)
11/11/15–12/11/15
(8 blocks)
22/11/15–27/11/15
(8 blocks)
28/11/15–29/11/15
(8 blocks)
30/11/15
(8 blocks)
12/12/15–13/12/15
(8 blocks)
8/1/16–11/1/16
(2 blocks wired in place)
22/1/16
(2 blocks wired in place)
24/1/16–25/1/16
(2 blocks wired in place)
26/1/16
(2 blocks wired in place)
3/2/16–6/2/16
(1 block wired in place)*
7/2/16–9/2/16
(1 block wired in place)*
20/2/16–21/2/16
(1 block wired in place)*
9/3/16–11/3/16
(1 block wired in place)
14/3/16
(1 block wired in place)
* Romax® Rat CP™
Interference with bait stations by non-target species
was low. Invertebrates, particularly slugs, consumed some
bait (1.78 kg). Sheep were estimated to consume 19.6 kg of
bait released by kicking up the stations. Ravens and crows
were observed to take bait (10.9 kg) by pulling out wires
and removing bait station lids, until a more secure wire
fastening system was established. No evidence was found
of non-target species being affected by the rodenticide.
142
No carcasses showing signs of anticoagulant ingestion
were collected. An adult golden eagle carcass discovered
on 15th November was autopsied and showed no signs of
anticoagulant poisoning, and the state of decomposition
suggested it had almost certainly died before the start of the
baiting operation. Diversionary food provided throughout
the operation was not removed by any species.
Main, et al.: Rat eradication on the Shiants
Table 2 Frequency of replenishment of bait stations
accessible by foot. Mean number of days (± SE)
between the first day of each bait station check; range in
number of days to complete check and total number of
checks given in parentheses.
Island
Eilean an Tighe
Garbh Eilean
Eilean Mhuire
Total (all islands
combined)
Number of days between the first
day of each bait station check
6.2 ± 1.0 days
(2–27 days; 23 checks)
6.7 ± 1.0 days
(3–22 days; 20 checks)
8.1 ± 1.8 days
(2–27 days; 16 checks)
6.9 ± 0.7 days
(2–27 days; 16–23 checks)
Table 3 Frequency of replenishment of bait stations
accessed by rope. Mean number of days (± SE) between
the first day of each bait station check; range in number
of days to complete each check and total number of
checks given in parentheses.
Island
Eilean an Tighe
Garbh Eilean
Eilean Mhuire
Total (all islands
combined)
Number of days between the first
day of each bait station check
10.8 ± 1.7 days
(4–24 days, 13 checks)
13.5 ± 2.3 days
(2–23 days,11 checks)
14.4 ± 3.2 days
(2–29 days, 9 checks)
12.7 ± 1.3 days
(2–29 days; 9–13 checks)
Table 4 Bait application for stations accessed by rope – bait check, dates for each bait check, quantities deployed and
schedule. All rodenticide used was Contrac® All-weather Blox™ unless indicated.
Bait No
1
2
3
4
5
6
7
8
9
10
11
12
13
Eilean an Tighe
Garbh Eilean
Eilean Mhuire
17/11/15–29/11/15
(8 blocks)
5/12/15
(2 blocks wired in place)
12/12/15–15/12/15
(2 blocks wired in place)
5/1/16–6/1/16
(2 blocks wired in place)
15/1/16
(2 blocks wired in place)
23/1/16–2/2/16
(1 block wired in place)*
11/2/16–13/2/16
(1 block wired in place)*
15/2/16
(1 block wired in place)*
25/2/16
(1 block wired in place)*
4/3/16–6/3/16
(1 block wired in place)
8/3/16
(1 block wired in place)
16/3/16
(1 block wired in place)
22/3/16–24/3/16
(1 block wired in place)
19/11/15–3/12/15
(8 blocks)
6/12/15–7/12/15
(8 blocks)
17/12/15
(8 blocks)
2/1/16–4/1/16
(2 blocks wired in place)
14/1/16
(2 blocks wired in place)
16/1/16–21/1/16
(2 blocks wired in place)
8/2/16–14/2/16
(1 block wired in place)*
21/2/16
(1 block wired in place)*
29/2/16
(1 block wired in place)
7/3/16
(1 block wired in place)
17/3/16
(1 block wired in place)
11/11/15
(8 blocks)
23/11/15
(8 blocks)
14/12/15–15/12/15
(8 blocks)
8/1/16–9/1/16
(2 blocks wired in place)
6/2/16
(1 block wired in place)*
8/2/16–9/2/16
(1 block wired in place)*
22/2/16
(1 block wired in place)*
11/3/16
(1 block wired in place)
14/3/16
(1 block wired in place)
* Romax® Rat CP™
Bait stations near to coasts were affected by weather,
with 60 washed away during large storms (resulting in a
loss of 9.5 kg of bait). A total of 84 kg of bait (from 583
stations) was removed by hand because of damage by
mould or dampness that could have rendered it unpalatable
to rats.
Rat sign and monitoring
Four rats were found dead at the surface on GE (one
fresh carcass on 7 November 2015, four days after baiting
commenced); two fresh carcasses on 11 November 2015
(seven days after baiting commenced) and one desiccated
carcass on 4 February 2016 (90 days after baiting
commenced). Rat sign was recorded on flavoured wax at
three monitoring points following the start of the initial
baiting phase. These were at three different cliff stations on
ET (on 13 December, on 14 December and on 26 February).
Monitoring at these cliff stations through March 2016 did
not yield any further rat sign. The detection of incisor
marks on the flavoured wax block in February 2016, when
other food sources were beginning to become available,
was treated as a possible survivor from the eradication. To
establish whether any rats had survived, in October 2016
rope access teams re-established 470 monitoring stations
143
Island invasives: scaling up to meet the challenge. Ch 1B Rodents: Review
Fig. 5 Locations of permanent monitoring stations.
Fig. 6 Quantities of bait taken by rats during the eradication.
Fig. 4 Rodent motel (lower) and ProtectaTM box (upper)
monitoring stations.
across the three islands with a focus on the cliff stations on
ET. Four checks were completed on EM and ET and three
checks on GE, finding no sign of rat presence. Permanent
monitoring stations on ET, GE and EM were checked
monthly in the summer (April–August) and every three
months outside the summer season until February 2018.
A month-long intensive monitoring check was carried out
on all islands and sea stacks in February 2018, with the
declaration of rat-free status made on 2 March 2018.
Rodenticide resistance testing
Although some mutations were present within the
section of genome sequenced, these mutations were not
144
Fig. 7 Quantity (kg) of bait consumed by rats following
each baiting application on each island. Solid black line
= Garbh Eilean; Dashed line = Eilean an Tighe; Grey
line = Eilean Mhuire.
the same as brown rat mutations known to confer genetic
resistance to bromadiolone.
DISCUSSION
Eradication of black rats at the Shiant Isles was one of
the four core aims of the Shiants Seabird Recovery Project,
now successfully achieved.
The Shiants black rat eradication was an ambitious
undertaking on a remote, uninhabited island group, with no
existing infrastructure or facilities except the bothy on ET.
The operation required the establishment of safe working
environments and the provision of shelter, for example,
Main, et al.: Rat eradication on the Shiants
on EM where there are no existing structures on an island
of 80 m height. The lack of sanitation, water supply and
electricity necessitated basic living conditions and robust
and appropriate waste management procedures. Weather
influenced operations throughout, sometimes becoming so
severe as to prevent access to various parts of the island
group. The terrain of the Shiant Isles is steep across many
sections of each island, hence the need for rope access
at various points. The use of extensive rope access, in
particular, allowed a neat solution to the problems of
challenging access and a ground-based operation. Other
methods, such as bait stations deployed at the end of lines,
may have been more difficult to monitor and would also
have necessitated the close approach to cliff edges (also
requiring support ropes in order to comply with health and
safety requirements). As well as the challenging access
on foot, the need to address separate islands at the same
time by boat presented further logistical and personnel
considerations.
New challenges had arisen since the feasibility study
(2012) including changes in the profile of the boulder
causeway connecting ET and GE resulting in reduced
access on foot to GE from the main camp on ET. White
tailed eagles had established a breeding site at the islands
and mitigation actions were required to address potential
secondary poisoning of these predators, which had not
been necessary at the time of the feasibility study. This
highlights the need to review feasibility studies as an
ongoing process within the planning phase.
A number of valuable lessons were learnt during the
course of the eradication. In light of the challenges faced
by working in difficult conditions, the operation was
delivered to a high standard by an effective team. Rope
access elements worked well as a result of the thorough
training, and of equipment and safety considerations
which were appropriate to the operation. Once established,
procedures concerning training, preparation, boating and
accommodation all worked well because conditions had
been considered thoroughly as part of a detailed health and
safety plan. Boat access arrangements allowed sufficient
access to the separate islands to achieve baiting and
monitoring throughout. The whole operation provided
positive input into the nearby economy of Lewis and
Harris over the winter months. At the end of the eradication
phase, the establishment of permanent monitoring stations
provided early detection capability, which is necessary as
part of delivering long term biosecurity at the islands.
Periods of heavy workload for personnel involved in
eradication preparations resulted from time and resource
pressures during the preparation phase. There was a need
for careful planning of logistics, the satisfying of legal
obligations, the need to train local personnel and set up
health and safety. As a result, UK-based capacity for
undertaking eradications for conservation purposes has
been greatly enhanced, and the need for detailed planning
from early on has been highlighted. A dedicated logistics
coordinator would have been a useful additional staff
member to have had in place.
Technical rope access training required a further
investment of time and resources and, although the number
of trained personnel was sufficient to be able to carry out
checks, this did limit the total number of checks possible.
Successful communication between team leaders and
volunteers took place regularly throughout the operation
but has been noted as an area in which continued focus is
important in complex operations. Lessons learnt from this
eradication will form part of a full project review planned
by RSPB.
A lack of work on the genetic resistance of black rats
to bromadiolone posed a potential problem, since it was
not possible from the start to confidently predict whether
alternative bait types might be needed. However, the
consistent lack of rat sign across the islands following
baiting with bromadiolone, and zero take of the alternative
rodenticide, indicates that there was no genetic resistance
of rodents to bromadiolone at the Shiant Isles.
The project has contributed to building UK capacity
for delivering rat eradications, biosecurity and incursion
response through its training of staff. Local community
members at the Western Isles were involved in bait
station assemblage, service provision (e.g. boats) and
volunteer work during eradication operations. Providing
safe breeding habitat and maintaining rodent-free status at
important island sites will be an important part of the longterm legacy of protection for UK seabirds.
ACKNOWLEDGEMENTS
The Shiants Seabird Recovery Project is a partnership
between the Nicolson family as island owners, RSPB
and SNH. Funding was provided by the EU LIFE+ fund
(LIFE13/NAT/UK/000209 LIFE+ SHIANTS), SNH and
RSPB.We are extremely grateful to all volunteers engaged
in preparations for the eradication and at the islands
during the campaign. Eradication volunteers were: Stuart
Burbidge, Alister Clunas, Alex Cropper, Cindy DenaroBarrett, Molly Heal, James Hedges, Amy King, Tara Proud,
Tegan Newman, Sandy Ogilvie, Ross Packman and Fritha
West. Kate Langley, and volunteers (Phil Bertin, Shona
Bertin, Heather Board, Cindy Denaro-Barrett, Richard
Denaro-Barrett, Dan Fitzsimmons, Fiona Lassen, Caroline
Magne, John Masterton, Anne Milne, Marie Newman and
Liz Struthers) assisted by RSPB staff (Vicky Anderson,
Laura Robertson and Niall Currie) helped to create
hundreds of bait stations at the Harris Volunteer Centre,
Tarbert. Paul St Pierre and Jaclyn Pearson assisted at the
islands during the set up of the bait station grid.
The help of Reading University Vetebrate Pest Unit,
who carried out rodenticide resistance testing work, is
gratefully acknowledged. Jill Harden (RSPB) and Kevin
Murphy (Comhairle nan Eilean Siar) provided essential
archaeological assessment and advice. Boat operators
Engebret Ltd and Sea Harris Ltd provided transport and
logistical support throughout the operation. Air-lifting
of bulk equipment to the Shiants was carried out at the
beginning of the operation by PDG helicopters. Remote
rope access technician training was provided by Stuart
Johnson of Climb Industries Ltd and co-workers, and
Adam Long of Access Techniques Ltd (at time of final
check for rats). Lee Ashton provided key guidance during
the establishment of health and safety procedures. We
would like to acknowledge the generous support of Bell
Laboratories for providing bait and traps.
REFERENCES
Atkinson, I.A.E. (1985). ‘The Spread of Commensal Species of Rattus
to Oceanic Islands and their Effects on Island Avifaunas’. In: P.J.
Moors (ed.) Conservation of Island Birds, pp. 35–81. ICBP Technical
Publication 3. Cambridge, UK: BirdLife International.
Bell, E. (2013). Shiant Isles Project: Feasibility Study and Operational
Plan for the Eradication of Black Rats (Rattus rattus) from the Shiant
Isles, Hebrides, Scotland. Unpublished report. Blenheim, New Zealand:
Wildlife Management International Ltd.
Bell, E. and Boyle, D. (2015). Black Rat Eradication, Shiants Seabird
Recovery Project, Shiant Isles, Scotland. Pre-assessment Visit 23 June–
9 July 2015. Unpublished report. Blenheim, New Zealand: Wildlife
Management International Ltd.
Bell, E.A. (2019). ‘It’s not all up in the air: the development and use of
ground-based rat eradication techniques in the UK’. In: C.R. Veitch,
M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) Island
invasives: scaling up to meet the challenge, pp. 79–87. Occasional
Paper SSC no. 62. Gland, Switzerland: IUCN.
145
Island invasives: scaling up to meet the challenge. Ch 1B Rodents: Review
Foster, P (2000). ‘Shiant Islands Project.’ In: P. Milburn (ed.) Discovery
and Excavation in Scotland NS Vol. 1 Council for Scottish Archaeology,
pp. 95–96. Edinburgh, UK: Archaeology Scotland.
Garg, N. and Singla, N. (2014). ‘Toxicity of second generation
anticoagulant bromadiolone against Rattus rattus: individual and sex
specific variations’. Cibtech Journal of Zoology 3(2): 43–48.
Gibb, F.G.F. and Henderson, C.M.B. (1996). ‘The Shiant Isles main sill
structure and mineral fractionation trends’. Mineralogical Magazine 60:
67–97.
Jones, H.P., Tershy, B.R., Zavaleta, E.S., Croll, D.A., Keitt, B.S.,
Finkelstein, M.E. and Howald, G.R. (2008). ‘Severity of the effects
of invasive rats on seabirds: A global review’. Conservation Biology
22(1):16–26.
Key, G. Fielding, A.H. Goulding, M.J. Holm, R.S. and Stevens-Woods,
B. (1998). ‘Ship rats Rattus rattus on the Shiant Islands, Hebrides,
Scotland’. Journal of Zoology 245: 228–233.
Love, J.A. (1983). The Return of the Sea Eagle. Cambridge, UK:
Cambridge University Press.
Meehan, A.P. (1978) ‘Rodenticidal activity of bromadiolone – a new
anticoagulant’. In: W.E Howard and R.E. Marsh (eds.) Proceedings of
the Eighth Vertebrate Pest Conference 1978, pp. 122–126. Davis, CA:
University of California.
Mitchell, P.I., Newton, S., Ratcliffe, N. and Dunn, T.E. (eds.) (2004).
Seabird Populations of Britain and Ireland: results of the Seabird 2000
census (1998–2002). London, UK: T and A.D. Poyser.
Murray, S. and Shewry, M. (2002). A Baseline Census of Manx
Shearwaters on Rum National Nature Reserve in 2000 and 2001.
Commissioned Report to Scottish Natural Heritage, Inverness, UK:
Scottish Natural Heritage.
Pelz, H-J., Rost, S., Hünerberg, M., Fregin, A., Heiberg, A-C., Baert,
K., MacNicoll, A.D., Prescott, C.V., Walker, A-S., Oldenburg, J. and
Müller, C.R. (2005). ‘The genetic basis of resistance to anticoagulants
in rodents’. Genetics 170: 1839–47.
Prescott, C.V., Buckle, A.P., Gibbings, J.G., Allan, E.N.W., and Stuart,
A.M. (2010). ‘Anticoagulant resistance in Norway rats (Rattus
norvegicus Berk.) in Kent – a VKORC1 single nucleotide polymorphism,
tyrosine139phenylalanine, new to the UK’. International Journal of
Pest Management 57(1): 61–65.
Ratcliffe, N., Mitchell, I., Varnham, K., Verboven, N. and Higson, P.
(2009). ‘How to prioritise islands for the benefit of petrels: A case study
of the UK, Channel Islands and Isle of Man’. Ibis 151: 699–708.
Sridhara, S. and Krishnamurthy, T.R. (1992). ‘Potentiation of anticoagulant
toxicity to Rattus rattus by two non-steroid anti-inflammatory drugs’.
In: J.E. Borrecco and R.E. Marsh (eds.) Proceedings of the Fifteenth
Vertebrate Pest Conference 1992, pp. 212–217. Davis, CA: University
of California.
Stapp, P (2002). ‘Stable isotopes reveal evidence of predation by ship
rats on sea birds on the Shiant Islands, Scotland’. Journal of Applied
Ecology 39: 831–840.
Taylor, P., MacMinn, R., Sophie Marsh, S., Brooke, M.de L., Macdonald,
M., Hughes, P., Donald, F., Docherty, H., Over, A., Beaton, J., Scridel,
D., Robertson, L. and Currie, N. (2018). Seabird population trends on
the Shiant Isles, Outer Hebrides, 2000–2015. Scottish Birds 38(1): 3–14.
Thomas, S., Brown, A., Bullock, D., Lock, L., Luxmoore, R., Roy, S.,
Stanbury, A. and Varnham, K. (2017a). ‘Island restoration in the UK –
past present and future’. British Wildlife 28: 231–242.
Thomas, S., Varnham, K., and Havery, S. (2017b). Current Recommended
Procedures for UK (Bait Station) Rodent Eradication Projects (Version
4.0). Sandy, Bedfordshire: Royal Society for the Protection of Birds.
Towns, D.R., Atkinson, I.A.E. and Daugherty, C.H. (2006). ‘Have
the harmful effects of introduced rats been exaggerated?’ Biological
Invasions 8: 863–91.
Walker, M.A. (1930). ‘The geology of the Shiant Isles’. Quarterly Journal
of the Geological Society 36: 355–398.
146
S. Poncet, K. Passfield, A. Kuepfer and M.A. Tabak
Poncet, S.; K. Passfield, A. Kuepfer and M.A. Tabak. The effect of Norway rats on coastal waterbirds of the Falkland Islands: a preliminary analysis
The effect of Norway rats on coastal waterbirds of the Falkland Islands:
a preliminary analysis
S. Poncet1, K. Passfield2, A. Kuepfer3 and M.A. Tabak4,5
Antarctic Research Trust, P.O. Box 685, Stanley, Falkland Islands. <sallyponcet@horizon.co.fk> 2Island LandCare,
P.O. Box 538, Stanley, Falkland Islands. 3Department of Natural Resources, Falkland Islands Government, Stanley,
Falkland Islands. 4Quantitative Science Consulting, LLC, 765 North 10th St., Laramie, WY 82072 USA. 5Department of
Zoology & Physiology, University of Wyoming, 1000 E. University Ave., Laramie, WY 82071 USA.
1
Abstract The Falkland Islands have been affected by anthropogenic-induced habitat modification including introduction
of invasive species and grazing by livestock. Introduced Norway rats are known to have a large effect on native Falklands
passerines but their effect on other native birds has not been explored. We investigated the effects of several environmental
variables, including the presence of Norway rats and chronic grazing by livestock, on an assemblage of 22 species of
coastal waterbirds by comparing species richness and relative abundance of birds among 65 rat-infested islands, 29 rateradicated islands and 76 historically rat-free islands. Bird count data from 299 km of coastline were used to estimate
relative bird abundance, expressed as the number of individuals per kilometre of coastline for each species. Our study
provided three key results. First, coastal waterbird abundance on islands historically without rats was twice as high as that
on islands where rats were present. Second, bird abundance on rat-eradicated islands was intermediate between that of
historically rat-free and rat-infested islands. Third, habitat modification by grazing appeared to reduce bird abundance in
both rat-free and rat-infested habitats. From a conservation perspective, this study suggests that rat eradication programmes
in the Falkland Islands are effective at restoring coastal waterbird abundance and would be even more so if carried out in
conjunction with restoration of native coastal plant communities.
Keywords: cat, ecotype, eradication, grazing, marine, mouse, Norway rat, Poa flabellata, relative abundance, tussac,
warrah
INTRODUCTION
The Falkland Islands are an archipelago of 477 islands
(Falkland Islands Government, 2014) that differ in size,
habitat modification and presence of introduced species.
This creates unique opportunities to examine the effect
of anthropogenic factors and stochastic events on the
distribution and abundance of native species (Hall, et al.,
2002; St. Clair, et al., 2011).
Human colonisation of the islands in the late 1700s led
to considerable changes in native flora and fauna. Grazing
by livestock caused a reduction of almost 80% in the
coastal grasslands of tussac (Poa flabellata) (Strange, et al.,
1988), which greatly affected bird species and populations
(Strange, 1992; Woods, 1984; Woods & Woods, 2006).
Major changes to bird populations are also attributed to
the introduction of Norway rats (Rattus norvegicus), black
rats (R. rattus), house mice (Mus domesticus), cats (Felis
catus) and Patagonian foxes (Lycalopex griseus) (Woods
& Woods, 2006; Falklands Conservation, 2006). Norway
rats are now present on about half of the archipelago’s
islands (Tabak, et al., 2015a) and are known to have a large
effect on the abundance and diversity of Falklands’ native
passerine species (Hall, et al., 2002; Tabak, et al., 2015b),
while the black rat has been recorded on one island only.
Cats are known to prey on thin-billed prions (Pachyptila
belcheri) (Matias & Catry, 2008), although nothing is
known of their impact on other Falkland bird species. Mice
impact small burrowing petrels and some passerine species
(Rexer-Huber, et al., 2013) and Patagonian foxes reduce
the breeding success of coastal waterbirds (Poncet, 1998).
Prior to the arrival of these invasive predators, the only
terrestrial mammal was the endemic warrah or Falklands
wolf (Dusicyon australis). Restricted to the two largest
islands (East Falkland and West Falkland), this native
canid may have been present for at least 70,000 years
before being exterminated in 1876; craniodental evidence
and first-hand accounts indicate that the warrah was an
efficient predator, subsisting on penguins, geese and seals
(Slater, et al., 2009). Beyond all doubt, it must have had
a major impact on the distribution and abundance of all
wildlife species on East Falkland and West Falkland.
The effects of introduced predators and their removal
on passerine and seabird species have been relatively well
documented for many islands world-wide (Ebbert & Byrd,
2002; Courchamp, et al., 2003; Rauzon, 2007; Kurle, et
al., 2008; Towns, 2011; Veitch, et al., 2011), including the
Falkland Islands (Woods, 1970; Strange, 1992; Woods &
Woods 1997; Hilton & Cuthbert, 2010; Poncet, et al., 2011;
Tabak, et al., 2014) where the eradication of Norway rats
from 80 islands since 2001 provides a large-scale experiment
for evaluating wildlife response to the removal of rats.
Studies have shown that successful eradications result in
a higher species richness of passerines (Hall, et al., 2002)
and an increase in abundance of both passerines (Tabak,
et al., 2015b) and invertebrates (St. Clair, et al., 2011).
However, nothing is known of the impact of introduced
predators on Falkland’s shags, gulls, wildfowl, waders and
birds of prey (referred to hereafter as coastal waterbirds) or
of the response of these birds to rat eradication.
The evaluation of the subsequent recovery of native
species and ecosystems following alien species eradication
is an essential part of the process of determining the success
of an operation (Courchamp, et al., 2011). In this study,
we examined the potential effect of several environmental
variables on the distribution and abundance of a coastal
waterbird assemblage consisting of 22 species of groundnesting, coastline-foraging birds. Using estimates of
relative bird abundance (individuals per unit of coastline
transect length) and species richness as the response
variables, we compared bird populations on tussac islands
that were historically rat-free, with those that were ratinfested and those where rats had been eradicated, and we
examined bird response to habitat modification by grazing.
We hypothesise that (1) the presence of rats would reduce
bird abundance and species richness; (2) bird abundance
on islands where rats had been eradicated would be higher
than on rat-infested islands; and (3) long-term grazing of
native habitats would reduce bird abundance and species
richness.
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 147–153. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
147
Island invasives: scaling up to meet the challenge. Ch 1B Rodents: Review
METHODS
Study area
Study species
The Falkland Islands archipelago (12,200 km ) is
situated approximately 500 km east of continental South
America between latitudes 51°S and 53°S in the South
Atlantic Ocean. It consists of the two large island masses
East Falkland (6,480 km2) and West Falkland (4,450 km2),
475 smaller vegetated offshore islands (an island being
defined as any vegetated land surrounded by water at low
tide) and numerous associated rocks and stacks (Falkland
Islands Government, 2014), totalling over 6,000 km of
coastline. Surveys of all bird species were conducted
on 168 offshore islands, and on East Falkland and West
Falkland. The majority of islands included in this study
were remote and uninhabited, requiring access by boat.
2
Coastlines surveyed were typical of most rocky
offshore tussac islands with upper littoral and intertidal
zones of gently to moderately sloping rock, shingle or
boulder, occasional small sand beaches and short stretches
of sheer cliff on exposed coasts (Strange, 1992). Coastline
vegetation on ungrazed islands was dominated by the
native grass tussac (Poa flabellata), which grows up to
3 m tall and forms dense canopies. On islands grazed by
livestock, plant communities were dominated by short
swards of grasses and herbs.
Data collection
Surveys were carried during the breeding season
(September to May) between 2008 and 2014. All surveys
were conducted in favourable weather conditions and
by the same two observers (S. Poncet and K. Passfield).
Environmental characteristics for each transect and each
island, and the identity of each bird species and the number
of birds detected were recorded following a standardised
data collection protocol (Tabak, et al., 2015b). Surveyors
walked along the coastline at a slow and consistent pace,
noting birds that moved ahead or accompanied the surveyor
to avoid counting the same bird multiple times. Counts
were of adults and subadults; breeding status and social
structure were also recorded. The geographical location of
individuals, pairs and groups of birds was recorded using
a hand-held global positioning system (GPS) receiver
(Garminmap 62). Introduced predators were detected
visually or by field signs typical for each species.
The sampling unit (transect) on each island consisted
of a 100 m-wide swathe of coastline extending from
approximately 20 m inland of the high tide mark out to
approximately 80 m offshore. The inland distance of 20
m was determined on the basis that this is the maximum
distance at which most birds would be visible or heard by
an observer walking along the shoreline. Transect length
was obtained from the surveyor’s GPS track. Surveys
involved walking at least 1 km of coastline.
For each transect we recorded the date, local time
at the start and end of survey, transect length, observer
name, wind speed and direction, cloud cover, temperature,
precipitation, tide state, geographic region, grazing
intensity and dominant vegetation.
For each island we recorded rat eradication status,
island surface area and percent of island covered in tussac
using data sourced from the Falkland Islands Biodiversity
Database (Falkland Islands Government, 2014). Island
coastline perimeter was obtained either from the observer’s
GPS track data or by using mapping software to measure
coastlines on map sheets 1–29 (Directorate of Overseas
Surveys, 1962).
148
All bird species (native, non-native, resident, vagrant
and migratory) encountered on transects were recorded.
The number of individuals of most species was also
counted. Most shoreline species are of high to very high
detectability and occur in habitats that are generally open
to view from long distances (Woods & Woods, 1997), there
being no trees or woodlands on the islands. Species that
were not easily detected (burrowing petrels which nest
underground) were not counted, and nor were penguins
and black-browed albatross for which the coastline survey
methodology was not suitable due to the amount of time
required to count large concentrations of colonial seabirds.
In this study, 22 shoreline species were grouped to
form an assemblage called coastal waterbirds. These
were defined as any non-passerine species that relies on
the littoral and sub-littoral zones for breeding, wintering
and/or foraging, and they include waterfowl, waders, a
bird of prey and some seabirds (Table 1). The majority of
species are common and widespread around Falklands’
coasts and forage predominantly on shoreline and inshore
coastal habitats (Woods & Woods, 1997). Species detected
on surveys that did not conform to this definition were
passerines, southern giant petrels (Macronectes giganteus),
upland geese (Chloephaga picta) and ruddy-headed geese
(C. rubidiceps). The five introduced mammalian species
recorded were feral cats, black rats, Norway rats, house
mice and Patagonian foxes.
Data analysis
Relative abundance and species richness
Relative abundance of each coastal waterbird species
for all 170 islands surveyed was estimated as the number of
individuals counted divided by the total length of transects
(in kilometres) walked on that island. Species richness
of an island was defined as the total number of coastal
waterbird species detected in transects on each island.
Ecotypes
The 170 islands were grouped into six different ‘ecotype’
categories based on several specific environmental
variables (Table 2). Variables were the presence or absence
of Norway rats (distinguishing between historically ratfree and rat-free following successful rat eradication), the
presence or absence of heavy grazing, and the presence or
absence of tussac-dominant vegetation along an island’s
coastline. The relative abundance and species richness
for the coastal waterbirds was calculated for each of these
ecotypes.
Effect of environmental variables on bird abundance and
species richness
We modelled the relationship between response
variables (total relative abundance and species richness of
the coastal waterbird assemblage) and a number of potential
driving environmental variables for a subset of 139 islands
using Generalised Linear Models (GLMs). The predictor
variables included in these models were the presence or
absence of rats (excluding islands where rats had been
eradicated), the presence or absence of heavy grazing,
the percent of an island covered in tussac grass (“tussac
cover”), island coastline perimeter and weather and tide at
the time of survey. Data from the two largest islands East
Falkland and West Falkland were excluded from these
GLM analyses because the difference in coastline perimeter
Poncet, et al.: Rats and waterbirds, Falkand Islands
Table 1 Summary of the abundance of all 42 passerine and non-passerine bird species recorded on coastal
transects on 170 islands in the Falkland Islands.
Species
Falkland Island steamer duck* Tachyeres bracypterus
Kelp goose* Chloephaga hybrida
Blackish oystercatcher* Haematopus ater
Rock shag* Phalacrocorax magellanicus
Crested duck* Anas specularioides
Kelp gull* Larus dominicanus
Turkey vulture* Cathartes aura
Magellanic oystercatcher* Haematopus leucopodus
Tussacbird Cinclodes antarcticus
Upland goose Chloephaga picta
Black-crowned night-heron* Nycticorax nycticorax
Austral thrush Turdus falcklandii falcklandii
Black-chinned siskin Carduelis barbata
Dark-faced ground-tyrant Muscisaxicola maclovianus
Grass wren Cistothorus platensis
Dolphin gull* Leucophaeus scoresbii
Black-throated finch Melanodera melanodera
Striated caracara* Phalcoboenus australis
Cobb’s wren Troglodytes cobbi
Snowy sheathbill* Chionis albus
King cormorant* Phalacrocorax albiventer
Long-tailed meadowlark Sturnella loyca
Ruddy-headed goose Chloephaga rubidiceps
Brown-hooded gull* Larus maculipennis
Southern giant petrel Macronectes giganteus
South American tern* Sterna hirundinacea
Falkland skua* Catharacta antarctica
Southern caracara* Caracara plancus
Two-banded plover* Charadrius falklandicus
Magellanic snipe* Gallinago magellanica
Speckled teal* Anas flavirostris
White-rumped sandpiper* Calidris fuscicollis
Rufous-chested dotterel* Charadrius modestus
Correndera pipit Anthus correndera grayi
Red-backed hawk Buteo polyosoma
White-tufted grebe* Rollandia rolland
Peregrine falcon Falco peregrinus
Silver teal Anas versicolor fretensis
Chiloe wigeon Anas sibilatrix
House sparrow Passer domesticus
Domestic goose Anser anser
Cattle egret Bubulcus ibis
Total
No. birds
counted
5,432
2,400
752
5,668
2,143
1,726
687
1,380
2,378
1,295
670
378
541
364
242
762
302
262
645
796
3,969
168
145
404
812
687
97
35
332
31
68
292
71
14
7
5
3
10
4
68
2
1
36,048
No. of
islands
170
144
140
131
133
125
120
105
95
95
93
84
80
70
68
60
66
55
52
51
43
37
35
34
33
29
27
24
21
17
16
11
8
6
6
4
3
2
2
1
1
1
170
% of
islands
100
85
82
77
75
74
71
62
56
56
55
49
47
41
40
35
33
32
31
30
25
22
21
20
19
17
16
14
12
10
9
6
5
<5
<5
<5
<5
<5
<5
<5
<5
<5
Species ranked by frequency of occurrence on the 170 islands surveyed.
* Species in the coastal waterbird assemblage analysed in this study.
149
Island invasives: scaling up to meet the challenge. Ch 1B Rodents: Review
Table 2 A comparison of relative abundance (mean ± s.e.) of the coastal waterbird assemblage counted
using standardised surveys on 170 islands and six ecotypes during the period 2008–2014.
Island ecotype
I ‘mainland’1
II rat-infested tussac2
III rat-free tussac3
IV rat-eradicated tussac
V rat-infested non-tussac5
4
VI rat-free non-tussac6
Total
No. birds
counted
1,387
6,051
Birds/km
2
57
Km of
coastline
44.98
82.27
No.
transects
18
64
31 ± 2.6
74 ± 4.8
11,775
5,521
156 ± 14.2
138 ± 9.8
70
29
75.49
40
81
32
2,642
1,257
28,633
60 ± 4
101 ± 7.3
6
6
170
43.97
12.41
299.12
23
10
228
No. islands
1
Mainland East and West Falkland: grazed and/or massively modified by past grazing; Norway rats, mice and
cats present; no Patagonian foxes.
2
Tussac islands with Norway rats: tussac dominant along the coastline; not permanently grazed and/or not
massively modified by past grazing; no cats, no ship rats, no Patagonian foxes, no mice.
3
Tussac islands without Norway rats: tussac dominant along the coastline; not permanently grazed and/or not
massively modified by past grazing; no cats, no ship rats, no Patagonian foxes, no mice.
4
Tussac islands where Norway rats have been eradicated: tussac dominant along the coastline; not
permanently grazed and/or not massively modified by past grazing; no cats, no ship rats, no Patagonian foxes,
no mice.
5
Non-tussac islands with Norway rats: little or no tussac; permanently grazed and/or massively modified by
past grazing; no cats, no ship rats, no Patagonian foxes, no mice.
6 Non-tussac islands without Norway rats: little or no tussac; permanently grazed and/or massively modified by
past grazing; no cats, no ship rats, no Patagonian foxes, no mice.
between these islands and the smaller islands resulted in a
disproportionate influence on model output. The full model
contained all covariates of interest, and an intercept term
and was described as where the response variable was
either relative abundance or species richness of the coastal
waterbird assemblage. A Gaussian distribution with an
identity link function was used to describe the relative
abundance response variable while a Poisson distribution
with a log link function was used to describe the species
richness variable. We also ran all possible reduced models,
containing all possible combinations using these covariates,
and calculated Akaike Information Criterion corrected for
small sample size (AICC) for each model. For each set of
models, we conducted model averaging on the best models
(those with DAICC < 7), to obtain a single ensemble model
for relative abundance and species richness (Burnham &
Anderson, 2002). All calculations and modelling were
conducted using R v.3.2.2 (R Core Team, 2015).
The individual effects of each of the above covariates
on the response variable were also analysed. Discrete
covariates (rats and grazing) were analysed using Welch’s
two sample t-test; continuous covariates (percent tussac
cover and coastline perimeter) were analysed using
Analysis of Variance (ANOVA).
The effect of Norway rat status on bird abundance and
species richness
We also analysed the relationship between response
variables and the potential driving environmental variable
of Norway rat status (rat-infested, rat-free and rateradicated) for a subset of 155 tussac islands of less than 10
km perimeter and less than 200 ha. Bird data for ecotypes
II, III and IV (Table 2) were analysed using a KruskalWallis test with a Bonferroni post-hoc test to assess the
effect of rat eradications on bird abundance. A one-way
ANOVA with Tukey HSD post-hoc test was used to assess
ecotype effect on species richness (Table 3).
RESULTS
Data overview
A total of 42 bird species (of which 22 were classed
as coastal waterbirds) and 36,000 individual birds were
recorded along 299.12 km of coastal transects on East
Falkland, West Falkland and 168 offshore islands (Table 1;
Fig. 1). The majority (156) of these offshore islands were
predominantly tussac-covered, uninhabited and ungrazed;
the other 12 islands had little or no tussac cover and were
Table 3 Results of the Kruskal-Wallis test with Bonferroni post-hoc tests of effects of rat status on relative
abundance and of the one-way ANOVA post-hoc tests of effects on species richness of coastal
waterbirds, on 155 tussac islands of which 70 were historically rat-free, 57 were rat-infested at the time
of survey and 28 had been successfully cleared of rats.
150
Coastal
waterbird
Test used
Relative
abundance
Kruskal-Wallis
with Bonferroni
Species richness
ANOVA with
Tukey HSD
Kruskal-Wallis and Post hoc results
ANOVA results
(p-value; difference in mean)
Rat-infested vs
Rat-infested vs
rat-free
rat-eradicated
2
χ = 166.339
0.000;
0.008;
df = 153.2
-3.866
-2.794
p = 0.000
F (153.2) = 2.715
0.485;
0.056;
p = 0.069
0.540
1.417
Poncet, et al.: Rats and waterbirds, Falkand Islands
Table 4 Model-averaged estimates for each parameter on relative abundance and species richness of the coastal waterbird
assemblage on 139 islands for 214.15 km of coastline surveyed. Low magnitude of covariate values (i.e. values that are
close to zero) indicate that the variable has a weak or insignificant effect on the response, while high absolute values
indicate that the variable is important in predicting the response.
Intercept
Relative abundance
Species richness
140.07
8.07
Model-averaged estimates for coastal waterbirds
heavy grazing
coastline
rats (present) (present)
perimeter (km)
-69.33
0.04
-4.46
-0.01
-0.13
1.34
percent tussac
cover
0.28
0.00
cats, mice, Norway rats, permanent human settlements and
year-round grazing by sheep and cattle. None of the islands
had black rats or Patagonian foxes.
Fig. 1 The Falkland Islands showing islands where coastal
bird surveys were undertaken.
grazed year-round by sheep (Table 2). Offshore islands
ranged in size from 0.1 to 5,600 ha. The average length
of coastline surveyed was 1.3 km; the entire coastline
perimeter was surveyed on 58 of the smallest islands of
less than 125 ha. Seventy six islands had one predator only
(Norway rat); 63 had never had any introduced predators;
29 had been formerly occupied by one predator only
(Norway rat) which had been subsequently eradicated and
was absent at the time of bird survey. East Falkland and
West Falkland, (collectively referred to as ‘mainland’) had
Coastal waterbirds accounted for 80% of all individual
birds detected; species with the highest number of birds
detected were the rock shag (Phalacrocorax magellanicus)
and Falkland steamer duck (Tachyeres brachypterus),
the latter being present on all islands (Table 1). Coastal
waterbird abundance varied considerably between
ecotypes. It was highest in predator-free tussac habitats,
followed closely by rat-eradicated tussac habitat and lowest
on the grazed (non-tussac) mainland coastlines where rats,
mice and cats were present. It was twice as high in ratfree tussac habitat than in rat-infested tussac. A similar
pattern emerged for non-tussac (i.e. grazed) islands where
abundance in rat-free habitat was also nearly twice as high
as that of rat-infested habitat. Overall, bird abundance in
rat-free habitats, regardless of their tussac and grazing
status, was higher than that in all rat-infested habitats.
However, grazing also appeared to exert some effect, in
that abundance was systematically lower in non-tussac
habitat compared to tussac (Table 2, Fig. 2).
Effect of environmental variables on bird abundance
and species richness
GLM analyses of the coastal waterbird relative
abundance data from the subset of 139 islands (214.15 km
of coastline) indicated that the model best supported by the
data included the presence of rats (Table 4). Heavy grazing
appeared to have a slightly negative effect on relative
abundance; coastline perimeter of an island and the percent
of an island covered in tussac had a negligible effect. Rats,
heavy grazing and percent tussac cover had no effect on
species richness. Coastline perimeter, however, may affect
species richness, with more species being present on larger
islands. The covariates depicting weather at the time of
Table 5 Results of Welch’s two sample t-tests and ANOVA
on the effect of rat presence, heavy grazing, percent of an
island covered in tussac and coastline perimeter on the
relative abundance of the coastal waterbird assemblage
on 139 islands for 214.15 km of coastline surveyed.
Welch’s t-test
ANOVA
t-value p-value F-value
Fig. 2 The relative abundance (birds per kilometre of
coastline surveyed) of the Falkland Islands coastal
waterbird assemblage in six different ecotypes.
Rats (present)
Heavy grazing (present)
Percent tussac cover
Coastline perimeter
-4.63
-2.85
0.0000
0.0069
13.37
6
151
Island invasives: scaling up to meet the challenge. Ch 1B Rodents: Review
Table 6 Results of Welch’s two sample t-test and ANOVA
on the effect of rats, heavy grazing, percent of an island
covered in tussac and coastline perimeter on the species
richness of the coastal waterbird assemblage on 139
islands for 214.15 km of coastline surveyed.
Welch’s t-test
ANOVA
t-value p-value F-value
Rats (present)
Heavy grazing (present)
Percent tussac cover
Coastline perimeter
0.47
2.05
0.6405
0.0513
0.77
0.30
survey (i.e., wind speed, precipitation, cloud cover, and
tide) did not appear in any of the best models (those with
DAICC < 7), so we concluded that they did not affect bird
abundance or species richness. The effects of individual
covariates analysed using Welch’s two sample t-test results
indicate a highly significant negative effect of both rats
and heavy grazing on the relative abundance of coastal
waterbirds. ANOVA results show that relative abundance
increased significantly with percent tussac grass cover on
an island and decreased with coastline perimeter (Table
5). Welch’s two sample t-test and ANOVA results indicate
that species richness is not affected by rats, heavy grazing,
percent tussac cover on an island or island coastline
perimeter (Table 6).
Effect of Norway rat status on bird abundance and
species richness
Kruskal-Wallis tests with Bonferroni post-hoc tests
showed that there was a significant effect of rat status on
coastal waterbird abundance (Table 3). Bird abundance
differed significantly between rat-infested and historically
rat-free islands and between rat-infested and rat-eradicated
islands, indicating that rat eradication resulted in an
increase in coastal waterbird abundance. There was no
significant difference in abundance between historically
rat-free and rat-eradicated islands, which may possibly
indicate that bird populations had nearly fully recovered
following eradications. Results from the one-way ANOVA
showed no effect of rat status on species richness.
DISCUSSION
The presence of Norway rats was the most important
factor in predicting the relative abundance of the coastal
waterbird assemblage (Table 4). Rat presence had a
strong and significant negative effect on bird abundance
(Tables 3, 4 and 5). This negative effect and the significant
recovery benefits of rat eradication are like those observed
for passerines (Hall, et al., 2002; Tabak, et al., 2015b). In
contrast, rats did not affect species richness of the coastal
waterbird assemblage.
Heavy grazing and the percent of tussac cover on
an island also had significant negative effects on coastal
waterbird abundance (Tables 2 and 3). Previous work has
shown that grazing has a negative effect on bird abundance
(Batáry, et al., 2007). In the Falklands, grazing led to the
disappearance of the majority of the coastline’s original
vegetation of tall native grasses and shrubs which, in a
landscape devoid of trees, provided optimal breeding
habitat for the majority of the islands’ bird populations. The
impact of grazing is reflected in the higher bird abundance
of rat-free tussac coastlines (ecotype III) compared to ratfree non-tussac (grazed) coastlines (ecotype V; Table 2).
152
An indication of the effectiveness of rat eradications
in restoring coastal waterbird populations is shown by the
large difference in bird relative abundance between rateradicated and rat-infested islands (Table 2). Abundance
levels on the former are nearly twice as high as on the
latter and approximate those of historically rat-free islands
(Table 3) indicating that significant increases in coastal
waterbird populations are likely when rats are eradicated.
The importance of rat-free tussac islands (ecotype II) for
Falkland bird populations is clearly demonstrated by the
statistically significant difference in relative abundance
of coastal waterbirds on these islands compared with
rat-infested tussac islands (ecotype III) where relative
abundance is 50% less. However, it is the coastlines of
East Falkland and West Falkland (ecotype I where Norway
rats, cats and mice are present and the impacts of grazing
and destruction of native habitats are widespread), that
show the largest response with a five-fold reduction in bird
abundance. The impact that mice and cats exert on coastal
waterbird species of the Falklands is largely unknown
(Matias & Catry, 2008; Rexer-Huber, et al., 2013),
although negative impacts have been assumed (Johnson
& Stattersfield, 1990). Suggestions that the likely impact
of the endemic warrah (Dusicyon antarcticus) on bird
abundance was continued by later anthropogenic mammal
introductions (notably cats) (Hall, et al., 2002) is an
important consideration in any assessment of the potential
of East Falkland and West Falkland for future vertebrate
pest eradications.
CONCLUSION
Our study of the effect of environmental variables, and
notably predators and over-grazing of native vegetation,
on coastal waterbirds has identified the potential for
differences in relative abundance of these species to
serve as indicators of ecosystem recovery following rat
eradications and habitat restoration activities. It is a first
step in understanding the range of environmental factors
that influence the distribution and abundance of Falkland
coastal waterbirds. However, caution is required when
interpreting differences as they may be caused not only
by the balance of predation effects of Norway rats upon
birds or by grazing impacts and habitat alteration but also
by other indirect effects such as annual oceanographic and
climate variations, biogeographical factors, island size and
mesopredator release of birds (Watari, et al., 2011).
Future work will aim to determine the impact of these
factors on individual species. Additionally, conducting
repeated visits on islands will allow for future models
to incorporate estimates of each species’ detection
probability. An improved understanding of how coastal
bird distribution and abundance is affected by ecosystem
processes is essential for informing future eradications
in the Falkland Islands and for monitoring other largescale landscape-level ecological changes to the Falklands
coastline and its inshore marine environment.
ACKNOWLEDGEMENTS
Fieldwork for this research has been supported by
funding from a number of organisations since 2008;
they include the Antarctic Research Trust, Shackleton
Scholarship Fund, Falkland Islands Government, Royal
Society for the Protection of Birds, United Kingdom
Overseas Territories Environment Programme, United
Kingdom Joint Nature Conservation Conservancy,
Falklands Conservation, Bleaker Island Farm and Falkland
Islands Company. We are grateful for permission from
landowners in the Falklands to access privately owned
islands and we thank the following individuals for their
Poncet, et al.: Rats and waterbirds, Falkand Islands
contributions to the development of this research: Ian
Fisher, Carlos Martinez del Rio, Nick Rendell, the late Ian
Strange, Clare Stringer, Eric Woehler and Robin Woods.
REFERENCES
Batáry, P., Báldi, A. and Erdős, S. (2007). ‘Grassland versus non-grassland
bird abundance and diversity in managed grasslands: Local, landscape
and regional scale effects’. Biodiversity and Conservation 16: 871–881.
Burnham, K.P. and Anderson, D.R. (2002). Model Selection and
Multimodel Inference: A Practical Information-theoretic Approach.
New York, NY: Springer.
Strange, I.J., Parry, C.J., Parry, M.C. and Woods, R.W. (1988). Tussac
Grass in the Falklands. Port Stanley: Falkland Island Foundation
Project Report.
Strange, I. J. (1992). Wildlife of the Falkland Islands and South Georgia.
London: HarperCollins.
Tabak, M.A., Poncet S., Passfield K. and Martinez del Rio, C. (2014).
‘Invasive species and land bird diversity on remote South Atlantic
islands’. Biological Invasions 16: 341–352.
Tabak, M.A., Poncet S., Passfield K., and Martinez del Rio, C. (2015a).
‘Modeling the distribution of Norway rats (Rattus norvegicus) on
offshore islands in the Falkland Islands’. NeoBiota 24: 33–48.
Courchamp, F., Chapuis, J.L. and Pascal, M. (2003). ‘Mammal invaders
on islands: Impact, control and control impact’. Biological Reviews 78:
347–383.
Tabak, M.A., Poncet S., Passfield K., Goheen, J.R. and Martinez del
Rio, C. (2015b). ‘Rat eradication and the resistance and resilience of
passerine bird assemblages in the Falkland Islands’. Journal of Animal
Ecology 84: 755–764.
Courchamp, F., Caut, S., Bonnaud, E., Bourgeois, K, Angulo, E. and
Watari, Y. (2011). ‘Eradication of alien invasive species: surprise
effects and conservation successes’. In: C.R. Veitch, M.N. Clout and
D.R. Towns (eds.) Island invasives: eradication and management, pp.
285–289. Occassional Paper SSC no. 42. Gland, Switzerland: IUCN
and Auckland, New Zealand: CBB.
Towns, D.R. (2011). ‘Eradications of vertebrate pests from islands around
New Zealand: What Have we Delivered and What Have we Learned?’.
In: C.R. Veitch, M.N. Clout and D.R. Towns (eds.) Island invasives:
eradication and management, pp. 364–371. Occassional Paper SSC no.
42. Gland, Switzerland: IUCN and Auckland, New Zealand: CBB.
Directorate of Overseas Surveys. (1962). The Directorate of Overseas
Surveys 453 (Series H791). UK: Directorate of Overseas Surveys.
Ebbert, S.E. and Byrd, G.V. (2002). ‘Eradications of invasive species
to restore natural biodiversity on Alaska Maritime National Wildlife
Refuge’. In: C.R. Veitch and M.N. Clout (eds.) Turning the tide: the
eradication of invasive species, pp. 102–109. Occasional Paper SSC
no. 28. IUCN SSC Invasive Species Specialist Group. IUCN, Gland,
Switzerland and Cambridge, UK.
Veitch, C.R., Gaskin, C., Baird, K. and Ismar S.M.H. (2011). ‘Changes
in bird numbers on Raoul Island, Kermadec Islands, New Zealand,
following the eradication of goats, rats and cats’. In: C.R. Veitch,
M.N. Clout and D.R. Towns (eds.) Island invasives: eradication and
management, pp. 373–376. Occassional Paper SSC no. 42. Gland,
Switzerland: IUCN and Auckland, New Zealand: CBB.
Falklands Conservation. (2006). Important Bird Areas of the Falklands.
UK: Falklands Conservation.
Watari, Y., Caut, S., Bonnaud, E., Bourgeois, K. and Courchamp, F.
(2011). ‘Recovery of both a mesopredator and prey in an insular
ecosystem after the eradication of rodents: a preliminary study’. In: C.R.
Veitch, M.N. Clout and D.R. Towns (eds.) Island invasives: eradication
and management, pp. 377–383. Occassional Paper SSC no. 42. Gland,
Switzerland: IUCN and Auckland, New Zealand: CBB.
Falkland Islands Government. (2014). The Falkland Islands Biodiversity
Database. Stanley.
Woods, R. W. 1970. The avian ecology of a tussac island in the Falklands
Islands. Ibis 112: 15–24.
Hall, J.R., Woods, R.W., Brooke, M.de L. and Hilton, G.M. (2002).
‘Factors affecting the distribution of landbirds on the Falkland Islands’.
Bird Conservation International 12: 151–167.
Woods, R. W. 1984. A census of breeding Falkland Islands passerine
birds in tussac grass. UK: Unpublished report to the Falkland Islands
Foundation.
Hilton, G.M. and Cuthbert, R.J. (2010). ‘The catastrophic impact of
invasive mammalian predators on birds of the UK Overseas Territories:
A review and synthesis’. Ibis 152: 443–458.
Woods, R.W. and Woods, A. 1997. Atlas of Breeding Birds of the Falkland
Islands. Oswestry, Shropshire, U.K.: Anthony Nelson.
Johnson, T.H. and Stattersfield, A.J. (1990). A’ global review of island
endemic birds’. Ibis 132: 167–180.
Woods, R.W., and Woods, A. 2006. Birds and Mammals of the Falkland
Islands. Hampshire: WILDGuides.
Kurle, C.M., Croll, D.A. and Tershy, B.R. (2008). ‘Introduced rats
indirectly change marine rocky intertidal communities from algae- to
invertebrate-dominated’. Proceedings of the National Academy of
Sciences 105(10): 3801–3804.
Matias, R. and Catry, P. (2008). The diet of feral cats at New Island,
Falkland Islands, and impact on breeding seabirds. Polar Biology 31:
609–616.
Poncet, S. (1998). ‘Eradicating Foxes from Beaver Island’. In: J. FisherSmith, & L. Wallace (eds). The Wool Press 100, pp. 15–16. Stanley:
Department of Agriculture.
Poncet, S., Poncet, L., Poncet, D., Christie, D., Dockrill, C. and Brown,
D. (2011). ‘Introduced mammal eradications in the Falkland Islands
and South Georgia’. In: C.R. Veitch, M.N. Clout and D.R. Towns (eds.)
Island invasives: eradication and management. Proceedings of the
International Conference on Island Invasives, pp. 332–336. Occassional
Paper SSC no. 42. Gland, Switzerland: IUCN and Auckland, New
Zealand: CBB.
R Core Team. (2015). R: A Language and Environment for Statistical
Computing. Vienna, Austria: R Foundation for Statistical Computing.
Rauzon, M.J. (2007). ‘Island restoration: Exploring the past, anticipating
the future’. Marine Ornithology 35: 97–107.
Rexer-Huber, K., Parker, G.C., Reeves, M., Stanworth, A.J. and Cuthbert,
R.J. (2013). ‘Winter ecology of house mice and the prospects for their
eradication from Steeple Jason (Falkland Islands)’. Polar Biology
36(12): 1791–1797.
St Clair, J.J.H, Poncet, S., Sheehan, D.K., Sze´kely, T. and Hilton, G.M.
(2011). ‘Responses of an island endemic invertebrate to rodent invasion
and eradication’. Animal Conservation 14: 66–73.
Slater, G.J., Thalmann, O., Leonard, J.A., Schweizer, R.M., Koepfil, K-P.,
Pollinger, J.P., Rawlence, N.J., Austin, J.J, Cooper, A. and Wayne, R.K.
(2009). ‘Evolutionary history of the Falklands wolf’. Current Biology
19(20): R937–R938.
153
D.J. Will, K. Swinnerton, S. Silander, B. Keitt, R. Griffiths, G.R. Howald, C.E. Figuerola-Hernandez and J.L. Herrera-Giraldo
Will, D.J.; K. Swinnerton, S. Silander, B. Keitt, R. Griffiths, G.R. Howald, C.E. Figuerola-Hernandez and J.L. Herrera-Giraldo. Applying lessons learnt
from tropical rodent eradications: a second attempt to remove invasive rats from Desecheo National Wildlife Refuge, Puerto Rico
Applying lessons learnt from tropical rodent eradications: a second
attempt to remove invasive rats from Desecheo National Wildlife Refuge,
Puerto Rico
D.J. Will1, K. Swinnerton1, 2, S. Silander3, B. Keitt1,4, R. Griffiths1, G.R. Howald1, C.E. Figuerola-Hernandez1
and J.L. Herrera-Giraldo1
Island Conservation, 2100 Delaware Ave Suite 1 Santa Cruz, CA 95060. <david.will@islandconservation.org>.
Current affiliation: The Island Endemics Foundation, P.O. Box 1908, Boquerón, PR 00622. 3U.S. Fish & Wildlife
Service, Caribbean Islands NWR, P.O. Box 510 Boquerón, PR 00622. 4Current affiliation: American Bird Conservancy,
4249 Loudoun Ave, The Plains VA 20198.
1
2
Abstract The introduction of invasive rats, goats, and rhesus macaques to Desecheo National Wildlife Refuge, Puerto
Rico led to the extirpation of regionally significant seabird colonies and negatively impacted plant and endemic reptile
species. In 2012, following the successful removal of goats and macaques from Desecheo, an attempt to remove black
rats using aerially broadcast rodenticide and bait stations was unsuccessful. A review of the operation suggested that the
most likely contributors to the failure were: unusually high availability of alternative foods resulting from higher than
average rainfall, and insufficient bait availability. In 2016, a second, successful attempt to remove rats was conducted
that incorporated best practice guidelines developed during a workshop that focused on addressing the higher failure rate
observed when removing rats from tropical islands. Project partners developed a decision-making process to assess the
risks to success posed by environmental conditions and established go/no-go decision points leading up to implementation.
Observed environmental conditions appeared suitable, and the operation was completed using aerial broadcast of bait in
two applications with a target sowing rate of 34 kg/ha separated by 22 days. Application rates achieved on the ground
were stratified such that anticipated high risk areas in the cliffs and valleys received additional bait. We consider the
following to be key to the success of the second attempt: 1) monitoring environmental conditions prior to the operation,
and proceeding only if conditions were conducive to success, 2) reinterpretation of bait availability data using the lower
99% confidence interval to inform application rates and ensure sufficient coverage across the entire island, 3) treating the
two applications as independent, 4) increasing the interval between applications, 5) seeking regulatory approval to give
the operational team sufficient flexibility to ensure a minimum application rate at every point on the island, and 6) being
responsive to operational monitoring and making any necessary adjustments.
Keywords: bait availability, environmental conditions, operational monitoring, regulatory approval
INTRODUCTION
Tropical islands are rich in biodiversity but are
susceptible to invasive species. Invasive species are the
leading threat to island biodiversity (Tershy, et al., 2015),
with invasive rodents known to be particularly harmful
(Towns, et al., 2006). Eradications have been successful in
removing invasive species (Veitch, et al., 2011), allowing
island species to recover (Jones, et al., 2016), however
there has been a greater record of success on temperate
islands than on tropical islands (Russell & Holmes, 2015).
The two rodent eradication attempts (failed, then
subsequently successful) on Desecheo Island, Puerto Rico
offer an opportunity to explore the challenges of tropical
rodent eradications. Here, we highlight the key changes that
were made to the operational strategy during the second
attempt, the role of the recently developed recommended
best practices for tropical rodent eradications from Keitt, et
al. (2015), and chronicle the recently confirmed successful
project.
home to three single-island endemic and two native reptile
species (Evans, et al., 1991) and a US Endangered Species
Act listed threatened cactus, higo chumbo (Harrisia
portoricensis). Desecheo was originally set aside as a
wildlife preserve in 1912, but the introduction of invasive
goats (Capra hircus), rhesus macaques (Macaca mulatta),
feral cats (Felis catus) and black rats (Rattus rattus), and
human uses of the island, had a substantial impact on the
island’s habitat, contributing to the collapse of the large
seabird populations (Evans, 1989; Meier, et al., 1989).
In 1976, the island was transferred to the US Fish and
Wildlife Service (USFWS) who currently manage it as
Study area
Desecheo is a small (117.1 ha) hilly island (18° 23’
N, 67° 29’ W) situated in the Mona Passage about 17 km
offshore of the west coast of Puerto Rico (Fig. 1). Desecheo
is composed of a peak of volcanic calcareous rock with
a mosaic of grassy patches, shrublands, woodlands with
candelabra cacti, and semideciduous forests dominated by
Bursera simaruba (Woodbury, et al., 1971). The highest
point is nearly 200 m with steep slopes ranging from 20 to
35 degrees
Historically, Desecheo was a major seabird rookery and
in the early 1900s tens of thousands of seabirds nested on
the island (Wetmore, 1918; Meier, et al., 1989) and it is
Fig. 1 Desecheo National Wildlife Refuge, located 17 km
west of Puerto Rico in the Mona Passage.
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
154
up to meet the challenge, pp. 154–161. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Will, et al.: Second rat removal Desecheo NWR
Desecheo National Wildlife Refuge. To restore the island,
the USFWS and collaborators removed feral cats in 1987
(Evans, 1989), feral goats in 2003 and the rhesus macaque
population was reduced to being functionally extinct (i.e.
reproduction ceased with only one wild macaque known
to remain on the island) between 2009 and 2015 (Hanson,
et al., 2019). In the absence of herbivory from goats and
macaques, island species showed evidence of recovery,
the higo chumbo resurged from the suppression caused
by herbivory (Figuerola-Hernández, et al., 2017) and
researchers detected seabirds prospecting for suitable
habitat and attempting to nest on the main island in small
numbers. However, recovery of the island ecosystem
would not be possible until rats were removed.
MATERIALS AND METHODS
Project planning
Planning for the removal of black rats on Desecheo
began in 2007 through the National Environmental Policy
Act (NEPA) review process. The Finding of No Significant
Impact (FONSI) identified aerial application of cereal bait
pellets containing brodifacoum as the preferred alternative.
The ‘dry’ season from January to April was considered
the ideal period for baiting because food for rats would be
limited due to the dry environmental conditions and there
was a higher likelihood of suitable weather for conducting
an aerial application. Field trials were conducted in
February and March 2009 and 2010 to evaluate rodent
breeding status, presence of naturally occurring foods,
abundance of non-target bait competitors, bait application
rates, and detection capabilities of rodent surveillance
devices.
Rats trapped during the 2009 (n = 33) and 2010 (n = 70)
field trials indicated that rat reproduction appeared to be
low during the dry season with no juvenile rats caught and
no captured females showing signs of lactation or foetal
development. The mean hermit crab density surveyed
in 2010 was 696 crabs/ha but densities were higher in
woodland sites (833 crabs/ha) than in shrubland sites (61
crabs/ha). Tomahawk live traps proved to be an effective
surveillance device for rats with a 25% capture rate in 2009
and 55% in 2010.
2012 eradication attempt
Brodifacoum Conservation 25-D (Bell Labs, Madison
WI) 2 g pellets were applied aerially by helicopter in March
2012 using a spreader bucket slung below the helicopter. To
minimise the risk of bait entering the marine environment,
bait was applied along the coastal zone with a directional
half swath bucket (deflector) and in the interior with a full
swath starting and stopping inside of the coast. A full coastal
swath was flown inland of the coast at the interface of the
coastal and interior zones to provide sufficient overlap or
‘safety buffering’ and reduce the risk of bait gaps and areas
of lower than target bait density (Fig. 2).
Additionally, to offset suspected ant consumption
and supplement aerial broadcast in high risk areas bait
stations were established at an interval of 25 m along two
parallel transects on the ridgelines. Ant stations armed with
Amdro®Pro fire ant bait (0.73% hydramethylnon) were
placed within 1.5 m of each bait station. Stations were
checked at least weekly and bait was replaced as needed
for six weeks.
Bait availability transects were established across two
of the same habitats as the trials (woodland and shrubland)
measuring 1 × 25 m. The number of pellets in each transect
was standardised and plots were sampled for seven
consecutive days after each aerial broadcast or until all
pellets had disappeared. At each visit, the number of pellets
remaining was counted.
A captive programme was undertaken to hold
representative samples of two endemic reptiles as a
preventive action to reduce the risk of population-level
impact from the application of rodenticide. A reptile markrecapture monitoring study was done between February
and April 2012 to confirm that the use of brodifacoum did
not cause any observed population-level impacts in wild
reptile populations on Desecheo (Herrera Giraldo, et al.,
2019).
A live rat was found and captured 12 days after the 2nd
application at the field camp and a buffer of bait stations
was deployed in trees surrounding the field camp. No bait
take was observed, and no additional rats were seen during
the next week staff were on island.
In 2009, bait availability trials using a placebo
biomarker in woodland habitat showed that bait applied
at 18 kg/ha remained available in most plots for at least
three days. The second trial in 2010 showed similar results
for the same habitat. However, plots located on ridges
in shrubland habitat exhibited a much faster rate of bait
disappearance. Bait consumption by ants, considered to be
in higher numbers on the island’s exposed ridgelines, was
suspected to be one of the key factors driving this result.
Monitoring during the trials also demonstrated, through
non-toxic biomarker bait, that native and endemic reptiles
could be exposed. Additionally, all surveyed hermit
crabs in woodland sites tested positive for the presence
of biomarker; this, together with high densities of crabs,
indicated that hermit crabs would be a significant consumer
of rodent bait.
Based on trial data, a bait application strategy was
designed to achieve a bait density on the ground of 18
kg/ha during the first application followed approximately
10 days later by a second application targeting 9 kg/ha.
Desecheo has a planar area (2-dimensional) of 117.1 ha
including the offshore islets, and a topographical surface
area (3-dimensional) of 134 ha; the surface area is 13%
higher than the planar area. To account for the island’s steep
topography, bait was sown at a rate of 20 kg/ha followed by
10 kg/ha to achieve the bait density required on the ground.
Fig. 2 Bait application strategy showing flight plan used to
minimise bait into the marine environment with interior
flight lines starting and stopping inside the coastal edge,
a coastal half swath (deflector) along the coastal edge,
and a full swath coastal overlap at the interface of the
interior and coastal zones.
155
Island invasives: scaling up to meet the challenge. Ch 1B Rodents: Review
Eradication failure
Rats were not detected during fieldwork in October
2012 (six months post-operation), but in March 2013 (oneyear post-operation) rats were observed and captured.
Subsequent analysis of remote cameras deployed in
2012 showed the first rat detection in November 2012.
Genetic testing indicated the eradication operation was not
successful and the presence of rats was not the result of a
reintroduction (i.e. operational failure).
To determine reasons why the operation may have
failed, a review of the project investigated if rats could
not eat a lethal dose because of gaps in bait coverage,
insufficient bait availability, resistance to the toxin in the
bait, or that the bait was not toxic enough; or if they would
not eat a lethal dose of bait because of the palatability of
the bait, availability of natural food resources, or breeding
behavioural changes (i.e. pregnant females or emerging
pups). Resistance to the toxin in the bait, bait toxicity,
and bait palatability were not considered likely because
the bait product had a proven record of success and rats
captured during the biomarker trials showed a high level
of acceptance. Despite implementing in the ‘dry season’,
rainfall leading up to the operation was significantly higher
than it was prior to either of the placebo bait trials, which
may have resulted in a subsequent increase in the availability
of natural food resources for rats and probable rat breeding.
Bait disappeared quickly in several of the woodland plots
with all bait disappearing within two to three nights of
each application, likely the result of the significant crab
densities in the woodland habitat. Finally, while there
were few true gaps in bait coverage some areas during
the first bait application received bait at less than half the
prescribed rate. Thus, insufficient bait availability due to
localised low bait densities during the first bait application
and invertebrate bait competition, and an increase in the
availability of natural food resources and rodent breeding
due to above average rainfall, were identified as factors
that could have individually or collectively contributed to
the failure.
Tropical rodent eradication failures
About the same time that the 2012 attempt failed
there were several other high-profile rat eradication
failures on tropical islands, including Wake Atoll, western
tropical pacific; Enderbury, Phoenix archipelago; and
Henderson Island, Pitcairn group (Keitt, et al., 2015).
A subsequent analysis of historical data showed that
tropical rat eradications fail more than twice as often
as temperate eradications (Russell & Holmes, 2015),
resulting in a workshop attended by global experts to
evaluate the possible reasons for this higher risk of failure
and recommend solutions. The result of this workshop
was a paper that provided recommended guidelines for rat
eradications on tropical islands using aerial broadcast of
brodifacoum (Keitt, et al., 2015).
Revised project approach
Starting in 2014, a steering committee of project
partners (USFWS, USDA, and Island Conservation) was
established to evaluate how to conduct a second attempt,
the available strategy options, and how to manage ongoing
project risk. A revised operational strategy was developed
based on information from the review of the 2012 attempt
and the recommended guidelines produced during the
workshop on tropical rodent eradications (Keitt, et al.,
2015). The following highlights the key changes:
156
1) Monitoring environmental conditions prior to the
operation and proceeding only if conditions were
conducive to success
A comprehensive review of factors influencing
environmental conditions on Desecheo was conducted
showing that rainfall and soil moisture content were key
drivers of resource availability, typical of Puerto Rican
subtropical dry forests. Inter-annual variability was
evaluated using monthly rainfall totals and vegetation
greenness, as a proxy for resource availability, between
2000 and 2013. Vegetation greenness was derived from
remote sensing analyses using 30 m resolution 16-day
MODIS Enhanced Vegetation Index (EVI) data. EVI data
were smoothed using the HANTS algorithm (Roerink, et
al., 2000) and mean monthly EVI were extracted from
pixels that intersected the island using R (R Core Team,
2016).
Four assessments were conducted between three
months and one week prior to implementation to evaluate
the risk that short-term climatic changes could trigger
higher biological productivity on the island prior to
an irretrievable commitment of resources. Increased
greenness represented more food availability via plants and
invertebrates, and thus, increased opportunities for rodent
breeding and increased bait competition due to invertebrate
abundance. Each assessment included a review of regional
climatic summaries, regional forecast products, and local
weather conditions. Additionally, four island site visits
were conducted to measure local rainfall, plant fruiting and
flowering productivity, canopy cover and rodent breeding.
To assist in data collection an automated logging rain gauge
(WeatherShop, California, U.S.), three time-lapse cameras
taking two photos per day (Day6Outdoors, Georgia,
U.S.), and eight standardised photo point locations, were
established on island.
A summary of conditions following each assessment
was provided to the project steering committee for
review. These summaries provided a subjective evaluation
based on the team’s knowledge of the island and the
recommendations were used as part of a holistic evaluation
of risk factors facing project implementation to make an
operational go/no-go decision.
2) Reinterpretation of bait availability data
Using recent guidelines from Pott, et al. (2015) and
data from the 2012 eradication attempt, bait availability
was recalculated based on the lower-limit for a 99%
t-based confidence interval. The linear rate at which
bait disappeared was estimated by calculating the slope
from four days of bait availability: 5.97 kg/ha per day in
the woodland plots during the 2012 eradication attempt.
This daily disappearance rate was used to calculate a
conservative target bait density on the ground of 30 kg/ha
to ensure that bait was available to rats for approximately
five consecutive days after each application.
3) Treat the two applications as independent events
Following the guidelines outlined in Keitt, et al.
(2015), the second attempt targeted the same application
rate for each application and the target interval between
application was increased to approximately 24 days. Two
critical habitats, the valleys and steep cliffs identified
in the review as areas of concern, were earmarked for
additional supplemental bait application. On Desecheo,
the predominant valleys and cliffs run perpendicular to
one another, such that flights that are parallel to one are
perpendicular to the other. To mitigate concerns about
Will, et al.: Second rat removal Desecheo NWR
the impact on bait density caused by bait shadows on
steep terrain (i.e. more bait downslope than upslope) and
the higher rat and non-target bait consumer densities in
the valleys observed during the 2012 attempt, additional
flights were flown parallel to the valleys and cliff features
to achieve higher application rates in these areas.
4) Seeking regulatory approval to give the operational
team sufficient flexibility to ensure a minimum application
rate at every point on the island
Some areas of Desecheo received bait below the desired
bait density during the 2012 attempt. For the second
attempt, the operational team sought regulatory approval
to achieve a minimum bait density across every point on
the island. This allowed for the retreatment of any areas
that were estimated to be below the desired minimum bait
density and limited the total amount of bait that could be
applied per application rather than the application rate.
The project review also noted that the bait application
strategy used to minimise bait in the marine environment
created a risk of bait gaps and/or lower than planned baiting
rates between coastal and interior zones. Regulatory
approval was sought to ensure sufficient bait was available
to treat the interface between the interior and coastal
zones. This provided the flexibility to achieve the desired
minimum bait density while also minimising bait entering
into the marine environment.
5) Responding to operational monitoring in real time
To ensure quality coverage, bait sowing rates were
carefully monitored and the helicopter shut down every
five loads to download GIS files and review progress.
During each application, a GIS specialist produced bait
application maps estimating bait densities achieved on
the ground. These data were used to identify any possible
errors in flight lines, GPS logging, or bait application
rates. Any gaps, identified as areas larger than 20 × 20 m
receiving less than 15% (5 kg/ha) of the target bait density,
were re-treated.
Greater emphasis was placed on operational monitoring
than in 2012, including the deployment of additional bait
availability monitoring transects and ground-truthing
of bait application rates across the island. Additionally,
communications between the environmental monitoring
and bait application teams were improved by conducting
the bait loading on island so that key project personnel were
in the same place. In 2012 bait loading was done in Rincón
approximately 17 km away on the main island of Puerto
Rico. Following the first bait application, and prior to the
second, a review of all operational data was conducted to
allow for adjustments to the operational strategy.
2016 eradication attempt
The second eradication attempt was conducted in March
and April 2016. The baiting strategy used was similar to
the 2012 attempt albeit with an increased application rate
and additional supplemental treatments along the cliffs and
valleys. As in 2012, the sowing rate during the 2016 attempt
was increased from 30 kg/ha to 34 kg/ha to accommodate
the 3-dimensional surface area to ensure the desired bait
density on the ground.
To allow comparisons with bait availability data
collected during the previous field trials and the 2012
attempt, 25 m2 sample transects were monitored in the
woodland and shrubland habitats using the previous
protocols. Additionally, a circular hoop sampling method
(1 m2) was used to estimate bait density on the ground
following each application and collect additional bait
availability across five different treatment zones.
Confirmation
In April 2017, one year after implementation, staff
returned to the island and deployed chew tags, tracking
tunnels, and live traps to confirm the absence of rats.
Additionally, images from trail cameras were collected and
analysed.
RESULTS
A summary of key differences between the two attempts
is outlined in Table 1.
Environmental conditions
On first arrival at Desecheo Island on February 19,
2012, initial impressions were that the island’s vegetation
was more lush and green than observed during the same
period in 2009 and 2010. Personnel recorded a total of 25.5
mm of precipitation on Desecheo between 10 March and 2
April 2012. Opportunistic necropsies of a small number of
rats (n = 6) found dead during the 2012 operation showed
one female rat with three embryos, and a male and the
same female showed subjectively significant abdominal
fat. However, the monitoring team did not observe any
small juvenile rats suggesting breeding was not widely
occurring for any prolonged period beforehand.
Retrospective analysis of precipitation recorded at the
Rincón, Puerto Rico station (the closest point to Desecheo)
showed that rainfall between January and March 2012
was above the annual average, and in February 2012
precipitation was 2.9 times higher than the 34-year average
and the third highest rainfall for the month of February
since 1968 (NOAA, 2015). Further, the remote sensing
Table 1 Summary of key differences between the 2012 and 2016 eradication attempts on Desecheo.
Factor
2012
2016
Month
Rainfall 6 months prior
Rainfall during
Rodent breeding
March
4603 mm
25.5 mm
One pregnant female
observed (n=6)
March/April
772 mm
35.56 mm
None observed (n=44)
Canopy Cover
Flush vegetation
Post-peak vegetation followed by unproductive
flowering after 31 mm rain event
Target bait density
18 kg/ha, 9 kg/ha
30 kg/ha, 30 kg/ha
Average application rate
Interval between applications
17.1 kg/ha, 9.1 kg/ha
9 days
40.3 kg/ha, 39.9 kg/ha
22 days
157
Island invasives: scaling up to meet the challenge. Ch 1B Rodents: Review
second application of bait, most flower and fruit production
had been abandoned. An irruption of caterpillars occurred
after the second application, consuming much of the fresh
Bursera growth (Shiels, et al., 2017).
Between 25 January and 10 April 2016, a total of 105.9
mm of precipitation was observed on Desecheo. Almost
half of this precipitation was the result of two single events.
In comparison, Rincón received a total of 239.3 mm of
precipitation in the same period.
Fig. 3 MODIS Enhanced Vegetation Index (EVI) showing
vegetation greenness vs rainfall in inches from Aguadilla,
Puerto Rico. Dashed line represents EVI where higher EVI
means greener and lower EVI means drier vegetation.
Stacked lines represent rainfall in millimetres. The solid
horizontal line represents the mean EVI for March, black
squares EVI during placebo trials and black diamond
EVI during the 2012 eradication.
analyses showed that March 2012 had significantly higher
EVI than the same period during either of the 2009 or 2010
trials (Fig. 3).
During the assessment trips leading up to the 2016
attempt, observations showed that the island, while lush
and green, was in post-peak greenness and starting to dry
out in January 2016. This corresponded with a delayed
and short wet season, likely the result of a record drought
throughout much of Puerto Rico in 2015 (NOAA, 2015).
By February there was a significant reduction in canopy
cover; however, a significant rain event (31 mm in 24 hours)
resulted in a large increase in canopy cover by March,
mostly restricted to Bursera trees. Increased flowering was
noticeable on some herbaceous shrubs and vine species;
however, the fruits produced were not considered to be an
alternative food source for rats. Increases in canopy cover
continued through March, but by April 2016, after the
A total of 44 rats was captured during the 2016
attempt, all animals captured were adult size and none of
the females showed signs of pregnancy, although some
females showed indications (fat deposits and engorged
uterine blood vessels) that breeding could have occurred
soon after.
Bait application
During the 2012 eradication attempt, 3,588 kg of bait was
applied on Desecheo as required by regulatory compliance,
which resulted in an average application rate of 17.1 kg/ha
and 9.1 kg/ha. An interval of nine days separated the first
and second bait applications. Additionally, a total of 127 kg
of bait was used in 107 bait stations placed along the ridges.
The target application rates (18 kg/ha and 9 kg/ha) were at
the upper limits allowed by regulatory requirements and the
operational team was cautious in their approach with 1,000
kg of available bait unused. While the average application
rates (total bait divided by island area) achieved were 17.1
kg/ha and 9.1 kg/ha, 76% of the island had a bait density
on the ground below the target, with 8% less than half the
target rate during the first application and 50% of the island
below the target, with 4% less than half the target, during
the second application.
In 2016, 10,650 kg of bait was applied according to
regulatory compliance, resulting in an average application
rate of 40.3 kg/ha and 39.9 kg/ha separated by 22 days.
Regulatory approval was sought to allow for the retreatment
of areas with less than the target bait density, ensuring a
minimum bait density at every point across the island.
Table 2 Bait availability results from placebo trials and both eradication attempts on Desecheo. Bait availability is
expressed as the 99% lower limit t-based confidence interval of mean bait availability to represent the “worst-case”
scenario rather than the average case.
Habitat
Year
1st Application
2009
Woodland
2010
2012
2016
2010
Shrubland
2012
2016
2nd Application
2009
Woodland
2010
2012
2016
2010
Shrubland
2012
2016
158
Plots
Target bait
Lower limit bait availability Lower limit bait availability
density (kg/ha) (kg/ha) after one day
(kg/ha) after three days
6
9
5
6
6
7
6
18
18
18
45
18
18
30
0.5
3.2
6.9
7.8
0
11.5
23.6
0
0
0
0
0
4.9
14.8
6
9
5
8
6
7
6
18
18
9
45
18
9
30
-
-
0
36.0
8.3
30.0
0
24.0
5.7
27.5
Will, et al.: Second rat removal Desecheo NWR
Using this strategy 11% of the island received less than the
target bait density and 2% received less than half the target
during the first application. During the second application
31% of the island received less than the target bait density
and1% less than half the target.
A review of operational monitoring data following
the first application during the 2016 attempt showed that
bait disappeared faster than anticipated in the woodland
valley habitat. In response, a total of 100 bait stations
were installed in the valleys spaced at least 25 m apart.
Bait stations were filled the day after aerial broadcast and
elevated in trees wherever possible to reduce bait take
by crabs. Each station was checked and replenished (if
needed) three times during a three-week period. A total of
22.25 kg of bait was used in bait stations.
There was a small increase in the number of non-target
carcasses observed between the 2012 (n = 4) and 2016
(n = 17) attempts, although a larger team was surveying
the island for a longer duration in 2016. Few non-target
species presented a high-risk exposure pathway so
significant mortality was not expected following the 2016
attempt despite the increase in total bait applied to the
island. Additionally, biological samples of rats, reptiles,
and invertebrates were collected before and after the
2016 attempt to evaluate the persistence of brodifacoum
in the environment four years after the 2012 attempt, if
still detectable, and following the 2016 attempt, results of
which will be reported elsewhere.
Bait availability
Observed bait availability was represented as the 99%
lower limit confidence interval of mean bait availability
(Table 2). The lower limit was used instead of mean
availability to represent the “worst-case scenario” of
bait availability rather than the average. During the first
applications of both the 2012 and 2016 attempts, estimated
bait availability reached zero in the woodland plots within
three 24-hour periods despite the difference in application
rates (Fig. 3).
Confirmation
A biosecurity monitoring trip was conducted seven
months after the second attempt in November 2016 during
which 10 A24 GoodNature traps, 40 bait stations, 10
Tomahawk live traps, 50 chew tags and 10 trail cameras
were placed near possible landing sites. In April 2017, a
total of 179 chew tags, 22 tomahawk live traps, 21 trail
cameras and 20 tracking tunnels were placed across the
island over a nine-day period for a total of 1,074; 124; 3,108;
and 114 detection nights, respectively. No signs of rats
were detected on any device during either monitoring trip.
Following confirmation, monthly biosecurity monitoring
trips between September 2017 and March 2018 continued
to check the surveillance devices with no detections of rats.
DISCUSSION
The failure of the rat eradication on Desecheo in 2012
provided an excellent opportunity to better understand
the reasons for failure, build upon the lessons learnt from
other failed projects, and design a second attempt that
addressed the key challenges. Keitt, et al. (2015) lays out
a suite of recommendations to increase the probability of
success for tropical rat eradications using aerial broadcast
of brodifacoum based on reviews of several failed projects
and input from a large group of experts. Desecheo was the
first of these failed projects to be implemented a second
time and enables review of the operational changes that
contributed to operational success.
Environmental conditions
On tropical islands rainfall is a key driver of primary
productivity and resulting elevated vegetation density is
associated with an increase in rodent population densities
(Harper & Bunbury, 2015). Like other dry tropical islands,
primary productivity and resulting resource availability
in the dry season on Desecheo (January–April) can be
variable and is highly dependent on the amount of soil
water recharge generated from successive rainfall events
in the previous year’s wet season (July–December) and
the timing and amount of rain during the dry season.
Environmental conditions leading up to the 2016 attempt
were drier than those in 2012, primarily as an artefact
of long term drought conditions experienced in 2015,
resulting in lower primary productivity, less resource
availability, and lower probability of rodent breeding.
We feel that these ‘favourable’ conditions contributed to
project success and had conditions leading up to the 2016
attempt been like those observed in 2012 the project would
have been postponed.
Even though environmental conditions, and their
subsequent implications for project success, are difficult to
predict, the subjective assessments conducted on Desecheo
were critical to the steering committee’s confidence in
proceeding with the bait application. They provided an
opportunity to critically evaluate project risk and, more
importantly, considered the consequences of postponement
in advance of a final go/no-go decision. Where possible,
future projects can improve stakeholder confidence by
identifying the primary environmental drivers that pose
risks to project success and developing a process that
evaluates these risks to inform a final go/no-go decision.
Projects should identify the worst-case scenario of
alternative resource availability, non-target bait competitor
abundance and rodent breeding, and plan accordingly.
Desecheo was relatively easy to access during day
trips, but the deployment of an automated rain gauge
and time-lapse cameras and use of remote sensing data
provided valuable information on climatic conditions that
could be replicated on remote islands. On islands where
variability in environmental conditions pose a risk to
operational efficacy projects should consider using these
tools and others to better evaluate these risks. At the very
least, projects can improve the collective knowledge
of the challenges facing tropical rodent eradications by
documenting and reporting observed environmental
conditions, and subsequent perceived risks, leading up to
and during implementation.
Bait availability
The review of the first attempt identified inadequate
overall or localised baiting rates as one of the more likely
causes of failure to eradicate rats. As described in Keitt, et
al. (2015) eradications should strive to make bait available
to rats for at least four consecutive 24-hour periods to
maximise the probability that all rats are exposed to a lethal
dose. The interpretation of the bait availability data for
the 2012 attempt used mean bait availability to determine
sufficient bait availability rather than the lower limit of
99% confidence intervals. Reinterpretation of the placebo
trials using the 99% lower limit confidence interval method
estimated that with a rate of 18 kg/ha the lower limit of bait
availability would reach zero within two to three days. This
was further supported by data from the 2012 attempt where
the lower limit of bait availability went to zero by the third
day after bait application (Fig. 3).
During the 2016 attempt bait availability observed in the
transect sampling (25 m2) roughly followed observations
from the 2012 attempt where bait disappeared more quickly
159
Island invasives: scaling up to meet the challenge. Ch 1B Rodents: Review
very near the maximum rate permitted by regulation and
the operational team needed to strike a balance between
achieving the desired application rate while staying within
permitted limits. This was not an issue unique to Desecheo
as, in general, aerial eradications conducted in the Unites
States tend to use less bait than planned, compared to
projects conducted elsewhere that use more than planned
(Will, et al., 2019).
Leading up to the 2016 attempt the operational team
aimed to address this challenge by engaging regulatory
partners early in the project process as part of the project
steering committee. The operational team justified, and
sought approval for, a strategy that focused on achieving
a site-specific minimum application rate based on the
best available science. The justified strategy estimated
the amount of bait needed to achieve a minimum rate at
every point across the island, the amount of bait needed
for overlapping flights necessary to minimise bait spread
into the marine environment while minimising the chance
of gaps along the coastal edge, and an additional amount
of bait to fill unanticipated gaps and undertreated areas.
Particularly in complex regulatory environments, future
projects should consider seeking site-specific regulatory
approval based on a justified strategy that maximises
project success and bait quantities derived from a predicted
flight plan.
Operational strategy
Fig. 4 The mean bait availability from the 25 m² bait
availability plots set in the woodland habitats in 2012
and 2016. Day 0 represents the broadcast date and Day
1 represents the first 24-hour period during which bait
is available to bait consumers (i.e. day 1 ends 24 hours
after the end of bait application). The error bars represent
the lower limit for a 99% t-based confidence interval. The
trend line represents the bait disappearance rate based
on the lower limit.
in the woodland than in the shrubland habitat. Although the
revised strategy intended for bait to be available for at least
four consecutive 24 hour periods after bait application, the
lower limit of bait availability in the woodland plots was
similar to the 2012 attempt and reached zero by the third day
after bait application despite the nearly two-fold increase
in application rate. This highlights some of the challenges
tropical rodent eradications face in the presence of nontarget bait competitors, supports the methods proposed in
Pott et al. (2015) for evaluating bait availability data, and
suggests that the higher application rate used in the 2016
attempt may have been necessary in the hermit crab-dense
woodland habitat to ensure sufficient bait availability for
all rats.
Regulatory approval
One of the criticisms of the 2012 attempt was that some
areas received lower than the prescribed rates during the
first application, particularly inside the coastal edge. This
was potentially a consequence of the complex regulatory
environment in the United States and the strategy employed
to minimise bait spread into the marine environment.
The desired target application rate on the ground was
160
The justification for increasing the interval between
applications was to reduce the risk posed by the scenario
of pups emerging three weeks after the first application.
The justification for using the same application rate in both
applications was to ensure bait availability in the presence
of non-target bait competitors. It should be noted that
several tropical island eradications elsewhere have been
successful with shorter gaps between bait applications. For
example, in Mexico seven projects were successful with
durations of 7–10 days between applications (SamaniegoHerrera, et al., 2014, 2017) even though rat breeding was
confirmed. Additionally, an interval of three weeks could
incur considerable operational costs while personnel and
equipment are on standby. Alternatively, rodent breeding
risks could be mitigated by increasing the application rate
so that bait was available for a longer period, or conduct a
third application; however, these would need to be balanced
against associated non-target risk. As Keitt, et al. (2015)
note, the recommendations should not be considered hard
and fast rules as every island is different and we still have
much to learn about tropical ecosystems.
The decision to apply additional bait in the valleys
and on cliffs was based on the perceived risks justified
from observations in 2012. These concerns appear
somewhat validated because bait disappeared quickly in
the woodland plots following the first application in 2016
despite the increased higher application rate from the 2012
attempt. It is difficult to evaluate what impact this strategy
decision had on operational success but this stresses the
importance of selecting bait application rates based on the
best available science. Additionally, future projects should
consider an additional treatment to increase confidence in
areas of concern.
Operational monitoring
Intentionally slowing down the bait application in 2016
and reviewing bait density estimate maps improved the
quality of the bait application and ensured that significantly
less of the island was below the minimum bait density than
during the 2012 attempt. Future projects should consider
this strategy particularly on small islands where a single
load treats a significant proportion of the island, and using
Will, et al.: Second rat removal Desecheo NWR
bait density estimate maps to identify gaps or low treatment
areas (Will, et al., 2019).
The second attempt put emphasis on near-real time
information sharing to inform decision-making during the
operation. While there is limited opportunity for adaptive
management during aerial eradications, where success or
failure is largely determined on the day, projects should
put processes in place to ensure that data from the field
are available to inform operational decision making
and risk assessments during project implementation.
Comprehensive operational monitoring allows managers
to implement any available response options and, more
importantly, allows stakeholders to understand project risk
as the implementation unfolds.
CONCLUSION
Although we are unlikely to determine the influence
environmental conditions, bait applications rates, or the
interval between applications have on project success
without experimentation, the variability in conditions
observed on Desecheo during the ‘dry’ season and the
consistently high rate of bait disappearance in crab-dense
areas highlight the importance of understanding an island’s
ecosystem prior to implementing tropical eradications.
The second attempt on Desecheo provided a significant
opportunity to reconsider operational strategies for
tropical eradications and marks the first of the high-profile
failures to be successfully redone following the global
review of tropical rodent eradications. The synthesis of
recommended guidelines in Keitt, et al. (2015), and the
process of reviewing project risks at pre-determined times,
were necessary for increasing stakeholder confidence to
make a second attempt. Ultimately, the rationale employed
during the successful 2016 attempt should increase global
confidence in rodent eradications on tropical islands.
ACKNOWLEDGEMENTS
The eradication of invasive rodents from Desecheo was
led and funded by USFWS (Caribbean Islands NWR) and
supported by Island Conservation and USDA – National
Wildlife Research Center. We are very grateful to the
donors who provided significant match funding including
the David and Lucile Packard Foundation and the National
Fish and Wildlife Foundation. In-kind assistance was
provided by the Commonwealth of Puerto Rico Department
of Natural and Environmental Resources, the Puerto Rico
Department of Agriculture, and the Puerto Rico Electric
Power Authority and a generous donation of bait by Bell
Laboratories Inc. We would also like to thank the many
other local companies, agencies, and individuals that
assisted during project implementation.
REFERENCES
Evans, M., Herbert, H.J. and Rohnke, K. (1991). ‘Observations on
the status of the herpetofauna of Desecheo Island National Wildlife
Refuge, Puerto Rico’. In: J.A. Moreno (ed.) Status y Distribución de los
Reptiles y Anfibios de la Región de Puerto Rico. Publicación Científica
Miscelánea No. 1, pp. 34–36. San Juan, Puerto Rico: Departamento de
Recursos Naturales.
Evans, M.A. (1989). ‘Ecology and removal of introduced rhesus monkeys:
Desecheo Island National Wildlife Refuge, Puerto Rico’. Puerto Rico
Health Sciences Journal 8(1): 139–156.
Figuerola-Hernandez, C., Swinnerton, K.J., Holmes, N., MonsegurRivera, O., Herrera Giraldo, J.L., Wold, C., Hanson, C., Silander, S. and
Croll, D. (2017). ‘Resurgence of Harrisia portoricensis (Cactaceae) on
Desecheo Island after the removal of invasive herbivores: Implications
and recommendations for the management of the species’. Endangered
Species Research 34: 339–347.
Hanson, C., Hall, T.J., DeNicola, A.J., Silander, S., Keitt, B. and
Campbell, K. (2019). ‘Rhesus macaque eradication to restore the
ecological integrity of Desecheo National Wildlife Refuge, Puerto
Rico’. In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J.
West (eds.) Island invasives: scaling up to meet the challenge, pp. 249–
255. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Harper, G.A. and Bunbury, N. (2015). ‘Invasive rats on tropical islands:
Their population biology and impacts on native species’. Global
Ecology and Conservation 3: 607–627.
Herrera-Giraldo, J.L., Figuerola-Hernández, C.E., Holmes, N.D.,
Swinnerton, K., Bermúdez-Carambot, E.N., González-Maya, J.F. and
Gómez-Hoyos, D.A. (2019) ‘Survival analysis of two endemic lizard
species before, during and after a rat eradication attempt on Desecheo
Island, Puerto Rico’. In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C.
Russell and C.J. West (eds.) Island invasives: scaling up to meet
the challenge, pp. 191–195. Occasional Paper SSC no. 62. Gland,
Switzerland: IUCN.
Jones, H.P., Holmes, N.D., Butchart, S.H., Tershy, B.R., Kappes, P.J.,
Corkery, I., Aguirre-Muñoz, A., Armstrong, D.P., Bonnaud, E. and
Burbidge, A.A. (2016). ‘Invasive mammal eradication on islands
results in substantial conservation gains’. Proceedings of the National
Academy of Sciences 113(15): 4033–4038.
Keitt, B., Griffiths, R., Boudjelas, S., Broome, K., Cranwell, S., Millett,
J., Pitt, W. and Samaniego-Herrera, A. (2015). ‘Best practice guidelines
for rat eradication on tropical islands’. Biological Conservation 185:
17–26.
Meier, A.J., Noble, R.E. and Raffaele, H.A. (1989). ‘The birds of
Desecheo Island, Puerto Rico, including a new record for Puerto Rican
Territory’. Caribbean Journal of Science 25 (1–2): 24–29.
NOAA. (2015). Climate Prediction Center: National Drought Predicition
Center. <http://www.cpc.ncep.noaa.gov/>. Accessed 11 November
2015.
Pott, M., Wegmann, A.S., Griffiths, R., Samaniego-Herrera, A., Cuthbert,
R.J., Brooke, M.d.L., Pitt, W.C., Berentsen, A.R., Holmes, N.D.,
Howald, G.R., Ramos-Rendón, K. and Russell, J.C. (2015). ‘Improving
the odds: Assessing bait availability before rodent eradications to aid in
selecting bait application rates’. Biological Conservation 185 :27–35.
R Core Team (2016). R: A Language and Environment for Statistical
Computing. Vienna, Austria: R Foundation for Statistical Computing.
Roerink, G., Menenti, M. and Verhoef, W. (2000). ‘Reconstructing
cloudfree NDVI composites using Fourier analysis of time series’.
International Journal of Remote Sensing 21(9): 1911–1917.
Russell, J.C. and Holmes, N.D. (2015). ‘Tropical island conservation: Rat
eradication for species recovery’. Biological Conservation 185: 1–7.
Samaniego-Herrera, A., Russell, J.C., Choquenot, D., Aguirre-Muñoz,
A. and Clout, M. (2014). ‘Invasive Rodents on Tropical Islands:
Eradication Recommendations from Mexico’. In: R.M. Timm and J.M.
O’Brien (eds.) 26th Vertebrate Pest Conference, pp. 43–50. Davis:
University of California.
Samaniego-Herrera, A., Aguirre-Muñoz, A., Bedolla-Guzmán, Y.,
Cárdenas-Tapia, A., Félix-Lizárraga, M., Méndez-Sánchez, F., ReinaPonce, O., Rojas-Mayoral, E. and Torres-García, F. (2017). ‘Eradicating
invasive rodents from wet and dry tropical islands in Mexico’. Oryx
10.1017/s0030605316001150: 1–12.
Shiels, A.B., Haines, W.P., Swinnerton, K.J., Silander, S., FiguerolaHernández, C., Will, D., García-Cancel, J.G. and Torres-Santana, C.W.
(2017). ‘Sudden appearance and population outbreak of Eunica monima
(Lepidoptera: Nymphalidae) on Desecheo Island, Puerto Rico’. Florida
Entomologist 100(1): 176–179.
Tershy, B.R., Shen, K.-W., Newton, K.M., Holmes, N.D. and Croll, D.A.
(2015). ‘The importance of islands for the protection of biological and
linguistic diversity’. BioScience 65(6): 592–597.
Towns, D.R., Atkinson, I.A. and Daugherty, C.H. (2006). ‘Have the
harmful effects of introduced rats on islands been exaggerated?’
Biological Invasions 8(4): 863–891.
Veitch, C.R., Clout, M.N. and Towns, D.R. (eds.) (2011). Island invasives:
eradication and management. Occassional Paper SSC no. 42. Gland,
Switzerland: IUCN and Auckland, New Zealand: CBB.
Wetmore, A. (1918). ‘The birds of Desecheo Island, Porto Rico’. Auk 35:
333–340.
Will, D., Howald, G., Holmes, N., Griffiths, R. and Gill, C. (2019).
‘Considerations and consequences when conducting aerial broadcast
applications during rodent eradications’. In: C.R. Veitch, M.N. Clout,
A.R. Martin, J.C. Russell and C.J. West (eds.) Island invasives: scaling
up to meet the challenge, pp. 71–78. Occasional Paper SSC no. 62.
Gland, Switzerland: IUCN.
Woodbury, R.C., Martorell, L.F. and Tuduri, J.G. (1971). ‘Flora of
Desecheo Island, Puerto Rico’. Journal of Agriculture of the University
of Puerto Rico 55(4): 478–505.
161
E. Bell, J. Daltry, F. Mukhida, R. Connor and K. Varnham
Bell, E.; J. Daltry, F. Mukhida, R. Connor and K. Varnham. The eradication of black rats
(Rattus rattus) from Dog Island, Anguilla, using ground-based techniques
The eradication of black rats (Rattus rattus) from Dog Island, Anguilla,
using ground-based techniques
E. Bell1, J. Daltry2, F. Mukhida3, R. Connor4 and K. Varnham5
Wildlife Management International Limited, PO Box 607, Blenheim 7240, New Zealand, <biz@wmil.co.nz>. 2Fauna
& Flora International, The David Attenborough Building, Pembroke Street, Cambridge CB2 3QZ, United Kingdom.
3
Anguilla National Trust, The Valley, Anguilla. 4 Department of Environment, Government of Anguilla, The Valley,
Anguilla. 5 RSPB, The Lodge, Sandy, Bedfordshire, United Kingdom.
1
Abstract Rat eradication techniques developed in New Zealand are a proven method for removing invasive rodents
from islands worldwide. This technology moved rapidly from ground-based bait station operations to aerial application
of rodenticides. Rat eradications on tropical islands using similar methods, have not always been as successful as those in
temperate regions. As most previous eradications in the Caribbean have been on islands smaller than 50 ha, the eradication
of black rats (Rattus rattus) from 207 ha Dog Island was a significant increase in size. Reptile and seabird populations on
Dog Island had been in decline for a number of years and black rats were identified as the most likely factor. Following
the feasibility study in 2007, the Dog Island Recovery Project was launched in 2011. This was a multiple-year project
incorporating a ground-based eradication with establishment of biosecurity procedures to prevent reinvasion, alongside
long-term monitoring of native species. Bait stations with cereal-based wax blocks containing brodifacoum at 0.005%
w/w were established on a 30–50 m grid over the island. Interference with bait stations by non-target invertebrates,
particularly crabs, was high and bait stations required moving or elevating to avoid this. However, there was no evidence
of any non-target animals being killed or injured by the bait. Eradication success was confirmed in 2014.
Keywords: biosecurity, black rat, brodifacoum, Dog Island, eradication, monitoring, Rattus rattus
INTRODUCTION
Dog Island (207 ha) is located 13 km north-west
of Anguilla (18.2783°N, 63.2533°W) in the north-east
Caribbean and consists of one main island and three
smaller offshore cays, East, Mid and West Cay (Sanders,
2006; Hodge, et al., 2008). Designated as an Important
Bird Area, the Royal Society for the Protection of Birds
(RSPB), Anguilla National Trust (ANT) and Fauna &
Flora International (FFI) have monitored the seabird
colonies on Dog Island for nearly 10 years and recorded
that seabird populations, particularly sooty terns
(Onychoprion fuscatus) and magnificent frigatebirds
(Fregata magnificens), had been declining (Campbell,
1991; Sanders, 2006; Holliday, et al., 2007; Hodge, et al.,
2008; Daltry, 2010). Dog Island also has a much reduced
endemic reptile community consisting of the Anguilla
Bank ground lizard (Pholidoscelis plei), the Anguilla Bank
tree lizard (Anolis gingivinus), two species of dwarf gecko
(Anguilla Bank dwarf gecko (Sphaerodactylus parvus)
and Leeward Island banded gecko (S. sputator)), and the
Anguilla Bank skink or slipperyback skink (Spondylurus
powelli); surveys in 2009 failed to observe any dwarf
geckos or skinks (Hodge, et al., 2003; Daltry, 2010; Hedges
& Conn, 2012). Black (ship) rats (Rattus rattus) were
identified as the most likely factor influencing this decline
through predation on eggs, and young or small individuals.
Rats are known to have devastating effects on seabird
and reptile populations, causing extinctions on numerous
islands worldwide (Moors & Atkinson, 1984; Atkinson,
1985; Towns, et al., 2006; Jones, et al., 2008, Harper &
Bunbury, 2015). Many islands have been successfully
cleared of rats, including more than 30 in the Caribbean,
with a subsequent increase in bird and reptile populations
(Day & Daltry, 1996; Daltry, 2000; Daltry, et al., 2001;
Thomas & Taylor, 2002; Towns & Broome, 2003; Jones,
et al., 2008; Varnham & Daltry, 2006; Howald, et al., 2007;
Varnham, 2010).
The Dog Island Restoration Project partnership
(consisting of Anguilla National Trust (ANT), Anguilla
Department of Environment (DOE), Fauna & Flora
International (FFI), the Royal Society for the Protection
of Birds (RSPB) and the island owner, Anguilla
Development Company) commissioned the development
of an operational plan to eradicate black rats from Dog
Island in 2011 (Bell, 2011) based on an earlier feasibility
assessment (Varnham, 2007). Wildlife Management
International Limited (WMIL) directed the eradication
with the assistance of international volunteers and ANT,
DOE, FFI and RSPB staff. The three-phase Dog Island
Recovery Project (Phase I eradication of black rats; Phase
II long-term monitoring of native species and Phase III
biosecurity) began in January 2012.
METHODS
Study area
Dog Island is a low-lying (highest point: 29 m asl),
rocky island with three small offshore islets (Mid Cay,
West Cay and East Cay). There are several long, sandy
beaches, two saline ponds and the rest of the coastline is
rocky or has low cliffs (< 8 m high). The island lies within a
Marine Protected Area, covering an area of approximately
10 km2 around the island (Hodge, et al., 2008). The island
is popular with visiting yachts and tourist vessels from
Anguilla or Saint Martin/Sint Maarten.
Dog Island was originally covered in dry forest or
woodland, with shorter vegetation in coastal areas exposed
to salt spray, but today is dominated by low, thorny
scrub (e.g. Lycium americanum and Castela erecta) and
prickly pear cacti (Opuntia spp.) due to herbivory by feral
goats (Capra hircus). Larger trees including manchineel
(Hippomane mancinella), sea grape (Coccoloba uvifera),
white cedar (Tabebuia heterophylla) and buttonwood
mangrove (Conocarpus erectus) can be found around the
coastline and occasionally inland.
The island is recognised as an Important Bird Area
because it is globally significant for a large number of
breeding seabirds, in particular sooty terns and magnificent
frigate birds, (Sanders, 2006; Hodge, et al., 2008). Other
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
162
up to meet the challenge, pp. 162–166. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Bell, et al.: Dog Island using ground-based techniques
seabirds include brown boobies (Sula leucogaster),
laughing gulls (Larus atricilla), masked boobies (S.
dactylatra), brown noddies (Anous stolidus) and redbilled tropicbirds (Phaethon aethereus) (Holliday, et al.,
2007; Hodge, et al., 2008; Daltry, 2010). The commonest
land-birds on Dog Island are Caribbean elaenias (Elaenia
martinica), bananaquits (Coereba flaveola) and black-faced
grassquits (Tiaris bicolor). The island is also frequently
used by migratory species travelling between North and
South America (Holliday, et al., 2007; Daltry, 2010; Ross,
2011). A total of 48 resident and migratory bird species
were confirmed on Dog Island by Richard Brown and
Giselle Eagle from January to March 2012 (Bell, 2012).
Dog Island also has an important, albeit reduced,
community of endemic lizards (Hodge, et al., 2003;
Hedges & Conn, 2012). Notable missing reptiles are the
globally threatened Anguilla racer (Alsophis rijgersmaei)
and Lesser Antillean iguana (Iguana delicatissima), which
were presumably present in the past. Dog Island would have
been connected by a land bridge to Anguilla and indeed
the rest of the Anguilla Bank well into the late Pleistocene,
likely until 5,000 years ago. Three globally threatened
species of marine turtles – hawksbill (Eretmochelys
imbricata), green (Chelonia mydas) and leatherback
(Dermochelys coriacea) – nest on and forage around Dog
Island. Freshly dug holes have been recorded on many of
the island’s beaches (Hodge, et al., 2003; Daltry, 2010).
There have been few studies of the invertebrate fauna
of Dog Island. Varnham (2007) collected samples using
pitfall traps, but her specimens have not been identified to
date (Daltry, 2010). Hermit crabs (Coenobita clypeatus)
are present in high numbers, particularly around the coast.
Feral goats are present on the island; a remnant of more
extensive grazing practices (Daltry, 2010). These goats are
the property of the landowner but are hunted regularly,
with or without permission. There are no known native
mammals on Dog Island, but it could potentially support
native bat species.
It is not known when black rats became established on
Dog Island; but this is likely to have occurred sometime
after 1613 when rats were first recorded in the Caribbean
region (Harper & Bunbury, 2015). There is a history
of human habitation on the island (i.e. stone walls and
ruins) and rats may have reached Dog Island during this
occupation or when ships were wrecked along the shores.
Rats have been implicated as causing major impacts on
island biodiversity (Towns, et al., 2006; Jones, et al., 2008)
and they are known to have effects on important species on
Dog Island. House mice (Mus musculus) have never been
recorded on Dog Island (Varnham, 2007; Hilton & Connor,
2008; Bell, 2011).
Eradication operation
The eradication operation was planned to take place in
the dry season (between January and May), when natural
foods for the rats are in short supply and when there
was little, or no risk of the operation being interrupted
by tropical storms or hurricanes. The eradication option
adopted for this project was a ground-based poison
programme using protective bait stations to reduce risk
to non-target species, particularly the reptile and feral
goat populations. The eradication programme ran from 8
February 2012 to 4 April 2012 and included establishing
the bait station grid, poisoning, monitoring and biosecurity.
Biosecurity monitoring ran monthly between April 2012
and February 2014. The final check, species monitoring,
and rat-free declaration ran from 10 to 19 February 2014.
A core team of ten people completed the eradication, six
people completed the biosecurity monitoring and a four-
person team completed the final check. Each operational
task was undertaken and completed as follows:
Bait station grid
A series of parallel tracks was cut through the vegetation
on Dog Island, by a local contracting firm from Anguilla,
between 20 November 2011 and 10 February 2012. Three
additional lines were completed between 27 February and
6 March 2012. The contracting firm used two mechanised
tools; machetes and rakes, to complete the task. One third
of the island (the north-eastern end) had lines that were 30
m apart where the scrub was lower and easier to cut and
the rest of the island, where scrub was much denser and
more difficult to clear, had lines that were 40 m apart; bait
stations were placed every 30 m along these lines. Areas
of manchineel were not cut by the contracting firm (under
arrangement with ANT as they did not want to deal with the
toxic plant) during the track cutting phase of the project,
but the main areas of manchineel were completed by the
eradication team over a one-week period (10–15 February
2012) during the grid establishment phase, and two smaller
stands of manchineel were completed over two four-day
periods during the baiting phase (2–6 and 12–16 March
2012). Protective gloves and clothing and full-face masks
were worn by the team when cutting tracks through the
manchineel to avoid the sap and fumes which can irritate
or blister the skin and cause breathing issues.
The bait station grid was established between 8
February and 15 February 2012. Bait stations were made
from 1.5-litre plastic bottles (with the top and bottoms
removed) donated by the public on Anguilla. These
stations were pegged to the ground with wire “legs” to
prevent movement by wind and/or stabilised with rocks or
other material to reduce interference by feral goats. Bait
was placed in the centre of the station through either end
of the bottle.
Bait stations were placed out on the baiting grid. Mid
and East Cays (not shown in the figures) were baited, but
bait was laid on the surface (i.e. under vegetation and
rocks) as feral goats were not present on the cays and there
were few other non-target species present on these offshore
islets. Each station was marked with flagging tape to ensure
visibility in thicker vegetation.
The entire grid of 1,714 stations was established before
being individually numbered and mapped using GPS and
added to a GIS-linked database (Fig. 1).
Poisoning
Brodifacoum was used in two formulations: Klerat®
(Syngenta, UK), a 20 g, wax-based wax block containing
the bittering agent Bitrex™, and Pestoff® (Animal Control
Products, NZ), a 24 g grain-based block bait. Both had
0.005% active ingredient and were dyed blue (or green/
Fig. 1 Bait station grid on Dog Island, Anguilla (bait station
positions are marked by a black dot).
163
Island invasives: scaling up to meet the challenge. Ch 1C Rodents: Lessons
blue), to be less attractive to birds (Caithness & Williams,
1971; Hartley, et al., 1999; Weser & Ross, 2013).
The poisoning programme commenced on 16 February
2012 and continued through to 30 March 2012. Baits
were present in each station throughout the poisoning
programme and replaced as required. Two bait blocks
were constantly available in each main island bait station
throughout the programme. Klerat® was used as the main
bait (16 February–20 March and 26–30 March). Pestoff®
was only used for checks 20 and 21 (21–25 March) to
target any surviving or rats that had avoided Klerat® for
any reason.
The bait stations on Dog Island were checked and
serviced every 1–4 days. However, the stations on the
offshore cays were only checked twice, during suitable
weather, on team changeover days when the boat was
available. Thus, they had more bait per station (10 blocks)
than on the main island. To present the data on bait-take
gained from these varied bait station checks we grouped
the data into 25 periods or checks (mean (±SEM) = 1.44 ±
0.14 days between checks, range 1–4 days) shown as days
from baiting (Fig. 2).
Bait take was recorded in field notebooks by bait
station number and the species believed to have consumed
or removed the bait as confirmed by sign in and around the
bait station (i.e. pieces or fragments with rat teeth marks
or crab claw marks, etc.). These data were entered into a
database and large-scale maps showing active stations
were produced in real-time to enable the team to effectively
monitor bait take activity and target any “hot spots”. All
rat carcasses found were collected and returned to base for
incineration to reduce risk for non-target scavengers.
Mitigation measures such as using bait stations to
prevent access by goats to the bait, moving bait stations
if crabs interfered with the stations and raising the bait
stations into vegetation were used to reduce the risk of
primary and secondary poisoning to non-target species.
Monitoring
Three distinct periods of monitoring were undertaken
as the project progressed. Monitoring points consisted of
materials attractive to rats (e.g. chocolate flavoured wax
or resin, candles and soap) and tracking tunnels. Intensive
monitoring using 3,428 points at 15–20 m spacing was
carried out from 12 March 2012 to 4 April 2012 to detect
any surviving rats. This was followed by a 22-month
period of long-term monitoring using 167 commercial
lockable plastic bait stations (placed around the coastline
Fig. 2 Amount (in kg) of bait consumed by black rats
(Rattus rattus) at each bait check (marked by black dot)
during the black rat eradication on Dog Island, Anguilla
(Day 1 was 16 February 2012).
164
of Dog Island as long-term biosecurity stations) from 5
April 2012 to 9 February 2014. These biosecurity stations
were established at high risk areas on the island; around
the coast and at seabird breeding sites (Bell, 2012). The
final check, using 626 monitoring points and biosecurity
stations, was carried out between 10 and 19 February 2014.
WMIL, FFI and ANT staff and volunteers carried out the
intensive and final checks and ANT staff and volunteers
maintained the long-term monitoring. All stations were
individually numbered and any evidence of activity (e.g.
teeth marks or foot prints) was recorded in field notebooks
by number and the species believed to have consumed the
wax or soap or marked the tracking plate.
Monitoring items were placed inside and outside
each biosecurity station as well as halfway between each
biosecurity station. Sand traps smoothed out to detect rat
foot prints were established on beaches and inner island
tracks. Checks for active rat runs and activity (i.e. identifying
evidence of predation or scavenging on carcasses, chews
on plants, droppings, etc.) at high-risk sites (i.e. ruins,
seabird colonies, etc.) were also undertaken.
Each monitoring site was checked regularly, either
separately, or during the poisoning phase, together with the
poisoning bait station grid. Any rat and non-target species
sign found on detection devices was recorded and added to
the database.
RESULTS
Bait take was high over most of the island. Green/blue
rat droppings appeared within three days and rats consumed
189 kg of bait. The bait take pattern was typical of other
bait station rat eradication campaigns (Thomas & Taylor
2002). It was very high in the days immediately after the
first bait loading (checks 1–3) and dropped to a relatively
low level 20 days after initial baiting (check 10). A small
increase was recorded at day 22 after initial baiting (check
12) but dropped away, reaching zero bait take on day 26
after the initial baiting (check 15) (Fig. 2).
Throughout the poisoning phase, 89% of bait stations
were visited by rats, with 58% active within nine days of
the initial baiting. The high number of active bait stations
during the first two bait rounds shows that the rats quickly
accepted the bait over most of the island.
The average number of blocks removed was 6.18 (±
0.07) blocks per station (Range: 0–16.05). As shown by
Fig. 3, bait take was not evenly distributed over the entire
island, with the greatest level of bait take at the eastern
end where the main sea bird colony was situated, and the
centre of the island. Bait take was also recorded on all the
offshore stacks. Rats were also quickly eradicated from the
cays as bait was still present when the second baiting visit
was undertaken.
Fig. 3 Distribution of total bait take by black rats (Rattus
rattus), as bait blocks consumed per station, during the
black rat eradication on Dog Island, Anguilla. Darker
shading indicates higher levels of bait uptake by rats.
Bell, et al.: Dog Island using ground-based techniques
baiting). Bait-take showed that the rat population was
not evenly distributed across the island. Apparently high
concentrations of rats where the seabird colonies are
present suggests rats were likely to have been having an
effect on these nesting seabirds.
Fig. 4 Amount of bait (in kg) consumed by hermit crabs
(Coenobita clypeatus) at each bait check (marked by
black dot) during the black rat (Rattus rattus) eradication
on Dog Island, Anguilla.
There was substantial interference by hermit crabs
(Fig. 4). The average number of blocks eaten by crabs was
estimated at 14.7 (± 0.3) blocks per station (Range 0–48.3).
Crabs took a few days to become habituated to the bait,
but then crab activity levels remained high (i.e. from days
19–41 over 50% of the stations were visited by crabs each
round) (Fig. 4). There were 1,586 (92.5%) bait stations
that had crab activity during the poisoning phase and only
128 stations were not affected. Crabs ate an estimated
467.4 kg of Klerat® bait throughout the eradication. There
was no evidence that crabs were adversely affected by
the bait. Anticoagulants are considered unlikely to affect
invertebrates, as most have an open circulation system
and have different physical and chemical clotting systems
compared to vertebrates (Pain, et al., 2000).
Other non-target species had interfered with the bait
to lesser amounts; goats consumed 0.04 kg (2 blocks)
of Klerat® bait, ground lizards 0.26 kg (10.8 blocks)
of Pestoff® bait, ants 10.9 kg of Klerat® bait and other
insects 0.33 kg of Klerat® bait.
No animals, other than rats, exhibited signs of poisoning
and no suspicious mortalities were recorded over the 11week operation. The team was trained to observe nontarget behaviour and collect any carcasses. There were 160
rat carcasses collected on the surface during the operation.
These carcasses were collected and incinerated on the
island to prevent availability to non-target species.
Monitoring for rat presence continued island-wide for
two years after the end of the poisoning operation. The last
rats were detected on 13 March 2012 during the overlap
between the poisoning and intensive monitoring phases
and these rats were successfully targeted using Klerat®
by 30 March 2012. No rats or sign were detected during
any phase of the long-term or final check monitoring. Dog
Island was declared rat-free in May 2014.
DISCUSSION
The success of the Dog Island black rat eradication
shows that a well-planned, adequately resourced, wellexecuted programme, supported by the landowner and
directed by experienced operators and completed during
the dry season can eradicate black rats from a large,
arid, tropical island using a ground-based bait station
operation. Dog Island is now the largest Caribbean island
to be cleared of invasive rats and we believe that similar
techniques could be utilised on other, even larger, islands
in the Caribbean region.
Once the poison grid was established, the island was
cleared of rats within four weeks (25 days from initial
Importantly there were no known non-target species
affected by this operation despite intensive searches for
carcasses and a high level of interference by land crabs
and to a lesser degree by ground lizards, invertebrates
and, on one occasion, a goat which ate two blocks of bait.
This stands in marked contrast to other operations that
have inadvertently poisoned a variety of birds and other
native wildlife (e.g. Howald, et al., 2007, Fisher, et al.,
2011, Pitt, et al., 2015). Our choice of bait was a critical
factor to this success; the primary bait used was Klerat®
which was consistently untouched by any vertebrate other
than rats (whereas the goat and lizards ate Pestoff® bait
only). Klerat® has been equally successful in almost all
the previous rat eradications in the Caribbean completed or
managed by the authors and others (Day & Daltry, 1996;
Daltry, 2000; Garcia, et al., 2002; Varnham, 2003; Varnham
& Daltry, 2006; Witmer, et al., 2007; Varnham, 2010).
Ecological surveys conducted prior to the rat
eradication operation identified a suite of ecological
indicators on Dog Island that were consistent with the
impacts of black rats, including the suppressed diversity
and abundance of land birds, lizards and plants (Daltry,
2010). Audubon’s shearwaters (Puffinus lherminieri) were
first confirmed nesting on the island in 2012, within a few
weeks of eradicating the rats (Bell, 2012). Preliminary
surveys in 2014 found significant increases in a number
of native species since the rats were eradicated; a twofold increase in the density of ground lizards, three-fold
increase in abundance of land-birds and a three-fold
increase in burrow occupancy of Audubon’s shearwaters
(Bell & Daltry, 2014). Further increases were recorded
during routine monitoring in 2016 and are predicted to
continue over the next 10–20 years. Birds, lizards, goats,
vegetation and invertebrates should be monitored for the
next 20 years to detect and assess longer-term changes to
the Dog Island ecosystem.
Unfortunately, as long as goats remain on Dog
Island, some of the benefits of removing rats may be
significantly reduced or fail to occur at all (Daltry, 2010).
By preferentially eating all but the most spiny and toxic
plants, the goats are maintaining an artificial, plagioclimax
vegetation of thorny scrub across most of the interior of
the island, which has low diversity and supports relatively
few animals. Our cross-island transects, for example,
revealed these interior areas had an extremely low density
of lizards (Bell & Daltry, 2014). Another major concern
about the goat herd is that it attracts parties of hunters who
pose a biosecurity risk because their vessels and gear could
provide pathways for rodents and other pests to invade the
island.
While eradicating rats from Dog Island is a considerable
achievement, it is important to stress that keeping this
island rodent-free will require constant vigilance and
commitment from all agencies, interested parties and the
Anguillan community to prevent, detect and respond to
any incursions. Prevention of rat re-infestation should be
the primary aim. The greatest risk of reinvasion by rats
reaching Dog Island is with private vessels, charter boats
and fishing boats, particularly those that moor overnight,
from Anguilla or the other nearby islands such as the Prickly
Pear Cays and Saint Martin. This is especially so when
equipment and food are brought to the island. Permanent
biosecurity stations have been established on Dog Island
and these will be maintained indefinitely by trained ANT
staff. An incursion response plan has also been developed
by ANT to deal with any rats that may be detected in the
165
Island invasives: scaling up to meet the challenge. Ch 1C Rodents: Lessons
future. This shows that the local conservation agencies
are totally committed to the restoration of this important
Caribbean island.
ACKNOWLEDGEMENTS
This project was carried out with funding from National
Fish and Wildlife Foundation (Grant No. 2011-0002000000 to FFI Inc.) and the Governor’s Discretionary
Fund (awarded to ANT). The Klerat™ bait was donated
by Syngenta Crop Protection AG. We would like to thank
Alan Gumbs (Anguilla Development Company, owner
of Dog Island), Jacqueline Cestero (Project Manager),
Clarissa Lloyd, Janeczka Richardson, Kim Gumbs, Louise
Soanes, Devon Carter and Michael Matthew (ANT staff),
Jenny Bright and James Millett (RPSB), Robert BenstedSmith, Sara Calçada, Nav Dayanand, A.E. Lovett and
Mary Rider (FFI), Scott Hall (NFWF Seabirds Keystone
Initiative), Judy Pierce (US Fish & Wildlife Service), Kirsty
Swinnerton and Richard Griffiths (Island Conservation),
Mark Birchmore, Alan Buckle, Andy Bywater and Alex
Cornish (Syngenta), Bill Simmons (Orillion, previously
Animal Control Products, New Zealand), Keith Broome
(Island Eradications Advisory Group, Department of
Conservation, New Zealand), Sharrine Gumbs (Uneek
Supplies), Lee Brooks (Ashley & Sons), Marshall
Mitchell (Safe Cargo), Maximo Hodge and Marvin Carty
(Trucking), Linda Bottlik (Sunfish Design), Jackie Pacsher
(Island Dream Rental), Kathy Haskins, Rachel HaskinsBrodie and Kieron Brodie (Sandy Feet Car Rental) and
Hon. Othlyn Vanterpool (Anguilla Scouts). Special thanks
to all of the volunteers (Thomas Aveling, Luke Banse,
Richard Brown, Chris Clarke, Giselle Eagle, Paul GarnerRichards, Melissa May, Tegan Newman, Matthew Rogers
and Sophie Thomas) who worked on Dog Island during
the eradication. Emile Lloyd, Rumple Lloyd and Irwin
Lloyd provided boat transport to Dog Island. Neville Carty,
Ashton Richardson, Khaloni Richardson, Giovani Hughes,
Leopold Reid, Clayton Reid, Andrew Brooks, Wells
Allwalters, Paul Gillis, Peter Liddie, Movis Connor, Evan
Hodge, and Desmond Brooks cut tracks across Dog Island
to establish the bait station grid. Kelvin Floyd (WMIL)
designed the GIS-linked database and provided Figs 1 and
2.
REFERENCES
Atkinson, I.A.E. (1985). ‘The Spread of Commensal Species of Rattus
to Oceanic Islands and their Effects on Island Avifaunas’. In: P.J.
Moors (ed.) Conservation of Island Birds, pp. 35–81. ICBP Technical
Publication 3. Cambridge, UK: International Council for Bird
Preservation (ICBP).
Bell, E. (2011). Dog Island Restoration Project: Operational Plan for the
Eradication of Black Rats (Rattus rattus) from Dog Island, Anguilla.
New Zealand: Wildlife Management International Ltd.
Bell, E. (2012). Dog Island Restoration Project: Technical Report for the
Eradication of Black Rats (Rattus rattus) from Dog Island, Anguilla.
New Zealand: Wildlife Management International Ltd.
Bell, E. and Daltry, J. (2014). Dog Island Restoration Project: Twoyear
Assessment Following the Eradication of Black Rats (Rattus rattus)
from Dog Island, Anguilla. New Zealand: Wildlife Management
International Ltd and UK: Fauna & Flora International.
Caithness, T. and Williams, G. (1971). ‘Protecting birds from poisoned
baits’. New Zealand Journal of Agriculture 122: 38–43.
Campbell, E. (1991). ‘The effect of introduced roof rats on bird diversity
of Antillean Cays’. Journal of Field Ornithology 62(3): 343–348.
Daltry, J. (2000). The Second Rat Eradication on Praslin Island, Saint
Lucia: August–September 2000. Jersey: Saint Lucia Forest and Lands
Department, Saint Lucia and the Durrell Wildlife Conservation Trust.
Daltry, J., Bloxam, Q., Cooper, G., Day, M., Hartley, J., Henry, M.,
Lindsay, K. and Smith, B. (2001). ‘Five years of conserving the ‘world’s
rarest snake’, the Antiguan racer Alsophis antiguae’. Oryx 35: 119–127.
Daltry, J. (2010). Observations on the Ecology of Dog Island, Anguilla,
2010. Anguilla Development Company, Anguilla National Trust,
Department of Environment (Government of Anguilla) and DEFRAFera (UK Government).
Day, M. and Daltry, J. (1996). ‘Rat eradication to conserve the Antiguan
Racer’. Aliens 3: 14–15.
166
Fisher, P., Griffiths, R., Speedy, C. and Broome, K. (2011). ‘Environmental
monitoring for brodifacoum residues after aerial application of baits
for rodent eradication’. In: C.R. Veitch, M.N. Clout and D.R Towns
(eds.) Island invasives: eradication and management, pp. 300–304.
Occassional Paper SSC no. 42. Gland, Switzerland: IUCN and
Auckland, New Zealand: CBB.
Garcia, M., Diez, C. and Alvarez, A. (2002). ‘The eradication of Rattus
rattus from Monito Island, West Indies’. In: C.R. Veitch and M.N. Clout
(eds.) Turning the tide: the eradication of invasive species, pp. 116–119.
Occasional Paper SSC no. 28. IUCN SSC Invasive Species Specialist
Group. IUCN, Gland, Switzerland and Cambridge, UK.
Harper, G. and Bunbury N. (2015). ‘Invasive rats on tropical islands: Their
population biology and impacts on native species’. Global Ecology and
Conservation 3: 607–627.
Hartley, L., O’Connor, C., Waas, J. and Matthews, L. (1999). ‘Colour
preferences in North Island robins (Petroica australis): Implications for
deterring birds from poisonous baits’. New Zealand Journal of Ecology
23: 255–259.
Hedges, S. and Conn, C. (2112). ‘A new skink fauna from Caribbean
islands (Squamata, Mabuyidae, Mabuyinae)’. Zootaxa 3288: 1–244.
Hilton, G. and Connor, R. (2008). Mouse Detection on Dog Island.
Anguilla: Anguilla National Trust and UK: Royal Society for the
Protection of Birds.
Hodge, K., Censky, E. and Powell, R. (2003). The Reptiles and
Amphibians of Anguilla, British West Indies. The Valley, Anguilla: The
Anguilla National Trust.
Hodge, K., Holliday, S. and Mukhida, F. (2008). ‘Anguilla’. In: D. Wege,
and V. Anadon–Irizarry (eds.) Important Bird Areas in the Caribbean,
pp. 27–34. Cambridge, UK: BirdLife International.
Holliday, S., Hodge, K. and Hughes, D. (2007). A Guide to the Birds of
Anguilla. Sandy, Bedfordshire, UK: Royal Society for the Protection
of Birds.
Howald, G., Donlan, C., Galvan, J., Russell, J., Parkes, J., Samaniego, A.,
Wang, Y., Veitch, D., Genovesi, P., Pascal, M., Saunders, A. and Tershey,
B. (2007). ‘Invasive rodent eradication on islands’. Conservation
Biology 21: 1258–1268.
Jones, H., Tershy, B., Zavaleta, E., Croll, D., Keitt, B., Finkelstein, M.
and Howald, G. (2008). ‘Review of the global severity of the effects of
invasive rats on seabirds’. Conservation Biology 22: 16–26.
Moors, P.J. and Atkinson, I.A.E. (1984). ‘Predation on Seabirds by
Introduced Animals and Factors Affecting its Severity’. In: J.P. Croxall,
P.G.H. Evans and R.W. Schreiber (eds.) Status and Conservation
of the World’s Seabirds, pp. 667–690. ICBP Technical Publication 2.
Cambridge, UK: International Council for Bird Preservation (ICBP).
Pain, D., Brooke, M., Finnie, J. and Jackson, A. (2000). ‘Effects of
brodifacoum on the land crabs of Ascension Island’. Journal of Wildlife
Management 64: 380–387.
Pitt, W., Berentsen, A., Shiels, A., Volker, S., Eisemann, J., Wegmann,
A. and Howald, G. (2015). ‘Non–target species mortality and the
measurement of brodifacoum rodenticide residues after a rat (Rattus
rattus) eradication on Palmyra Atoll, tropical Pacific’. Biological
Conservation 185: 36–46.
Ross, T. (2011) Dog Island: Brief Information and Data. Cambridge, UK:
Fauna & Flora International.
Sanders, S. (2006). Important Bird Areas in the United Kingdom Overseas
Territories. Sandy, UK: Royal Society for the Protection of Birds.
Thomas, B. and Taylor, R. (2002). ‘A history of ground-based rodent
eradication techniques developed in New Zealand, 1959–1993’. In:
C.R. Veitch and M.N. Clout (eds.) Turning the tide: the eradication of
invasive species, pp. 301–310. Occasional Paper SSC no. 28. IUCN
SSC Invasive Species Specialist Group. IUCN, Gland, Switzerland and
Cambridge, UK.
Towns, D. and Broome, K. (2003). ‘From small Maria to massive
Campbell: Forty years of rat eradications from New Zealand islands’.
New Zealand Journal of Zoology 30: 377–398.
Towns, R., Atkinson, I. and Daugherty, C. (2006). ‘Have the harmful
effects of introduced rats on islands been exaggerated?’. Biological
Invasions 8: 863–891.
Varnham, K. (2003). Eradication of Black Rats (Rattus rattus) from Sandy
Cay, British Virgin Islands. UK: Royal Society for the Protection of
Birds.
Varnham, K. and Daltry, J. (2006). Eradication of Invasive Black Rats
(Rattus rattus) from Green and York Islands, Antigua: June–July 2006’.
UK: Royal Society for the Protection of Birds and UK: Fauna & Flora
International.
Varnham, K. (2007). Eradicating Introduced Black Rats (Rattus rattus)
from Dog Island, Anguilla: A Feasibility Study. UK: Royal Society for
the Protection of Birds.
Varnham, K. (2010). ‘Invasive rats on tropical islands: their history,
ecology, impacts and eradication’. RSPB Research Report No. 41. UK:
Royal Society for the Protection of Birds.
Witmer, G., Boyd, F. and HillisStarr, Z. (2007). ‘The successful
eradication of introduced roof rats (Rattus rattus) from Buck Island
using diphacinone, followed by an irruption of house mice (Mus
musculus)’. Wildlife Research 34: 108–115.
Weser, C. and Ross, J. (2013). ‘The effect of colour on bait consumption
of kea (Nestor notabilis): Implications for deterring birds from toxic
baits’. New Zealand Journal of Zoology 40(2): 137–144.
A.L. Bond, R.J. Cuthbert, G.T.W. McClelland, T. Churchyard, N. Duffield, S. Havery, J. Kelly, J.L. Lavers, T. Proud, N. Torr, J.A. Vickery and S. Oppel
Bond, A.L.; R.J. Cuthbert, G.T.W. McClelland, T. Churchyard, N. Duffield, S. Havery, J. Kelly, J.L. Lavers, T. Proud, N. Torr, J.A. Vickery and S. Oppel.
Recovery of introduced Pacific rats following a failed eradication attempt on subtropical Henderson Island, South Pacific Ocean
Recovery of introduced Pacific rats following a failed eradication
attempt on subtropical Henderson Island, South Pacific Ocean
A.L. Bond1,4, R.J. Cuthbert1,5, G.T.W. McClelland1, T. Churchyard1, N. Duffield1, S. Havery1, J. Kelly2,
J.L. Lavers1, T. Proud1, N. Torr1, J.A. Vickery1 and S. Oppel3
RSPB Centre for Conservation Science, Royal Society for the Protection of Birds, The Lodge, Sandy, Bedfordshire,
SG19 2DL, UK. <a.bond@nhm.ac.uk>. 2Royal Society for the Protection of Birds, The Lodge, Sandy, Bedfordshire,
SG19 2DL, UK. 3RSPB Centre for Conservation Science, Royal Society for the Protection of Birds, David Attenborough
Building, Pembroke Street, Cambridge, CB2 3EZ, UK. 4Present address: Bird Group, Department of Life Sciences,
The Natural History Museum, Akeman Street, Tring, Hertfordshire, HP23 6AP, UK. 5Present address: Conservation
Solutions, 9 Prospect Drive, Belper, Derbyshire, DE56 1UY, UK.
1
Abstract Rodent eradications in tropical environments are often more challenging and less successful than those in
temperate environments. Reduced seasonality and the lack of a defined annual resource pulse influence rodent population
dynamics differently than the well-defined annual cycles on temperate islands, so an understanding of rodent ecology
and population dynamics is important to maximise the chances of eradication success in the tropics. Here, we report on
the recovery of a Pacific rat (Rattus exulans) population on Henderson Island, South Pacific Ocean, following a failed
eradication operation in 2011. We assessed changes in the rat population using capture rates from snap-trapping and
investigated seasonality by using capture rates from live-trapping. Following the failed eradication operation in 2011, rat
populations increased rapidly with annual per capita growth rates, r, of 0.48–5.95, increasing from 60–80 individuals to
two-thirds of the pre-eradication abundance within two years, before decreasing (r = -0.25 – -0.20), presumably as the
population fluctuated around its carrying capacity. The long-term changes in rat abundance may, however, be confounded
by short-term fluctuations: four years after the eradication attempt we observed significant variation in rat trapping rates
among months on the plateau, ranging from 36.6 rats per 100 corrected trap-nights in mid-June to 12.6 in late August.
Based on mark-recapture, we also estimated rat density fluctuations in the embayment forest between 20.4 and 42.9 rats
ha-1 within one month in 2015, and a much lower rat density on the coral plateau fluctuating between 0.76 and 6.08 rats
ha-1 in the span of two months. The causes for the short-term density fluctuations are poorly understood, but as eradication
operations on tropical and subtropical islands become more frequent, it will be increasingly important to understand the
behaviour and ecology of the invasive species targeted to identify times that maximise eradication success.
Keywords: introduced species, island restoration, Pitcairn Islands, rodents, spatially explicit capture-recapture
INTRODUCTION
The removal of introduced rodents from islands is an
increasingly important tool for the conservation of island
biodiversity, and has been successful in hundreds of cases
(Lorvelec & Pascal, 2005; Howald, et al., 2007; Bellingham,
et al., 2010; Russell & Holmes, 2015). Introduced rodents
have been eradicated from >580 islands (Keitt, et al., 2015;
DIISE, 2016) and rodent eradications are one of the most
cost-effective methods of preserving island biodiversity
(Howald, et al., 2007; Jones, et al. 2016).
The success rate of rodent eradications has improved
as eradication tools and methods become more refined.
However, failures still occur, especially on tropical islands
where conditions that can increase the risk of eradication
failure, such as aseasonal breeding, are more likely
(Varnham, 2010; Holmes, et al., 2015). While undesirable,
these unsuccessful projects still provide an opportunity
to advance conservation science, often through post hoc
review of operational planning and implementation (Keitt,
et al., 2015). However, they also present potentially unique
occasions to further understand invasion biology. For
example, the population dynamics of surviving rodent
populations following such failed eradication attempts are
seldom studied (Hein & Jacob, 2015) despite being useful
for predicting population dynamics during new invasions
(Nathan, et al., 2015).
In particular, there is currently little knowledge on
how much time elapses before tropical rodent populations
can reach an island’s presumed carrying capacity after a
severe population bottleneck, but such information could
be useful to inform the post-operation monitoring interval
that determines whether an eradication operation has
been successful or not (Samaniego-Herrera, et al., 2013).
On temperate islands, two years encompasses two rat
breeding seasons, and is typically sufficient to determine
an eradication operation’s success. In the tropics, rats have
a less constrained timing of breeding, and a breeding cycle
as short as four months, so a shorter time may be required
to reliably detect a recovering rat population, particularly
in wetter conditions (Keitt, et al., 2015).
For many widespread invasive rodents, however, there
is a lack of basic ecological knowledge about densities, and
the factors affecting the large variation in abundance that
is evident for highly versatile invasive rodents (Harper &
Bunbury, 2015). Henderson Island (24°20’S, 128°19’W),
in the subtropical Pitcairn Islands of the South Pacific, was
subject to an aerial poison bait-based eradication attempt
of the introduced Pacific rat (Rattus exulans) in 2011 (Torr
& Brown, 2012). The eradication was unsuccessful, but the
cause of the failure was neither operational shortcoming
nor due to resistance of rats to brodifacoum pellets (Torr &
Brown, 2012; Amos, et al., 2016, Brooke, 2019).
Here, we report on the population recovery of R.
exulans on Henderson Island up to four years following
a failed eradication attempt, and provide information on
short-term seasonality in density of Pacific rats using live
trapping and a spatially explicit capture-recapture (SECR)
framework. We use the obtained estimates in a rapid
eradication assessment (Russell, et al., 2017) to provide
guidance on the length of a post-operation monitoring
period after which an eradication could be considered
successful with 95% certainty. These data provide a robust
overview of the short- and long-term population variability
on an aseasonal sub-tropical island that will inform future
conservation management.
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 167–174. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
167
Island invasives: scaling up to meet the challenge. Ch 1C Rodents: Lessons
MATERIALS AND METHODS
Study area
Henderson Island, a UNESCO World Heritage Site, is
a 43 km2 raised coral atoll in the Pitcairn Islands, South
Pacific Ocean, with a tropical climate (Spencer, 1995;
Weigelt, et al., 2013). The island was subjected to an
unsuccessful aerial eradication attempt of the introduced
Pacific rat (Rattus exulans) in 2011 (Torr & Brown,
2012). While the ultimate cause of the eradication failure
remains unknown, resistance of rats to brodifacoum pellets
or operational errors are not considered factors (Torr &
Brown, 2012; Amos, et al., 2016).
We conducted our study at the northern end of the
island’s plateau and along the two accessible beaches, North
Beach and East Beach. The plateau substrate is fossilised
coral with a uniform, dense native vegetation consisting
of mostly of Pandanus tectorius, Xylosma suaveolens
and Psydrax odorata (Waldren, et al., 1995). The beach
and embayment forest (“beach back”) areas have a sandy
substrate with a mixed low vegetation and small stands
of introduced coconut (Cocos nucifera) (Waldren, et al.,
1995).
Rat snap-trapping and long-term abundance indices
We estimated rat abundance indices in 2009
(September), 2012 (May and November), 2013 (August),
and 2015 (October and November) on the plateau and
embayment forest areas of North and East beaches of
Henderson using snap-traps, though the precise methods
differed because of logistical and time constraints. In all
years, however, we set traps between 16:00–18:00 h (all
times UTC-8), and checked the following morning between
08:00–10:00 h (Table 1).
We recorded the traps’ contents (rat, crab, or snapped
and empty) to calculate an index of abundance as the
number of rats caught per 100 corrected trap-nights (100
CTN; Nelson & Clark, 1973), where
size and the snap-trapping rate to extrapolate population
sizes in other years. For 2011, when no snap-trapping
data were available, we used the population estimate of
60–80 individuals that was estimated to have survived
the eradication attempt in 2011 based on genetic markers
(Amos, et al., 2016). In each year, we proportioned the
total population to the three different habitats in which we
measured rat abundance based on their relative area: North
Beach embayment forest (7 ha), East Beach embayment
forest (7 ha), and the island plateau (4,290 ha), and the
initial population based on the abundance indices in these
three habitats in 2009. We assumed that trapability was
constant among years.
Rat live trapping and density estimation
To obtain a robust estimate of rat density and to
document short-term fluctuation in rat density over six
months, we implemented a spatial capture-mark-recapture
programme in 2015 (Oppel, et al., 2019). Rats were livetrapped on the plateau from 28 May to 16 October 2015
during seven primary sessions of 10 consecutive trapnights each, followed by a window of 10–15 days with no
trapping between primary sessions. We established a trap
network placed along 3.5 km of cleared paths (Fig. 1),
and traps were arranged at distances from 3–20 m along
343 locations, with a different subset of 250 trap locations
used during each primary session. Because our original
traps (Sherman and Elliott aluminum boxes; model LFA,
23 × 9 × 8 cm, H.B. Sherman Traps Inc., Tallahassee,
Florida, USA) were easily damaged by crabs, they were
replaced by larger and more robust Tomahawk cage traps
in September (27 × 16 × 13 cm, Key Industries, Auckland,
New Zealand).
In the embayment forest at North Beach, rats were livetrapped during three primary occasions of 6–10 trap-nights
each between 1 August and 19 September 2015 using 38
CTN = Total trap-nights – Trap-nights lost (equation 1);
Trap-nights lost = ½ × (crab captures + snapped traps)
(equation 2).
Estimating long-term rat population change
Based on the rat abundance indices derived
from snap-trapping, we estimated the annual
per capita growth rate, r, using the formula:
(equation 3), where N is the population
estimate at time t and t-1, and t is the elapsed time, in years,
between the two estimates.
To estimate a population growth rate, which requires
non-zero values in each time interval if no immigration
is assumed, we scaled abundance indices derived from
snap-trapping to an island population size. This approach
allowed us to have all the population estimates on the same
scale, and to include the very small population size in 2011.
We extrapolated population size based on live- and snaptrapping data from 2009: based on live-trapping, there
were approximately 28 rats/ha (95% confidence interval:
23–40 rats/ha) in the embayment forest of North Beach on
Henderson (Cuthbert, et al., 2012), which corresponded to
31.7 rats 100 CTN-1 in the same habitat. We extrapolated
the density estimate to an approximate population size
of 120,000 (range: 104,000–172,000) rats on the island,
assuming equal density across all habitat types, and used
the relationship between this extrapolated population
168
Fig. 1 Map of the path network on the north end of
Henderson Island showing the locations of all live traps
used in 2015 (+), snap-trapping grid on North Beach ()
and the location of the research camp ().
Bond, et al.: Rat recovery, Henderson Island
Table 1 Snap trap arrangements and bait used on Henderson Island from 2009–2015.
Habitat
Start date
End date
Bait
Trap spacing
Embayment forest
– East Beach
12 Sep 09
21 Sep 09
Coconut, peanut butter, rolled oats
10–15 m
14 Aug 11
23 Aug 11
23 Nov 12
30 Nov 12
Peanut butter, coconut
10–15 m
21 Aug 13
28 Aug 13
Peanut butter, coconut
10–15 m
12 Sep 09
21 Sep 09
Coconut, peanut butter, rolled oats
10–15 m
14 Aug 11
23 Aug 11
03 May 12
06 May 12
Peanut butter, rolled oats, chocolate
8–10 m
23 Nov 12
30 Nov 12
Peanut butter, coconut
10–15 m
21 Aug 13
28 Aug 13
Peanut butter, coconut
10–15 m
21 Oct 15
31 Oct 15
Coconut, Pandanus
20 m
01 Nov 15
14 Nov 15
Coconut, Pandanus
20 m
12 Sep 09
21 Sep 09
Coconut, peanut butter, rolled oats
10–15 m
14 Aug 11
23 Aug 11
03 May 12
06 May 12
Peanut butter, rolled oats, chocolate
8–10 m
23 Nov 12
30 Nov 12
Peanut butter, coconut
10–15 m
Embayment forest
– North Beach
Plateau
21 Aug 13
28 Aug 13
Peanut butter, coconut
10–15 m
21 Oct 15
31 Oct 15
Coconut, Pandanus
20 m
01 Nov 15
14 Nov 15
Coconut, Pandanus
20 m
traps arranged in a 6 × 6 configuration with traps spaced 10
m apart, and we expanded the trapping grid to 63 traps in a
7 × 9 configuration for the last primary session.
Before the first primary session in each habitat, traps
were deployed, but not opened, for approximately five
days to allow rats to overcome neophobia. For each
ten-day trapping period, traps were baited with a small
(approximately 1 × 1 cm) cube of fresh coconut between
16:00–18:00 h, and checked the following morning
between 08:00–10:00 h.
During trap checks each captured rat was fitted with a
uniquely numbered ear tag, or the number of an existing
ear tag was recorded, and the rat was released next to the
trap. We recorded the trap location for each capture and
recorded whether traps were available to capture rats or
had been rendered ineffective (e.g. by crabs). We estimated
a capture index (rats/100 corrected trap nights) for the
plateau and the embayment forest for each trap-night using
the same equation as above to correct for inactivated traps.
To estimate rat densities, we used spatially-explicit
capture-recapture models, which have been used
successfully for other rat density estimations on islands
(Russell, et al., 2011; Ringler, et al., 2014; Harper, et al.,
2015). We assumed that rat home ranges were randomly
located with respect to trap locations and stationary within
a given primary session, and that the central location of the
home range was adequately described by a homogenous
Poisson distribution (Efford, 2004; Borchers, 2012).
Capture probability of rats at a given trap was based on
the distance of the rat’s home range centre from the trap
and was modelled with a half-normal function in the
embayment forest (Borchers & Efford, 2008; Harper, et al.,
2015) and a negative exponential function on the plateau
where the distribution of rat movements included a long
tail of some very large movements >500 m. We estimated
density using the function ‘secr.fit’ in the package ‘secr’
(Efford, 2016) using R 3.2.5 (R Core Team, 2017) for each
habitat and primary session separately, thus allowing for
density, capture probability, and the movement parameter,
σ, to vary over time and habitat. We did not consider
trap dependence. We report estimates of density, capture
probability and σ with 95% confidence intervals.
Rapid eradication assessment
During the eradication operation in 2011, a team
remained on the island for three months after the bait drop
(from August to November), and any future eradication
operation will require a similar post-operational period
to monitor non-target species (Oppel, et al., 2016). We
therefore estimated whether rat monitoring at all 406
trap locations of our two networks could conclude that
an eradication had been successful with 95% certainty if
no rat was detected during three 10-day trapping sessions
up to three months after the bait drop. We also explored
whether certainty could be increased if a larger area was
covered with traps, and simulated a 30 × 30 m trapping
grid over 10%, 30%, and 50% of Henderson Island. We
used our empirical estimates of population growth rate and
rat roaming behaviour in a rapid assessment tool (REA
Shiny; Russell, et al., 2017) assuming a prior probability of
success of 83.9% (Russell & Holmes, 2015), no reinvasion
(Amos, et al., 2016), and rat dispersal distances of up to
500 m (Oppel, et al., 2019). We present the probability of
successful eradication that could be inferred given that no
rat was detected during the specified survey effort.
RESULTS
Rat abundance estimates and long-term population
recovery
We trapped rats from 11 August to 21 September 2009,
catching 233 rats in 734.5 corrected trap-nights overall, or
31.7 rats/100 CTN, with little difference among habitats
169
Island invasives: scaling up to meet the challenge. Ch 1C Rodents: Lessons
(29.0–33.4 rats/100 CTN; Table 2). The eradication attempt
in August 2011 reduced the Henderson rat population to
60–80 individuals (Amos, et al., 2016), and eight months
after the eradication one rat was caught on the plateau in
96.5 corrected trap nights. From 23–30 November 2012,
we caught 9.2–14.8 rats/100 CTN across all three habitats
(Table 2).
In 2013, we caught 20.0–73.2 rats/100 CTN, and the
abundance index exceeded the pre-eradication estimate
in the embayment forest (by more than 100% on North
Beach), while the population on the plateau was 62% of
pre-eradication levels.
In October 2015, we caught again more rats/100
CTN in embayment forest habitat on North Beach (42.9
rats/100 CTN) than on the plateau (13.2 rats/100 CTN)
corresponding to an abundance index similar to preeradication conditions in the embayment forest, but only
41% of pre-eradication levels on the plateau (Table 2).
Based on these rat abundance indices, the rat population
appears to have recovered rapidly, with annual per capita
growth rates ranging from 0.48 to 5.95 (Table 2) during
the recovery phase. The estimated number of rats reached
peaks of 113%, 219%, and 62% of pre-eradication levels
on East Beach, North Beach, and the plateau, respectively,
by 2013 (Table 2), two years after the eradication attempt.
The annual population growth rate has decreased since
2013 and was slightly negative between 2013 and 2015
(Table 2).
Short-term fluctuation in rat density
Overall in 2015, we recorded a total of 2,826 rat
captures in 7,552 corrected trap-nights in our live-trapping
network on the plateau and 319 captures in 684 corrected
trap-nights in the embayment forest. Trapping rates in the
embayment forest were much higher than on the plateau
and less variable over time (Fig. 2). On the plateau, the
trapping index declined from 36.6 rats/100 CTN in early
July to 12.6 rats/100 CTN in late August (Fig. 2). The
subsequent increase to 75.8 rats/100 CTN occurred after
switching Sherman traps with Tomahawk traps, and any
population increase is therefore confounded by a potentially
more effective trap type. To account for habitat- and timespecific variation in capture probability, we estimated rat
density using spatially explicit capture-recapture models
for each primary session.
Rat density in the embayment forest was about 10×
higher than that on the plateau (Fig. 2), and there were
significant temporal fluctuations in both habitat types:
apparent rat densities declined by 50% within one month in
the embayment forest, and by 85% within two months on
the plateau before recovering to 80% of the original density
another three months later (Fig. 2). Lower rat densities on
the plateau coincided with increased rat roaming distances
(σ), which were generally larger on the plateau than in the
embayment forest (Fig. 2). Despite some very long rat
movements on the plateau, only three individuals were
recorded in both the embayment forest and on the plateau
(Fig. 1).
Table 2 Abundance indices of Rattus exulans on Henderson Island increased markedly following an eradication attempt
in August 2011. August 2011 population estimate from Amos, et al. (2016); 2009 rat population from Brooke, et al.,
(2010a), and resulting population estimates are calculated from the relationship between rats/100 corrected trap nights
(CTN, see text for details) and population size. Annual per capita growth rate, r, is based on exponential population
growth.
Habitat
Embayment
forest – East
Beach
Embayment
forest – North
Beach
Plateau
Start – end date
Rats
Corrected Rats
100
trap nights caught CTN-1
12 – 21 Sep 2009
14 – 23 Aug 2011
252
23 – 30 Nov 2012
21 – 28 Aug 2013
12 – 21 Sep 2009
14 – 23 Aug 2011
No of
ratsa
% of
Time between Annual
original surveys
growth
popn
(months)
rate r
73
-
29.0
-
210
5
2%
22.8
83
92
11
30
13.3
32.6
96
236
46%
113%
15.0
8.7
374
-
2.36
1.24
125
33.4
210
03 – 6 May 2012
88.5
0
0.0
5
7
2%
3%
22.8
8.4
0.48
23 – 30 Nov 2012
21 – 28 Aug 2013
21 – 31 Oct 2015
01 – 14 Nov 2015
67.5
82
49
149.5
10
60
21
36
14.8
73.2
42.9
24.1
93
460
269
151
44%
219%
128%
72%
6.6
8.7
25.8
1.0
4.70
2.21
-0.25
NAb
12 – 21 Sep 2009
14 – 23 Aug 2011
03 – 6 May 2012
23 – 30 Nov 2012
108.5
96.5
272
35
1
25
32.3
1.0
9.2
120,120
60
3,859
34,225
0%
3%
28%
22.8
8.4
6.6
5.95
3.97
21 – 28 Aug 2013
21 – 31 Oct 2015
01 – 14 Nov 2015
703.5
836.5
511.5
141
110
50
20.0
13.2
9.8
74,633
48,967
36,400
62%
41%
30%
8.7
25.8
1.0
1.08
-0.20
NAb
a
Number of rats extrapolated from the relationship between the abundance index and original population estimate and the mean (95%
credible interval) from the state-space model used to calculate r. See text for details.
b
Populations and population changes for October–November 2015 were not calculated because snap trapping removes individuals
from the population and biases abundance indices for short time periods.
170
Bond, et al.,: Rat recovery, Henderson Island
Fig. 2 Rat abundance (a), density (b), capture probability (c), and spatial shape parameters (c) from spatially-explicit
capture-recapture analysis of Pacific rats in two habitats on Henderson Island in 2015 (black circles: embayment forest,
open circles: plateau).
Rapid eradication assessment
Using the same trapping array as in 2015 for three
10-day trapping sessions at monthly intervals following
a future hypothetical eradication attempt on Henderson
Island would be insufficient to declare an eradication
successful with 95% certainty. When we simulated a larger
trapping array, we found that we could only conclude with
95% confidence that the eradication had been successful
if no rat was detected within three months on a trapping
grid covering at least 30% of Henderson Island. The model
indicated that in order to be confident that the eradication
had been successful within three months, we would require
a 30 × 30 m trapping grid covering one third of the island.
DISCUSSION
Population recovery after a failed eradication
The rat population on Henderson increased rapidly
for at least the first 15 months following the eradication
operation, with high annual per capita growth rates up
to August 2013 (Table 2). As the population approached
or exceeded pre-eradication abundances, the growth
rate decreased between August 2013 and October 2015,
possibly as a result of the population fluctuating around
a carrying capacity. These growth rates are broadly
similar to the maximum annual growth rates of other rat
species (Hone, et al., 2010), and are useful to estimate the
probability of success of an eradication during follow-up
monitoring (Russell, et al., 2017). Owing to variability
in trapping methods and locations, there is considerable
uncertainty about the exact size of the rat population;
however, our density estimates in the embayment forest in
2015 indicate that rat density in this habitat was very similar
to the pre-eradication density estimated in the same habitat
at 24–40 rats/ha in 2009 (Cuthbert, et al., 2012). Based on
the updated density estimates from the plateau in 2015, we
estimate the rat population on North Beach in 2015 to be
~150–300 rats, which is similar to the estimates from the
extrapolated relationship between density and abundance
indices (Table 2). On the plateau, however, density
fluctuated considerably throughout the year (Table 3), and
extrapolating to the 4,290 ha of plateau habitat resulted in
an estimate of ~3,300–26,000 rats, which is lower than the
48,000 estimate from the relationship with the abundance
index. We assumed the relationship between density (rats
ha-1) and the abundance index (rats 100 CTN-1) was linear
but, on the plateau, this is clearly not the case. Estimating
171
Island invasives: scaling up to meet the challenge. Ch 1C Rodents: Lessons
density, however, entails significant work over several
months, whereas an abundance index can be determined
fairly quickly, in a matter of days. Further work should
investigate factors that influence the relationship between
these two metrics.
There have been few studies on the recovery of Rattus
spp. following eradication attempts. In urban Baltimore,
Maryland, USA, R. norvegicus recovered to pre-control
numbers within about 12–18 months (Emlen Jr., et al., 1948),
and R. fuscipes in Australian eucalypt patches returned to
pre-removal densities within two years (Lindenmayer, et
al., 2005). In both cases, immigration was the likely cause
of the rapid increase (though see Banks, et al., 2011).
Genetic analysis from Henderson shows that there was no
reinvasion, and that all rats present are descended from 60–
80 survivors of the failed 2011 eradication operation (Amos,
et al., 2016). Our results demonstrate the rapid recovery of
an island population of introduced rodents in the absence
of immigration. The time for rodent populations to either
recover or reach pre-eradication levels (15–24 months),
was similar to the experimental invasion of Saddle Island,
New Zealand by mice (Mus musculus), where immigration
may have supplemented mouse populations (Nathan, et al.,
2015), and the time from arrival to near-saturation of black
rats (Rattus rattus) on Taukihepa, New Zealand (24–36
months; Bell, et al., 2016).
Temporal and spatial variation in rat population
density
The shape of the recovery curve of the rat population
on Henderson is difficult to determine from the intermittent
trapping efforts and due to the high short-term variability.
In 2015 we documented three-fold fluctuations in live
trapping indices and even larger differences in rat density
within just two months (Fig. 2), indicating that there may
be pronounced seasonal changes among the rat population
that could potentially mask or confound any long-term
trajectories. There may also be considerable spatial
heterogeneity in rat densities, and rats on Henderson do
travel large distances (Oppel, et al., 2019), which further
complicate interpretations from sampling a relatively
small area of the available habitat. Some rats are relatively
territorial, moving <200 m, and others roaming >1000 m
(Oppel, et al., 2019).
Our finding that a decrease in rat density coincided with
increasing movement rates of rats (Fig. 2) adds a further
complication to the long-term comparison of simple trap
indices that do not account for capture probability and
rat movements. Tropical rodent populations are known to
undergo large population fluctuations, which can be driven
by short-term changes in resource availability (Adler,
1998; Madsen & Shine, 1999) or extreme climatic events
(Ujvari, et al., 2016). We did not observe pronounced plant
resource fluctuations in 2015, and most tree species had
individuals at various stages of flowering and fruiting
between June and October 2015, though invertebrate
abundance likely varied through the season (Lavers, et al.,
2016). There was also neither a noticeable drought, nor an
unusually heavy rainfall event during that period that could
have explained the apparent intermittent reduction in the
rat population. In the beach embayment forest, the major
reduction of rat density between early and late August
coincided with the temporal availability of Murphy’s
petrel (Pterodroma ultima) chicks, which may have led
to temporary immigration of rats, but would have been
unable to sustain a rat population for more than a few days
(Brooke, et al., 2010b). Although we do not know whether
higher mortality, lower fecundity, or both contributed to
the apparent temporal fluctuation that we observed, or
whether rats’ probability of capture changed significantly
over time, the timing of any future eradication operation
should coincide with a naturally occurring nadir in the
population trajectory to improve the probability of success.
The eradication operation in 2011 therefore appears to
have been optimally timed if rat population fluctuations are
similar every year, but more research is required to examine
whether rat populations exhibit predictable seasonality on
sub-tropical islands such as Henderson.
The population abundance indices of R. exulans on
Henderson (14–32 rats/100 CTN with snap-trapping, 12–
75.8 rats/100 CTN with live-trapping) appear to be higher
than abundance indices of other island rat population. For
example, the R. exulans abundance index on Hawaii was
only 5.65 rats /100 CTN (Sugihara, 1997), presumably
because the species is subject to competition and predation
(Moller & Craig, 1987); on the Marianas, the trapping
rate was also much lower than on Henderson with 3.7
rats/100 CTN (Yackel Adams, et al., 2011). On Honuea,
French Polynesia, indices ranged from 5–20 rats/100 CTN,
often lower than conspecific R. rattus (up to 35 rats/100
CTN; Russell, et al., 2015). Abundance indices of the
much larger R. norvegicus ranged from 3–9 rats /100 CTN
(Drever, 2004; Harper, et al., 2005; Bond & Eggleston,
2015), and those of R. rattus from 1.6–35 rats/100 CTN
across their range (Blackwell, et al., 2002; Shiels, 2010;
Russell, et al., 2015), but reached up to 94.1/100 CTN
on some nearshore islands in New Zealand (Russell &
MacKay, 2005), and ranged from 60–80 rats/100 CTN on
Surprise Island, New Caledonia (Caut, et al., 2009). The
Table 3 Mean (± 95% confidence interval) rat density, capture probability, and movement parameter, σ, in
two habitats of Henderson Island in June–October 2015 estimated with spatially-explicit capture-recapture
models. Note that different detection functions were used in the embayment forest (half-normal) and on the
plateau (negative exponential), and that σ values are not directly comparable.
Habitat
Time period
Density (rats ha-1)
Capture probability
Beach embayment –
North Beach
Early August
Late August
September
Early June
Late June
July
Early August
Late August
September
October
42.92 (27.92–65.98)
20.37 (12.67–32.73)
27.2 (19.73–37.48)
6.08 (4.76–7.76)
2.33 (1.95–2.78)
1.77 (1.49–2.11)
0.76 (0.6–0.95)
0.94 (0.75–1.17)
3.73 (3.15–4.4)
4.36 (3.92–4.85)
0.13 (0.09–0.19)
0.18 (0.13–0.24)
0.08 (0.06–0.1)
0.13 (0.1–0.17)
0.47 (0.38–0.56)
0.53 (0.41–0.64)
0.54 (0.41–0.67)
0.17 (0.11–0.25)
0.07 (0.05–0.1)
0.43 (0.38–0.49)
Plateau
172
σ
12.48 (9.67–16.11)
19.35 (15.41–24.29)
22.08 (18.85–25.86)
18.06 (15.39–21.19)
24.6 (22.74–26.61)
29.67 (27.57–31.92)
33.64 (30.52–37.09)
38.78 (33.92–44.33)
34.05 (31.75–36.51)
28.92 (27.56–30.34)
Bond, et al.: Rat recovery, Henderson Island
habitable by rats (Samaniego-Herrera, et al., 2013; Russell,
et al., 2017) but because Henderson Island is a fairly
large island with impenetrable vegetation it is logistically
unrealistic to install a monitoring network across an area
sufficiently large to enable an early declaration of success.
Depending on where rats that survive an eradication attempt
occur in relation to the trap array at the northern end of the
island, the potential benefit of post-operational monitoring
to facilitate a rapid assessment of success is questionable.
By the time surviving rats may be discovered with the
limited trap array, the population would have likely grown
to a size that would require a new eradication rather than
allow a rapid follow-up to kill any remaining survivors.
Although post-operational rat trapping on Henderson
Island may be useful to rapidly discover eradication failure,
it is unlikely that it would allow the confident conclusion of
eradication success.
much higher trapping rate of R. exulans on Henderson is
possibly because the species is smaller than congeners and
is not subject to either predation or competition because no
other mammals or avian predators exist on Henderson, and
there is minimal dietary overlap with Henderson’s birds
(Brooke & Jones, 1995; Jones, et al., 1995; Trevelyan,
1995; Lavers, et al., 2016). In addition, the relatively high
temperature and greater resource availability on tropical
islands is generally well known to increase rat population
size compared to temperate islands (Harper & Bunbury,
2015; Russell & Holmes, 2015). It is important to note,
however, that abundance indices of rats exhibit a wide
range depending on the rat species, environment, and the
presence of competitor or predator species, seasonality,
trap type, and the layout and spacing of traps.
Different approaches used to estimate densities also
complicate the comparison across different islands (Harper
& Bunbury, 2015). Despite the relatively high snap- and
live-trapping rates on Henderson, our estimate of rat
density is surprisingly low, especially on the coral plateau,
where large rat movements were observed (Oppel, et al.,
2019) that may have led to high trapping indices despite
low density. But even the 10-times higher density in the
beach embayment forest appears to be at the lower end of
the range found for R. exulans on tropical islands (1.2–288
rats/ha; Harper & Bunbury, 2015). A potential explanation
for this apparent discrepancy might be that Henderson
Island is a relatively nutrient-poor coral atoll, where the
maximum population size could be lower compared to
more fertile tropical islands. Due to the potential nutrient
limitation, the use of a highly attractive bait (coconut) may
result in relatively high trapping rates, especially on the
plateau where coconut is generally unavailable. Coconut
has been implicated as an important factor affecting
the eradication success on tropical islands (Holmes, et
al., 2015). Our data also suggest that coconut may have
facilitated a rapid recovery of the surviving rat population:
In 2009, trapping rates were similar in the beach
embayment forest and on the plateau, but on all occasions
after the eradication attempt the snap-trapping rates on the
plateau were considerably lower than in the embayment
forest (Table 2). Rats on Henderson would often gnaw into
de-husked coconuts, or those opened by land crabs on the
beaches. We speculate that the lush embayment forest with
abundant coconut may have facilitated a faster return to
pre-eradication rat population densities than the scrubby
plateau forest where coconut, though present, is scarce
compared to the embayment forest.
Previous investigations using live- and snap-trapping
indicated that live traps do not have a higher capture
probability than snap traps for R. exulans and R. rattus
(Russell, et al., 2015). We observed a much higher trapping
rate with live traps than with snap traps in 2015, and our
spatially explicit capture-recapture model indicated that
rats living along our trail network had an almost 100%
probability of being captured at least once in a trap. The
variation in capture probability between different trap
types highlights the need for consistent monitoring using
identical approaches to facilitate valid comparisons over
time.
We thank the Government of the Pitcairn Islands
for permission to work on Henderson Island, M. de L.
Brooke, A. Brown, G. Harper, G. Harrison, S. O’Keefe,
M. Rodden, J. Warren, and P. Warren for assistance in
the field, and J. Hall, A. Schofield, and C. Stringer for
logistic support. The crews of the Braveheart, Claymore
II, Teba, and Xplore, provided transportation to and from
Henderson. The David and Lucile Packard Foundation,
UK Overseas Territories Environment Programme, Darwin
Plus: Overseas Territories Environment and Climate
Fund, British Birds, generous donors, and the RSPB, the
UK partner in Birdlife International, helped to fund our
research. We appreciate the advice of J. Russell and M.
Efford on data analysis. Scientific and ethical approval was
granted by the Government of the Pitcairn Islands, and the
RSPB Council (paper 2/13/62 and protocol EAC 2015/01).
Comments from two anonymous reviewers improved this
manuscript.
Assessing the probability of eradication success
REFERENCES
The failure of the 2011 eradication attempt was
opportunistically recorded eight months after the bait drop.
Our assessment indicated that an intensive monitoring
programme for three months following a bait drop would
not have allowed the reliable conclusion that the eradication
had been successful, unless at least 30% of the island were
covered with a 30 × 30 m trapping grid. Such a trapping
effort is unrealistic on Henderson Island. Rapid eradication
assessments have so far only been conducted for small
islands where the survey effort covered the entire area
Adler, G.H. (1998). ‘Impacts of resource abundance on populations of a
tropical forest rodent’. Ecology 79(1): 242–254.
Amos, W., Nichols, H.J., Churchyard, T. and Brooke, M.de L. (2016).
‘Rat eradication comes within a whisker! A case study from the South
Pacific’. Royal Society Open Science 3: 160110.
Banks, S.C., Dujardin, M., McBurney, L., Blair, D., Barker, M. and
Lindenmayer, D.B. (2011). ‘Starting points for small mammal population
recovery after wildfire: Recolonisation or residual populations?’ Oikos
120: 26–37.
Bell, E.A., Bell, B.D. and Merton, D.V. (2016). ‘The legacy of Big South
Cape: Rat irruption to rat eradication’. New Zealand Journal of Ecology
40(2): 212–218.
In summary, we found that rat abundance increased
rapidly between the failed eradication operation in August
2011 and August 2013 before decreasing from August
2013 to October 2015. Rats on Henderson Island reached
two-thirds of their pre-eradication abundance 24 months
following their failed eradication, but our estimates of
rat density on the plateau of the island suggest that rat
density may have been substantially lower than previously
assumed (Brooke, et al., 2010a). Studies of failed
eradication operations, and the recovery of introduced
rodent populations are rare, but of great conservation
and operational importance. Our study highlights rodent
population fluctuations on a relatively short timescale, and a better understanding of the regularity and
underlying drivers of these fluctuations would be useful
to schedule an eradication operation so as to maximise
eradication success. As eradication operations on tropical
and subtropical islands become more frequent, it will be
increasingly important to understand the behaviour and
ecology of the invasive species targeted, and more work
in this area is required if we are to replicate success on
temperate islands.
ACKNOWLEDGEMENTS
173
Island invasives: scaling up to meet the challenge. Ch 1C Rodents: Lessons
Bellingham, P.J., Towns, D.R., Cameron, E.K., Davis, J.J., Wardle, D.A.,
Wilmshurst, J.M. and Mulder, C.P.H. (2010). ‘New Zealand island
restoration: Seabirds, predators, and the importance of history’. New
Zealand Journal of Ecology 34(1): 115–136.
Blackwell, G.L., Potter, M.A. and McLennan, J.A. (2002). ‘Rodent
density indices from tracking tunnels, snap-traps and Fenn traps: Do
they tell the same story?’ New Zealand Journal of Ecology 26(1): 43–51.
Bond, A.L. and Eggleston, C.J. (2015). ‘Application of a non-invasive
indexing method for introduced Norway rats (Rattus norvegicus) in the
Aleutian Islands, Alaska’. Biodiversity and Conservation 24: 2551–
2563.
Borchers, D.L. and Efford, M.G. (2008). ‘Spatially explicity maximum
likelihood methods for capture-recapture studies’. Biometrics 64: 377–
385.
Borchers, D. (2012). ‘A non-technical overview of spatially explicit
capture–recapture models’. Journal of Ornithology 152(2): 435–444.
Brooke, M. de L. and Jones, P.J. (1995) ‘The diet of the Henderson fruitdove Ptilinopus insularis. Island Field observations of fruit choice’.
Biological Journal of the Linnean Society 56: 149–165.
Brooke, M.de L., Cuthbert, R.J., Henricson, A., Torr, N., Warren, P. and
O’Keefe, S. (2010a). Towards Rat Eradication on Henderson Island:
Fieldwork Report, August – September 2009. Unpublished report.
Sandy, UK: Royal Society for the Protection of Birds.
Brooke, M.de L., O’Connell, T.C., Wingate, D., Madeiros, J., Hilton,
G.M. and Ratcliffe, N. (2010b). ‘Potential for rat predation to cause
decline of the globally threatened Henderson petrel Pterodroma atrata:
Evidence from the field, stable isotopes and population modelling’.
Endangered Species Research 11: 47–59.
Brooke, M. de L. (2019). ‘Rat Eradication in the Pitcairn Islands, South
Pacific: a 25-year Perspective’. In: C.R. Veitch, M.N. Clout, A.R.
Martin, J.C. Russell and C.J. West (eds.) Island invasives: scaling up
to meet the challenge, pp. 95–99. Occasional Paper SSC no. 62. Gland,
Switzerland: IUCN.
Cuthbert, R.J., Brooke, M.de L. and Torr, N. (2012). ‘Overcoming hermitcrab interference during rodent-baiting operations: A case study from
Henderson Island, South Pacific’. Wildlife Research 39(1): 70–77.
DIISE (2016). ‘The Database of Island Invasive Species Eradications,
developed by Island Conservation, Coastal Conservation Action
Laboratory UCSC, IUCN SSC Invasive Species Specialist Group.
University of Auckland and Landcare Research New Zealand’. <http://
diise.islandconservation.org>. Accessed 1 July 2016.
Drever, M.C. (2004). ‘Capture rates of Norway rats (Rattus norvegicus)
in a seabird colony: A caveat for investigators’. Northwestern Naturalist
85: 111–117.
Efford, M. (2004). ‘Density estimation in live-trapping studies’. Oikos
106(3): 598–610.
Efford, M.G. (2016). ‘secr: Spatially Explicit Capture-recapture in R. R
Package Version 2.10-3’. <https://CRAN.R-project.org/package=secr>.
Emlen Jr., J.T., Stokes, A.W. and Winsor, C.P. (1948). ‘The rate of
recovery of decimated populations of brown rats in nature’. Ecology
29(2): 133–145.
Harper, G.A., Dickinson, K.J.M. and Seddon, P.J. (2005). ‘Habitat use by
three rat species (Rattus spp.) on Stewart Island/Rakiura, New Zealand’.
New Zealand Journal of Ecology 29(2): 251–260.
Harper, G.A. and Bunbury, N. (2015). ‘Invasive rats on tropical islands:
Their population biology and impacts on native species’. Global
Ecology and Conservation 3: 607–627.
Harper, G.A., van Dinther, M., Russell, J.C. and Bunbury, N. (2015).
‘The response of black rats (Rattus rattus) to evergreen and seasonally
arid habitats: Informing eradication planning on a tropical island’.
Biological Conservation 185: 66–74.
Hein, S. and Jacob, J. (2015). ‘Recovery of small rodent populations after
population collapse’. Wildlife Research 42(2): 108–118.
Holmes, N.D., Griffiths, R., Pott, M., Alifano, A., Will, D., Wegmann, A.S.
and Russell, J.C. (2015). ‘Factors associated with rodent eradication
failure’. Biological Conservation 185: 8-–16.
Hone, J., Duncan, R.P. and Forsyth, D.M. (2010). ‘Estimates of maximum
annual population growth rates (rm) of mammals and their application
in wildlife management’. Journal of Applied Ecology 47(3): 507–514.
Howald, G.R., Donlan, C.J., Galván, J.P., Russell, J.C., Parkes, J.,
Samaniego, A., Wang, Y., Veitch, D., Genovesi, P., Pascal, M., Saunders,
A. and Tershy, B. (2007). ‘Invasive rodent eradication on islands’.
Conservation Biology 21(5): 1258–1268.
Jones, H.P., Holmes, N.D., Butchart, S.H.M., Tershy, B.R., Kappes,
P.J., Corkery, I., Aguirre-Muñoz, A., Armstrong, D.P., Bonnaud, E.,
Burbidge, A.A., Campbell, K., Courchamp, F., Cowan, P.E., Cuthbert,
R.J., Ebbert, S.E., Genovesi, P., Howald, G.R., Keitt, B.S., Kress, S.W.,
Miskelly, C.M., Oppel, S., Poncet, S., Rauzon, M.J., Rocamora, G.,
Russell, J.C., Samaniego-Herrera, A., Seddon, P.J., Spatz, D.R., Towns,
D.R. & Croll, D.A. (2016). ‘Invasive mammal eradication on islands
results in substantial conservation gains’. Proceedings of the National
Academy of Sciences of the United States of America 113: 4033–4038.
Jones, P.J., Schubel, S., Jolly, J.N., Brooke, M. de L., and Vickery, J.A.
(1995) ‘Behaviour, natural history, and annual cycle of the Henderson
Island rail Porzana atra (Aves: Rallidae)’. Biological Journal of the
Linnean Society 56: 167–183.
Keitt, B.S., Griffiths, R., Boudjelas, S., Broome, K.G., Cranwell,
S., Millett, J., Pitt, W. and Samaniego-Herrera, A. (2015). ‘Best
practice guidelines for rat eradication on tropical islands’. Biological
Conservation 182: 17–26.
174
Lavers, J.L., McClelland, G.T.W., MacKinnon, L., Bond, A.L., Oppel, S.,
Donaldson, A.H., Duffield, N.D., Forrest, A.K., Havery, S.J., O’Keefe,
S., Skinner, A., Torr, N. and Warren, P. (2016). Henderson Island
Expedition Report: May–November 2015. RSPB Research Report 57.
Sandy, UK: RSPB Centre for Conservation Science.
Lindenmayer, D.B., Cunningham, R.B. and Peakall, R. (2005). ‘The
recovery of populations of bush rat Rattus fuscipes in forest fragments
following major population reduction’. Journal of Applied Ecology 42:
649–658.
Lorvelec, O. and Pascal, M. (2005). ‘French attempts to eradicate
non-indigenous mammals and their consequences for native biota’.
Biological Invasions 7: 135–140.
Madsen, T. and Shine, R. (1999). ‘Rainfall and rats: Climatically-driven
dynamics of a tropical rodent population’. Australian Journal of
Ecology 24(1): 80–89.
Moller, H. and Craig, J.L. (1987). ‘The population ecology of Rattus
exulans on Tiritiri Matangi Island, and a model of comparative
population dynamics in New Zealand’. New Zealand Journal of
Zoology 14(3): 305–328.
Nathan, H.W., Clout, M.N., MacKay, J.W.B., Murphy, E.C. and Russell,
J.C. (2015). ‘Experimental island invasion of house mice’. Population
Ecology 57: 363–371.
Nelson, L. and Clark, F.W. (1973). ‘Correction for sprung traps in catch/
effort calculations of trapping results’. Journal of Mammalogy 54(1):
295–298.
Oppel, S., Bond, A.L., Brooke, M.de L., Harrison, G., Vickery, J.A.
and Cuthbert, R.J. (2016). ‘Temporary captive population and rapid
population recovery of an endemic flightless rail after a rodent
eradication operation using aerially distributed poison bait’. Biological
Conservation 204B: 442–448.
Oppel, S., McClelland, G.T.W., Lavers, J.L., Churchyard, T., Donaldson,
A.H., Duffield, N.D., Havery, S., Kelly, J., Proud, T., Russell, J., Vickery,
J.A. and Bond, A.L. (2019). ‘Seasonal variation in movements and
survival of invasive Pacific rats on a sub-tropical island: implications
for eradication’. In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell
and C.J. West (eds.) Island invasives: scaling up to meet the challenge,
pp. 200–208. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
R Core Team (2017). R: A Language and Environment for Statistical
Computing. Version 3.4.0.Vienna, Austria: R Foundation for Statistical
Computing.
Ringler, D., Russell, J., Jaeger, A., Pinet, P., Bastien, M. and Le Corre, M.
(2014). ‘Invasive rat space use on tropical islands: Implications for bait
broadcast’. Basic and Applied Ecology 15(2): 179–186.
Russell, J. and MacKay, J. (2005). Ship Rat Reinvasion of Tawhitinui
Island. Unpublished report. Auckland, NZ: University of Auckland.
Russell, J.C., Ringler, D., Trombini, A. and Le Corre, M. (2011).
‘The island syndrome and population dynamics of introduced rats’.
Oecologia 167: 667–676.
Russell, J.C., Caut, S., Anderson, S.H. and Lee, M. (2015). ‘Invasive rat
interactions and over-invasion on a coral atoll’. Biological Conservation
185: 59–65.
Russell, J.C. and Holmes, N.D. (2015). ‘Tropical island conservation: Rat
eradication for species recovery’. Biological Conservation 185: 1–7.
Russell, J.C., Binnie, H.R., Oh, J., Anderson, D.P. and SamaniegoHerrera, A. (2017). ‘Optimizing confirmation of invasive species
eradication with rapid eradication assessment’. Journal of Applied
Ecology 54: 160–169.
Samaniego-Herrera, A., Anderson, D.P., Parkes, J.P. and Aguirre-Muñoz,
A. (2013). ‘Rapid assessment of rat eradication after aerial baiting’.
Journal of Applied Ecology 50: 1415–1421.
Shiels, A.B. (2010). ‘Ecology and impacts of introduced rodents (Rattus
spp. and Mus musculus) in the Hawaiian Islands’. Ph.D. thesis. Manoa:
University of Hawai’i.
Spencer, T. (1995). ‘The Pitcairn Islands, South Pacific Ocean: Plate
tectonic and climatic contexts’. Biological Journal of the Linnean
Society 56: 13–42.
Sugihara, R.T. (1997). ‘Abundance and diets of rats in two native
Hawaiian forests’. Pacific Science 51(2): 189–198.
Torr, N. and Brown, D. (2012). ‘Henderson Island Restoration Project
Post-operational Report. Unpublished report. Sandy, UK: Royal Society
for the Protection of Birds.
Trevelyan, R.J. (1995) ‘The feeding ecology of Stephen’s lory and nectar
availability in its food plants’. Biological Journal of the Linnean Society
56: 185–197.
Ujvari, B., Brown, G., Shine, R. and Madsen, T. (2016). ‘Floods
and famine: Climate-induced collapse of a tropical predator-prey
community’. Functional Ecology 30(3): 453–458.
Varnham, K. (2010). Invasive Rats on Tropical Islands: Their History,
Ecology, Impacts and Eradication. RSPB Research Report No. 41.
Sandy, UK: Royal Society for the Protection of Birds.
Waldren, S., Florence, J. and Chepstow-Lusty, A.J. (1995). ‘A comparison
of the vegetation communities from the islands of the Pitcairn Group’.
Biological Journal of the Linnean Society 56: 121–144.
Weigelt, P., Jetz, W. and Kreft, H. (2013). ‘Bioclimatic and physical
characterization of the world’s islands’. Proceedings of the National
Academy of Sciences of the United States of America 110(38): 15307–
15312.
Yackel Adams, A.A., Stanford, J.W., Wiewel, A.S. and Rodda, G.H.
(2011). ‘Modelling detectability of kiore (Rattus exulans) on Aguiguan,
Mariana Islands, to inform possible eradication and monitoring efforts’.
New Zealand Journal of Ecology 35(2): 145–152.
A.L. Bond, S. O’Keefe, P. Warren and G.T.W. McClelland
Bond, A.L.; S. O’Keefe, P. Warren and G.T.W. McClelland. Bait colour and moisture do not affect bait
acceptance by introduced Pacific rats (Rattus exulans) at Henderson Island, Pitcairn Islands
Bait colour and moisture do not affect bait acceptance by introduced
Pacific rats (Rattus exulans) at Henderson Island, Pitcairn Islands
A.L. Bond1,3, S. O’Keefe2, P. Warren2 and G.T.W. McClelland1
RSPB Centre for Conservation Science, Royal Society for the Protection of Birds, The Lodge, Sandy, Bedfordshire,
SG19 2DL United Kingdom. <a.bond@nhm.ac.uk>. 2Department of Environment, Conservation, and Natural
Resources, Government of the Pitcairn Islands, Adamstown, Pitcairn Islands, PCRN 1ZZ South Pacific Ocean. 3Current
address: Bird Group, Department of Life Sciences, The Natural History Museum, Akeman Street, Tring, Hertfordshire,
HP23 6AP United Kingdom.
1
Abstract Rodent eradications are a useful tool for the restoration of native biodiversity on islands, but occasionally these
operations incur non-target mortality. Changes in cereal bait colour could potentially mitigate these impacts but must not
compromise the eradication operation. Changing bait colour may reduce mortality of Henderson crakes (Zapornia atra),
an endemic globally threatened flightless bird on Henderson Island, Pitcairn Islands, South Pacific Ocean. Crakes had
high non-target mortality in a failed 2011 rat eradication operation and consumed fewer blue than green cereal pellets.
We examined which cereal bait properties influenced its acceptance by captive Pacific rats (Rattus exulans) on Henderson
Island. We held 82 Pacific rats from Henderson Island in captivity and provided them with non-toxic cereal bait pellets
of varying properties (blue or green, moist or dry). We estimated the proportion of rats consuming bait using logistic
generalised linear mixed models. We found no effect of sex, females’ reproductive status, bait colour or bait moisture on
rats’ willingness to consume baits. Rats’ bait consumption was unaffected by cereal bait properties (colour or moisture).
The use of blue bait is unlikely to affect future eradication operational success but may reduce non-target mortality of
Henderson crakes. Timing cereal bait distribution in relation to precipitation may also reduce crake mortality without
compromising palatability to rats.
Keywords: baits, Henderson Island, island restoration, non-target safety, rat eradication
INTRODUCTION
The eradication of introduced rodents is a common
conservation intervention, especially on islands, and has
been accomplished on > 580 islands worldwide (DIISE,
2016), with benefits to native biodiversity (Lavers, et al.,
2010; Buxton, et al., 2014; Jones, et al., 2016). In some
cases, eradication operations may result in non-target
morality, and mitigation can include housing captive
populations of species likely to be affected, or by using
cereal pellets that are less palatable to non-target species
(Empson & Miskelly, 1999; Hoare & Hare, 2006; Pitt, et
al., 2015; Oppel, et al., 2016a; Oppel, et al., 2016b). In
such cases, it is crucial that the mitigation measures do not
decrease the likelihood of a successful rodent eradication
operation, and that the rodents are exposed to a sufficient
quantity of cereal bait, are susceptible to the rodenticide
used, and will consume a sufficient dose.
Rarely, though more so in the tropics and subtropics,
these eradication operations fail to remove rodents for
a variety of operational, biological, and environmental
reasons (Holmes, et al., 2015). Eradication success in
the tropics is generally lower than in temperate systems
because there is less seasonal change in the environment,
and consequently a less predictable period of food-resource
limitation, which is the ideal time for an eradication
operation (Holmes, et al., 2015; Russell & Holmes, 2015).
Consequently, understanding which factors may influence
rodents’ acquisition of a lethal dose of bait are crucial for
improving the probability of success in tropical systems
(Lamoreux, et al., 2006).
Factors affecting bait acceptance by Pacific rats (Rattus
exulans), a common introduced rodent in the Pacific
Ocean tropics and subtropics (Atkinson, 1985; Varnham,
2010; Keitt, et al., 2015), are poorly known. A number of
factors can influence rat food choice, including physical
characteristics such as bait colour and hardness (Booth, et
al., 1974; Clapperton, 2006; Hegab, et al., 2014). Murine
rodents (including rats and mice) have colour vision,
including sensitivity in the UV range (Jacobs, 1993;
Jacobs, 2009), and there is evidence that cereal bait colour
does affect the likelihood of acceptance by rats (Hegab, et
al., 2014).
Blue or green cereal pellets are the most effective
at reducing avian non-target mortalities, but there is
considerable variation in bait attraction among species.
Kea (Nestor notabilis) and weka (Gallirallus australis)
were less likely to eat green pellets than blue (Hartley,
et al., 2000; Weser & Ross, 2013), whereas North
Island robins (Petroica longipes) and Henderson crakes
(Zapornia atra) were less likely to consume blue pellets
than green (Hartley, et al., 1999; Oppel, et al., 2016b).
Henderson Island, part of the Pitcairn Islands in the South
Pacific Ocean, was the site of a failed eradication operation
for Pacific rat in 2011, which also resulted in non-target
mortality of Henderson crakes (Amos, et al., 2016; Oppel,
et al., 2016a). Subsequent work found that Henderson
crakes consumed less blue bait than green, and did not
consume dry pellets (Oppel, et al., 2016b).
Here we report on the factors affecting Pacific rat bait
acceptance on Henderson Island, Pitcairn Islands, South
Pacific Ocean. Our goal was to compare Pacific rats’
acceptance of both moist and dry cereal pellets of these two
colours to determine whether measures to reduce the nontarget mortality of Henderson crakes might affect success
of future eradication operations on Henderson.
METHODS
Rat capture & acclimatisation
Rats were captured on Henderson Island’s plateau using
either Tomahawk (27 × 16 × 13 cm, Metal Rat Cage Trap,
Key Industries, Auckland, New Zealand) or Sherman (22.9
× 8.9 × 7.6 cm, Sherman Traps Inc., Tallahassee, Florida,
USA) live traps baited with a 2 × 2 cm piece of coconut
(Cocos nucifera). Individual rats’ knowledge of coconut
prior to its presentation during the cage trial is presumed
to be limited, because the areas where rats were captured
were > 500 m away from the nearest coconut grove on
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 175–179. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
175
Island invasives: scaling up to meet the challenge. Ch 1C Rodents: Lessons
the island. We chose coconut as bait because it was easily
available and highly effective in attracting rats, while
alternative baits (Pandanus tectorius fruit, peanut butter,
chocolate, semolina, soap, and mixtures thereof) largely
failed to capture sufficient individuals. Although the choice
of coconut as trap bait may have predisposed some trapped
rats to exhibit universal acceptance of coconut in the trials
(see Results), particularly given the large movements
possible in this population (Oppel, et al., 2019), the use of
a non-natural food source (e.g., peanut butter, chocolate)
may have resulted in capturing only curious or bold rats
who will readily accept new food items, which may have
biased our assessment of acceptance rate of bait pellets,
another novel food item (Booth, et al., 1974).
We held 81 rats in captivity for 12 days each during
October–November 2015. Rats were weighed using a
spring balance to the nearest 1 g, fitted with a uniquely
numbered ear tag, and their sex was determined from
external anatomy. They were first allowed to acclimatise
for four days in sex-specific communal wire cages (70 ×
60 × 30 cm) of up to four individuals, where they were
fed commercial rodent food (Rabbit and Guinea Pig
Muesli, Topflite, Oamaru, New Zealand) ad libitum in a
single ceramic food dish and water was provided by both
a commercial water dispenser (Critter Canteen, SuperPet,
Walnut Creek, California, USA) and a large clamshell.
Each cage contained four hollow Pandanus tectorius logs,
with small, loose pieces of coconut bark providing cover
and visual barriers. After four days, rats were weighed
as before, and moved to individual wire cages (70 ×
50 × 30 cm) with the same food and water regime, and
environmental enrichment. Rats were considered to have
acclimated if, after three days in individual cages, their
mass differed by < 10% compared to their mass at capture,
and we observed no anomalous behaviour. Rats that had
lost > 10% of body mass were allowed two additional
days to acclimate and were then reassessed using the same
criteria. Any individuals that did not acclimate were not
subjected to the experimental trial and were euthanised by
cervical dislocation and used in other research (Lavers, et
al., 2016).
of trial rats consuming bait. This procedure was continued
for two more days if needed, with rats receiving 10 and
20 pellets on Days 3 and 4, respectively. Rats that had not
eaten bait after Day 4 were given five bait pellets and all
natural foods, with the exception of coconut (which was
accepted universally), on experimental Day 5. The trial
ended the next day regardless of outcome.
Food consumption was monitored daily and all
remaining food from the previous night removed and
the cage inspected to ensure no natural food item was
completely consumed, and any remaining bait pellets
counted. The remains of any partially eaten pellets were
inspected and the amount eaten estimated to the nearest
25%. Where a natural food item was completely consumed,
the result was ignored and the test repeated with the same
individual.
All captive rats were humanely euthanised by cervical
dislocation at the end of the trial. Females were examined
internally to determine reproductive status: breeding
was indicated by the presence of foetal pups, a highly
vascularised uterus, or lactation.
Statistical analysis
We used logistic generalised linear mixed-effects
models (GLMMs) with a logit link to test whether bait
acceptance (yes/no) varied as a function of the following
fixed factors: sex (female/male), bait colour (blue/green),
and bait moisture (moist/dry), and trial day. We treated
‘individual’ as a random effect to account for potential
serial autocorrelation (Bolker, et al., 2009). We included
main effects only, as the effective sample size would reduce
the statistical power to detect the effect of interactions in
our dataset. We constructed a series of models with varying
biologically meaningful combinations of the terms above,
as well as an intercept-only model (Table 1) and evaluated
Table 1 The ranked set of candidate models for examining
captive Pacific rats’ acceptance of bait on Henderson
Island. Models with ∆AICc < 2 were considered
competitive (i.e. the top 3).
Bait acceptance trial
Model
k
Individuals which acclimated were assigned randomly
to one of four treatment groups based on combinations of
bait colour (green or blue) and moisture (moist or dry);
cereal bait pellets (nontoxic ~2 g Pestoff 20R, Animal
Control Products, Whanganui, New Zealand) were
otherwise identical. The pellets were surface coated with
dye and did not lose their colour after soaking in water.
The green dye was a composite of tartrazine powder (with
the Chemical Abstracts Service (CAS) number 1934-210), Brilliant Blue powder (CAS number 3844-45-9) and
sodium sulphate (CAS number 7757-82-6), and the blue
dye was Hexacol Indigo Carmine Supra Blue R2613 (CAS
number 860-22-0). Cereal pellets were presented to rats
either dry, as manufactured, or moist, after being soaked in
water for three hours. Rats received fresh cereal bait pellets
and natural foods daily.
AICc
Intercept only
Moisture
Sex
Colour
Sex + Moisture
Moisture + Colour
Sex + Colour
Day
Sex + Colour + Moisture
Day + Moisture
Day + Sex
Day + Colour
Day + Sex + Moisture
Day + Moisture + Colour
Day + Sex + Colour
Day + Sex +
Colour+Moisture
2
3
3
3
4
4
4
5
5
6
6
6
7
7
7
8
Between 06:00-08:00 (UTC-8) on Day 1 of the trial,
rats were provided with sufficient natural food (coconut,
Myrsine hosakae seeds, and whole Pandanus tectorius
fruit) to ensure that they were sated but not consuming any
single food item completely, and a single cereal bait pellet.
Food consumption was assessed between 06:00-08:00 on
Day 2: if the rat had consumed a significant portion of a
pellet (~1 g), the individual’s trial was ended and the rat
was euthanised. If the bait was not eaten, the rat received a
fresh batch of natural food of the same amount as on Day
1 while the number of pellets was increased to five in an
attempt to overcome neophobia and increase the number
176
wi
214.80
215.89
216.62
216.85
217.72
217.96
218.68
219.15
219.80
220.54
221.11
221.30
222.49
222.66
223.23
ΔAICc
0.00
1.10
1.82
2.05
2.92
3.16
3.89
4.35
5.00
5.75
6.31
6.50
7.69
7.87
8.43
0.33
0.19
0.13
0.12
0.08
0.07
0.05
0.03
0.03
0.02
0.01
0.01
0.01
0.01
<0.01
224.62
9.82
<0.01
k: number of parameters, AICc: Akaike’s Information Criterion
adjusted for small sample size, ∆AICc: difference between each
model and the most parsimonious model, wi: Akaike model
weight.
Bond, et al.: Bait colour and moisture
them in a multi-model selection framework using Akaike’s
Information Criterion adjusted for small sample size
(AICc) (Burnham & Anderson 2002). Models with ΔAICc
< 2 were considered competitive. All models were fit using
Laplace approximation in the package lme4 (Bates, et al.,
2014) in R 3.3.0 (R Core Team, 2017), and we present
mean parameter estimates (β) ± standard error.
RESULTS
We captured 82 rats of which 81 acclimated to
the captive trial. Overall, 48% of captive rats (n = 39)
consumed the non-toxic cereal pellets. The intercept-only
model, where bait consumption varied among individuals,
but not with any other factors, received the most support,
but models that included the single terms for sex, bait
colour, and bait moisture had ΔAICc < 2.0 (Table 1). Using
each of these single-factor models, there was no difference
in bait acceptance between sexes (males: β = 0.532 ±
0.012, females: β = 0.536 ± 0.014), and no effect of bait
colour (blue: β = 0.533 ± 0.018, green: β = 0.534 ± 0.018),
or moisture (dry: β = 0.540 ± 0.015, moist: β = 0.529 ±
0.012; Fig. 1). Bait acceptance did not differ with females’
reproductive status (calculated parameter estimates for
breeding: β = 0.539 ± 0.033, n = 17; not breeding: β =
0.543 ± 0.025, n = 22; one individual not of breeding age:
β = 0.520). All models that included trial day had ΔAICc >
4, so were not considered further (Table 1).
DISCUSSION
We found no evidence for bait colour, moisture, sex,
or reproductive status affecting the consumption of bait
pellets by captive Pacific rats. Blue and green bait pellets
are frequently used in rodent eradication operations
(Clapperton, et al., 2015), and the use of blue pellets may
therefore reduce the reported non-target mortality among
Henderson crakes (Oppel, et al., 2016b) without affecting
the efficacy of rat eradication operations.
Rats ingested dry and moist pellets equally, which
is important operationally as Henderson crakes do not
consume dry pellets (Oppel, et al., 2016b). While rainfall
patterns on Henderson are unpredictable and aseasonal
(Spencer, 1995), targeting any future eradication operation
at periods of low rainfall is unlikely to affect the outcome
for rodents, but may reduce the risk of non-target mortality.
In the longer term (i.e. longer than the four days used in our
captive trial), rainfall will increase the degradation rates of
bait, thereby reducing rats’ exposure to bait, regardless of
their inherent preference to consume bait that is dry or wet
(Berentsen, et al., 2014).
The aseasonal breeding often found on tropical islands
means that baiting operations are more likely to include
breeding females than operations in temperate regions.
Concern has been expressed that pregnant and lactating
rodents are less likely to eat bait if their nutritional needs
are not met by the bait matrix (Keitt, et al., 2015), though
this also assumes that rats could identify the nutritional
content of bait pellets without consuming a lethal dose
(i.e., one pellet; Amos, et al., 2016). Our results suggest
that not only are female Pacific rats as likely to consume
bait pellets as males, but that, at least for this particular
bait formulation, females’ reproductive status is unlikely to
influence bait acceptance.
Importantly, while only 48% of trial rats consumed
bait pellets, this general acceptance rate cannot be used to
infer potential acceptance rates in free-ranging rats during
an eradication operation. Evidence of higher or lower bait
acceptance rates in the wild than in cage trials is equivocal
(Clapperton, 2006) but several important limitations
Fig. 1 There was no effect of rats’ sex, breeding status, bait colour, or bait moisture on the proportion of Pacific rats
consuming bait. Values are from generalised linear mixed models. Dark bars are the median, boxes are the interquartile
range, and whiskers the range.
177
Island invasives: scaling up to meet the challenge. Ch 1C Rodents: Lessons
of cage trials have the potential to lower acceptance
below what would be typical in the wild. For example,
the provisioning of commercial rodent food during the
acclimatisation phase ensures an unbiased test of food
preference because test subjects are not food stressed
(which results in selection of food items based on dietary
deficiencies rather than food palatability). However, the
chemical composition of food plays an important role in
diet selection in free-ranging rats, with individuals selfselecting food based on physiological need (Rozin, 1976).
Another major limitation of cage trials, specifically with
regard to acceptance of novel food items such as bait, is the
absence of social learning. While rats are predominantly
solitary foragers, the identification and adoption of novel
food items is heavily influenced by social interactions with
conspecifics (Galef Jr, 1996). Cage trials should therefore
only be considered as useful tools for identifying potential
problems that can be explored further by field trials. On
Palmyra Atoll, for example, rats were found to prefer
coconut over cereal bait pellets in cage trials, but later field
trials found adequate bait uptake, and a toxic cereal bait
eradication was successful (Buckelew, et al., 2006; Alifano
& Wegmann, 2011).
Burnham, K.P. and Anderson, D.R. (2002). Model Selection and
Multimodel Inference: A Practical Information-theoretic Approach,
New York: Springer.
CONCLUSIONS
DIISE (2016). ‘The Database of Island Invasive Species Eradications’.
Island Conservation, Coastal Conservation Action Laboratory UCSC,
IUCN SSC Invasive Species Specialist Group. University of Auckland
and Landcare Research New Zealand. <http://diise.islandconservation.
org>. Accessed 1 July 2016.
Bolker, B.M., Brooks, M.E., Clark, C.J., Geange, S.W., Poulsen, J.R.,
Stevens, M.H.H. and White, J.-S.S. (2009). ‘Generalized linear mixed
models: a practical guide for ecology and evolution’. Trends in Ecology
and Evolution 24: 127–135.
Booth, D.A., Stoloff, R. and Nicholls, J. (1974). ‘Dietary flavor
acceptance in infant rats established by association with effects of
nutrient composition’. Physiological Psychology 2(3A): 313–319.
Buckelew, S.L., Howald, G.R., Wegmann, A.S., Sheppard, J., Curl, J.,
McClelland, P., Tershy, B., Swift, K., Campbell, E. and Flint, B. (2006).
Progress in Palmyra Atoll Restoration: Rat Eradication Trial 2005.
Santa Cruz, California, USA: Island Conservation.
Buxton, R.T., Jones, C.J., Moller, H. and Towns, D.R. (2014). ‘Drivers
of seabird population recovery on New Zealand islands after predator
eradication’. Conservation Biology 28(2): 333–344.
Clapperton, B.K. (2006). ‘A Review of the Current Knowledge of Rodent
Behaviour in Relation to Control Devices’. Science for Conservation
263. Wellington, NZ: Department of Conservation.
Clapperton, B.K., Day, T.D., Morgan, D.K., Huddart, F., Cox, N. and
Matthews, L.R. (2015). ‘Palatability and efficacy to possums and rats
of pest control baits containing bird repellents’. New Zealand Journal
of Zoology 42: 104–118.
Our findings suggest Pacific rats have no preference
between green or blue bait pellets, nor if bait is moist or
dry. This suggests that individual variation is a significant
driver of bait acceptance, regardless of other demographic
parameters (Nathan, 2016). While a baiting operation
timed when rats are breeding carries increased risks and
is preferably avoided, pregnant or lactating females are as
likely to accept bait as non-pregnant females. Any future
operation on Henderson Island should use blue bait pellets,
and time the operation for dry conditions, in order to reduce
non-target mortality of Henderson crakes without affecting
rat bait acceptance.
Galef Jr, B.G. (1996). ‘Social enhancement of food preferences in Norway
rats: a brief review’. In: C.M. Heyes and B.G. Galef Jr (eds.) Social
Learning in Animals: The Roots of Culture, pp. 49–64. Amsterdam:
Elsevier.
ACKNOWLEDGEMENTS
Hartley, L., Waas, J., O’Connor, C. and Matthews, L. (2000). ‘Colour
preferences and coloured bait consumption by weka Gallirallus
australis, an endemic New Zealand rail’. Biological Conservation
93(2): 255–263.
Empson, R.A. and Miskelly, C.M. (1999). ‘The risks, cost and benefits of
using brodifacoum to eradicate rats from Kapiti Island, New Zealand’.
New Zealand Journal of Ecology 23(2): 241–254.
Hartley, L., O’Connor, C., Waas, J. and Matthews, L. (1999). ‘Colour
preferences in North Island robins (Petroica australis): Implications for
deterring birds from poisonous baits’. New Zealand Journal of Ecology
23(2): 255–259.
The authors declare no conflicts of interest. We thank
N. Duffield, S. Havery, and A. Skinner, for assistance
in the field, J. Kelly, A. Schofield, C. Stringer, and J.
Vickery for logistical support, and the Government of the
Pitcairn Islands for permission to work on Henderson.
Methods were approved by the RSPB Ethical Advisory
Committee (protocol EAC 2015/01). The David and Lucile
Packard Foundation, Darwin Plus: Overseas Territories
Environment and Climate Fund, British Birds and generous
donors helped to fund this expedition. Comments from
J. Lavers, S. Oppel, I. Robinson, A. Schofield, and three
anonymous reviewers improved earlier drafts.
Hoare, J.M. and Hare, K.M. (2006). ‘The impact of brodifacoum on nontarget wildlife: gaps in knowledge’. New Zealand Journal of Ecology
30(2): 157–167.
REFERENCES
Jacobs, G.H. (2009). ‘Evolution of colour vision in mammals’.
Philosophical Transactions of the Royal Society B: Biological Sciences
364(1531): 2957–2967.
Alifano, A. and Wegmann, A.S. (2011). The Ecotoxicology and
Palatability of Two Rodenticide Bait Products: Field-based Assessment
at Palmyra Atoll. Santa Cruz, California: Island Conservation.
Amos, W., Nichols, H.J., Churchyard, T. and Brooke, M.de L. (2016).
‘Rat eradication comes within a whisker! A case study from the South
Pacific’. Royal Society Open Science 3: 160110.
Atkinson, I.A.E. (1985). ‘The Spread of Commensal Species of Rattus
to Oceanic Islands and their Effects on Island Avifaunas’. In: P.J.
Moors (ed.) Conservation of Island Birds. pp. 35–81. Cambridge: ICBP
Technical Publication No. 3.
Bates, D., Machler, M., Bolker, B. and Walker, S. (2014). ‘Fitting
Linear Mixed-Effects Models using lme4’. <arXiv:http://arxiv.org/
abs/1406.5823>.
Berentsen, A.R., Pitt, W.C., Eisemann, J.D. and Engeman, R.M. (2014).
‘Longevity of rodenticide bait pellets in a tropical environment
following a rat eradication program’. Environmental Science and
Pollution Research 21(3): 2283–2288.
178
Hegab, I.M., Pan, H., Dong, J., Wang, A., Yin, B., Yang, S. and Wei, W.
(2014). ‘Effects of physical attributes and chemical composition of
novel foods on food selection by Norway rats (Rattus norvegicus)’.
Journal of Pest Science 87(1): 99–106.
Holmes, N.D., Griffiths, R., Pott, M., Alifano, A., Will, D., Wegmann, A.S.
and Russell, J.C. (2015). ‘Factors associated with rodent eradication
failure’. Biological Conservation 185: 8–16.
Jacobs, G.H. (1993). ‘The distribution and nature of colour vision among
the mammals’. Biological Reviews 68(3): 413–471.
Jones, H.P., Holmes, N.D., Butchart, S.H.M., Tershy, B.R., Kappes,
P.J., Corkery, I., Aguirre-Muñoz, A., Armstrong, D.P., Bonnaud, E.,
Burbidge, A.A., Campbell, K., Courchamp, F., Cowan, P.E., Cuthbert,
R.J., Ebbert, S.E., Genovesi, P., Howald, G.R., Keitt, B.S., Kress, S.W.,
Miskelly, C.M., Oppel, S., Poncet, S., Rauzon, M.J., Rocamora, G.,
Russell, J.C., Samaniego-Herrera, A., Seddon, P.J., Spatz, D.R., Towns,
D.R. and Croll, D.A. (2016). ‘Invasive mammal eradication on islands
results in substantial conservation gains’. Proceedings of the National
Academy of Sciences of the United States of America 113(15): 4033–
4038.
Keitt, B.S., Griffiths, R., Boudjelas, S., Broome, K.G., Cranwell,
S., Millett, J., Pitt, W. and Samaniego-Herrera, A. (2015). ‘Best
practice guidelines for rat eradication on tropical islands’. Biological
Conservation 182: 17–26.
Lamoreux, J.F., Morrison, J.C., Ricketts, T.H., Olson, D.M., Dinerstein,
E., McKnight, M.W. and Shugart, H.H. (2006). ‘Global tests of
biodiversity concordance and the importance of endemism’. Nature
440(7081): 212–214.
Bond, et al.: Bait colour and moisture
Lavers, J.L., Wilcox, C. and Donlan, C.J. (2010). ‘Bird demographic
responses to predator removal programs’. Biological Invasions 12:
3839–3859.
Lavers, J.L., McClelland, G.T.W., MacKinnon, L., Bond, A.L., Oppel, S.,
Donaldson, A.H., Duffield, N.D., Forrest, A.K., Havery, S.J., O’Keefe,
S., Skinner, A., Torr, N. and Warren, P. (2016). Henderson Island
Expedition Report: May–November 2015. RSPB Research Report 57.
Sandy, UK: RSPB Centre for Conservation Science.
Nathan, H.W. (2016) ‘Detection Probability of Invasive Ship Rats:
Biological Causation and Management Implications’. Ph.D. thesis.
Auckland, NZ: University of Auckland.
Oppel, S., Bond, A.L., Brooke, M.de L., Harrison, G., Vickery, J.A.
and Cuthbert, R.J. (2016a). ‘Temporary captive population and
rapid population recovery of an endemic flightless rail after a rodent
eradication operation using aerially distributed poison bait’. Biological
Conservation 204B: 442–448.
Oppel, S., Lavers, J.L., Bond, A.L. and Harrison, G. (2016b). ‘Reducing
the primary exposure risk of Henderson crakes (Zapornia atra) during
aerial broadcast eradication by selecting appropriate bait colour’.
Wildlife Research 43: 298–303.
Oppel, S., McClelland, G.T.W., Lavers, J.L., Churchyard, T., Donaldson,
A.H., Duffield, N.D., Havery, S., Kelly, J., Proud, T., Russell, J., Vickery,
J.A. and Bond, A.L. (2019). ‘Seasonal variation in movements and
survival of invasive Pacific rats on a sub-tropical island: implications
for eradication’. In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell
and C.J. West (eds.) Island invasives: scaling up to meet the challenge,
pp. 200–208. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Pitt, W.C., Berentsen, A.R., Shiels, A.B., Volker, S.F., Eisemann,
J.D., Wegmann, A.S. and Howald, G.R. (2015). ‘Non-target species
mortality and the measurement of brodifacoum rodenticide residues
after a rat (Rattus rattus) eradication on Palmyra Atoll, tropical Pacific’.
Biological Conservation 185: 36–46.
R Core Team (2017). R: A Language and Environment for Statistical
Computing. Version 3.4.0 [computer program]. Vienna, Austria: R
Foundation for Statistical Computing.
Rozin, P. (1976). ‘The selection of foods by rats, humans, and other
animals’. Advances in the Study of Behavior 6: 21–76.
Russell, J.C. and Holmes, N.D. (2015). ‘Tropical island conservation: Rat
eradication for species recovery’. Biological Conservation 185: 1–7.
Spencer, T. (1995). ‘The Pitcairn Islands, South Pacific Ocean: plate
tectonic and climatic contexts’. Biological Journal of the Linnean
Society 56 :13–42.
Varnham, K. (2010). Invasive Rats on Tropical Islands: Their History,
Ecology, Impacts and Eradication. RSPB Research Report No. 41.
Sandy, UK: Royal Society for the Protection of Birds.
Weser, C. and Ross, J.G. (2013). ‘The effect of colour on bait consumption
of kea (Nestor notabilis): Implications for deterring birds from toxic
baits’. New Zealand Journal of Zoology 40(2): 137–144.
179
J. Bryce and M. Tonkin
Bryce, J. and M. Tonkin. Containment of invasive grey squirrels in Scotland: meeting the challenge
Containment of invasive grey squirrels in Scotland:
meeting the challenge
J. Bryce1 and M. Tonkin2
Scottish Natural Heritage, Great Glen House, Leachkin Road, Inverness, IV3 8NW. <jenny.bryce@nature.scot>.
2
Scottish Wildlife Trust, Harbourside House, 110 Commercial Street, Edinburgh EH6 6NF, UK.
1
Abstract Saving Scotland’s Red Squirrels (SSRS), launched in 2009, is a project to stop the decline of core populations
of Scotland’s native red squirrel. It is a partnership project between Scottish Wildlife Trust, Scottish Natural Heritage,
Forestry Commission Scotland, RSPB Scotland, Scottish Land & Estates and the Red Squirrel Survival Trust. The aim is
the containment of the invasive non-native grey squirrel, which poses a dual threat to red squirrels through competition
and disease transmission. Grey squirrels have replaced red squirrels over much of their former range in England, Wales,
Ireland and central Scotland. SSRS controls grey squirrels at a landscape-scale in three strategically selected zones: in
north-east Scotland, where the aim is eradication of an isolated grey squirrel population; coast to coast along the Highland
Boundary Fault the aim is to prevent northwards incursion of grey squirrels into the Scottish Highlands and Argyll, where
red squirrel is still the only species; and in southern Scotland, the aim is now to prevent replacement of priority red squirrel
populations by focussing control in areas identified as having the best prospects for the long-term maintenance of red
squirrel populations. Control methods involve live cage-trapping combined with humane dispatch. The control network
comprises SSRS and Forestry Commission controllers, private landowners supported by EU/government funding and a
large number of individual volunteers. The work is dependent on wide public acceptance and active volunteer support.
To date SSRS has been successful at significantly reducing grey squirrel geographic range and occupancy in NE Scotland
and as well as reducing the incidence of grey squirrels north of the ‘Highland Line’ to no more than the occasional
occurrence. In southern Scotland grey squirrel control has contributed to the maintenance of red squirrel populations
despite the continued spread of squirrelpox in grey squirrels. The major challenge now is sustaining the level of grey
squirrel control needed to secure Scotland’s red squirrel populations in the long term. A new project phase started in
2017, focused on building community action networks until such a time as alternatives means of controlling grey squirrel
numbers and disease impacts become widely available.
Keywords: adaptive management, community engagement, land manager, sustainability, trapping effort, volunteer
INTRODUCTION
Grey squirrels (Sciurus carolinensis) were introduced
into Britain in the 1890s from the US and Canada, including
to several release sites in Scotland (Middleton, 1930;
Middleton, 1931). The impact of grey squirrels on native
red squirrel (Sciurus vulgaris) populations was documented
relatively early after their introduction (Middleton, 1931;
Shorten, 1962), but their range expansion was initially
quite modest (Gurnell, 1987). The role of grey squirrels
in the replacement of red squirrels was possibly not fully
recognised until the 1980s (Lloyd, 1983; Skelcher, 1997;
Reynolds, 1998), by which time grey squirrels occupied
much of southern and central England and Wales and
central Scotland (Lloyd, 1983).
The factors leading to the replacement of red by grey
squirrels have been the subject of extensive research
(reviewed by Gurnell, et al., 2014b). The evidence indicates
that competition with grey squirrels for food resources
alone can account for the loss of red squirrels from many
forests (Bryce, et al., 2001; Wauters, et al., 2002). However,
added to this is the threat of squirrelpox virus, which is
carried by grey squirrels and is highly pathogenic to red
squirrels (Sainsbury, et al., 2000, Thomas, et al., 2003),
greatly enhancing the speed of replacement (Rushton, et
al., 2006).
Red squirrels have been protected under UK law since
the 1930s and bounty schemes were enlisted to combat
increasing grey squirrel numbers in the 1950s (Sheail,
1999). However, low-level, sporadic control has failed
to halt the spread of grey squirrels (Lawton & Rochford,
2007). Grey squirrels are already widespread and abundant
throughout much of the UK and eradication is not
considered to be a realistic option (Gurnell & Pepper, 1993;
Pepper & Patterson, 1998). EU Regulation 1143/2014 on
Invasive Alien Species lists grey squirrels as species of
Union concern, hence Member States are required to take
concerted management action to ensure they do not spread
any further and to minimise the harm they cause to the
environment.
Large-scale control and containment of grey squirrels
was originally seen as an interim approach, whilst more
sustainable, long-term control measures were developed
(Scottish Squirrel Group, 2004; Scottish Natural Heritage,
2010). However, in the absence of a squirrelpox vaccine
or immuno-contraceptive, there has been growing support
for targeted grey squirrel control to protect red squirrel
populations (Scottish Natural Heritage, 2010). Following
public consultation, a draft strategy for grey squirrel control
in Scotland was published (Scottish Natural Heritage,
2010), which focuses on targeted control to maximise the
benefits for red squirrels.
A collaborative project under the heading ‘Saving
Scotland’s Red Squirrels’ (SSRS), was formalised in 2009.
It is a partnership comprising the Scottish Wildlife Trust
(SWT), Scottish Natural Heritage, Forestry Commission
Scotland, RSPB Scotland, Scottish Land & Estates and
the Red Squirrel Survival Trust. SSRS has become the
principal means of coordinating red squirrel protection in
Scotland. The focus of SSRS is on applied conservation
action, but SSRS has made a concerted effort to collate
records and monitor squirrel populations and has worked
closely with researchers to inform an adaptive approach.
A key challenge to SSRS has been to assess the efficacy
and sustainability of control measures. This paper explores
some of the work carried out to address these challenges
and highlights some of the learning to date.
THE SSRS APPROACH
Co-ordination of grey squirrel control
The main focus of SSRS activity is the co-ordination
of grey squirrel control. SSRS aims to co-ordinate grey
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
180
up to meet the challenge, pp. 180–186. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Bryce & Tonkin: Grey squirrels in Scotland
squirrel control across three strategic control zones (Fig.
1). The aims vary between zones reflecting the degree
to which grey squirrels are already established. The
Highlands of Scotland are currently free of grey squirrels;
grey squirrels are long established in central Scotland
(with introductions between 1892–1919) and had spread
throughout much of the Central Lowlands by the 1980s;
grey squirrels have spread into South Scotland from
both the north and from northern England in recent years
(Gurnell, et al., 2014a). The grey squirrel population in NE
Scotland (Aberdeen and Aberdeenshire) has been recorded
for about 30 years (Lloyd, 1983; Staines, 1986), and there
is evidence it originates from a separate introduction
rather than having spread from grey squirrels elsewhere in
Scotland (Signorile, 2013). Squirrelpox was first reported
to have crossed from northern England into Scotland in
2005, with the first cases in red squirrels observed in 2007
(McInnes, et al., 2009). The original SSRS aims are
listed as follows, although these have been adapted in
light of experience as is discussed later.
North-east (NE) Scotland – the original aim was to
halt the spread of grey squirrels outwards from the city of
Aberdeen;
Central Lowlands – SSRS’s work in the Central
Lowlands aims to contain the northward spread of grey
squirrels into the Highlands and Argyll by carrying out
control from coast to coast; along what is referred to as
the ‘Highland Line’ a zone of control extending for some
160 km from just north of Glasgow to Montrose on the
east coast.
Southern Scotland – the initial aim was to contain the
spread of squirrelpox virus in south Scotland.
Grey squirrel control is currently carried out by a
mixture of project staff, land managers and volunteers:
● Working with up to 197 landowners under five-year
EU and Scottish Rural Development Programme
funding (SRDP);
Fig. 1 Map of Saving Scotland’s Red Squirrels project
control zones.
● Co-ordination via five conservation officers, six fulltime and five part-time grey squirrel control officers;
● A trap-loan scheme involving up to 200 landowners,
and 500 individuals.
Trapping is also carried out by Forest Enterprise
Scotland at key sites on the National Forest Estate. Figure
2 illustrates the coverage of grey squirrel control initiated
by SSRS by 2012 (from Tonkin, et al., 2015).
With a view to the long term, SSRS has sought to
encourage land managers to carry out grey squirrel control
on their own land. Regional conservation staff provide
support to landowners applying for funding available
through the SRDP to help cover the costs. SRDP contracts
require land managers to operate an appropriate number of
traps (as advised by SSRS staff) for a minimum number of
sessions per year (usually five or six). All the traps are to
be set for a minimum number of days (usually 10). SSRS
control officers trap with landowner permission, in key
gaps in the landowner protection network (Fig. 2). Most
of the grey squirrel control occurs between April and the
end of September, when grey squirrels are easier to catch.
Due to the lack of specificity of other methods and animal
welfare considerations (Central Science Laboratory, 2009),
the SSRS Standard Operating Procedures specify the use
of cage trapping and humane dispatch.
Trapping by SSRS control officers has typically
followed the approach of five days pre-baiting and then
trapping continuously until no or few further grey squirrels
are caught. Traps are then revisited on a rotation. The
traps are not located at a standardised density, but instead
are grouped in areas of preferred grey squirrel habitats in
the target zones. In South Scotland grey squirrel control
by SSRS control officers was initially carried out in the
area buffering known squirrelpox seropositive cases. The
grey squirrels were then sent for laboratory testing for
squirrelpox. Hence control effort was reactive and did not
take place in the same locations across time.
In NE Scotland, in particular, SSRS is working in periurban and urban areas. Initially this created a challenge
because large areas of wooded habitats in private gardens
and parks were difficult to access, leaving reservoirs of
grey squirrels. To address this challenge and to harness
the community enthusiasm in other areas, SSRS has
successfully instigated a trap-loan scheme. Under this
scheme householders take responsibility for setting and
monitoring traps (supplied by SSRS) in their garden. They
Fig. 2 Red squirrel protection network established across
Scotland by 2012 (from Tonkin, et al., 2015).
181
Island invasives: scaling up to meet the challenge. Ch 1C Rodents: Lessons
are matched with a local trapper (gamekeeper or control
officer) who is available to carry out humane dispatch.
Project staff have also developed innovative trap designs
for trapping grey squirrels in city parks to avoid drawing
attention to traps.
Grey squirrel control in public spaces and, in particular,
the involvement of volunteers requires a high degree of
public acceptance, which is not guaranteed (Bertolino
& Genovesi, 2003). A targeted approach to control was
broadly supported in a public consultation (Scottish
Natural Heritage, 2010) and public surveys in Aberdeen
and Aberdeenshire have established that despite residents
enjoying seeing grey squirrels, there is an appreciation of
the need for grey squirrel control due to their impact on
red squirrels (Ashbrook Research Consultancy Ltd, 2010).
Evaluation of control measures
Alongside establishing the network of grey squirrel
control, SSRS has sought to collect evidence that this
work is benefitting red squirrels. This has been critical for
securing public funding for grey squirrel control. Three
methods were employed by the SSRS in order to evaluate
the efficacy of grey squirrel control:
● Evaluation of grey squirrel capture probability from
trapping data;
● Annual (presence/absence) monitoring of red and
grey squirrel occupancy in the three project areas,
and
● Public sightings of squirrels across Scotland that
have been catalogued since 2007.
Annual (presence /absence) monitoring of red and grey
squirrels has been co-ordinated by SSRS in NE Scotland
and the Central Lowlands since 2011 and since 2013 in
South Scotland. The surveys are intended to assess if there
are changes in squirrel distributions that can be attributed
to the project. Nearly 200 volunteers have been mobilised
to carry out these surveys. A sample of 2 km ×2 km grid
squares or ‘tetrads’ are surveyed across each control zone.
Four baited feeder-boxes are permanently located in
woodland within each tetrad. Each feeder-box is checked
by volunteers three times over a period of six weeks each
spring. Hairs are identified under a microscope and each
tetrad is consequently allocated to one of the following
four categories: “red squirrels only”, “grey squirrels only”,
“both species”, or “neither” species (Fig. 3, Shikhorshidi
& Tonkin, 2018). The number of tetrads has increased over
time, but comparison of the same tetrads over time enables
detection of changes in squirrel distributions. Changes
between years have been explored using a replacement
index as per Usher, et al. (1992). A positive index represents
a change in tetrad occupancy in favour of grey squirrels
and a negative index, a change in favour of red squirrels
(Usher, et al., 1992).
A programme of squirrelpox surveillance has also
helped guide the work in South Scotland. In 2012, grey
squirrels were sampled from a systematic sample of
locations across the whole of South Scotland to try and
establish the full extent of exposure to the virus (10 grey
squirrels are sampled from one 10 km square in every 20
km × 20 km square across the region).
RESULTS
Evaluation of grey squirrel capture probability
The project initially aimed to gather data on grey
squirrel trapping across all three control zones, however,
inconsistencies in recording between the different project
delivery models has made it problematic to fully assess
the cumulative trapping effort that has been achieved. The
182
control officers’ data are the most reliable. The format of
other records varies, effort is not always systematically
recorded and problems have been encountered (data
gathering and ownership) in accessing results of trapping
from the land mangers supported by SRDP funding.
Forestry Enterprise Scotland controllers’ data are included
with the control officers’ data where this has been possible.
The minimum total number of grey squirrels controlled
and the trapping effort achieved have been estimated
from collated data (SWT pers. comm.). It is estimated that
between 2009 and 2016:
● Control officers provided c. 214,000 trap days (the
number of traps multiplied by the number of days
for which traps are set) and controlled c. 13,000 grey
squirrels (Table 1); and
● Up to 197 SRDP contracts were established (Scottish
Government Statistics, 2014), with those reporting
accounting for 1.1 million trap days and having
controlled c. 18,000 grey squirrels Table 1).
Those in receipt of trap loans have not consistently
reported trapping effort, however, the trap loan scheme
in NE Scotland has made a larger contribution to the red
squirrel protection network than elsewhere, with trap loans
here accounting for the removal of more than 1,700 grey
squirrels between 2010 and 2016 (SWT pers. comm.).
Due to the scale of the task to follow up on missing
trapping information, SSRS have sought to collect as
complete trapping data as possible for four demonstration
areas in order to assess the cumulative impact on grey
squirrel capture probability (a proxy for abundance). The
size of demonstration areas is not equivalent but as an
illustration of control effort (control officer and landowner
data), the total number of trap days in 2014 in NE Scotland
demonstration area (55 km2) was 6,614 trap days, in Tayside
(222 km2) was 15,004 trap days, in Argyll & Trossachs
(278 km2) was 6,482 trap days and in South Scotland (604
km2) was 15,206 trap days (Table 2). Only the NE Scotland
and Tayside demonstration areas had generated sufficient
time series of data for detailed analysis by 2013 as reported
in Tonkin, et al. (2013).
Using all the available trapping data for the NE
Scotland Demonstration Area between 2007 and June 2013
and the Tayside Demonstration Area from 2010 to 2012,
a GLMM was used to explore the relationship between
the probability of grey squirrel capture and a range of
explanatory variables including the cumulative control
effort for each trap location, taking account of nearby
captures. There was found to be a significant negative
effect of cumulative control effort on the probability of
grey squirrel capture in both areas (Tonkin, et al., 2013).
In Tayside the GLMM coefficient was -1.54 (CI -1.99 –
-1.09), and in NE Scotland was -0.34 (CI -0.51 – -0.18)
(both on link scale of logit model). Effects were found to be
stronger in areas with the highest cumulative control effort.
In these areas in Tayside the mean capture probability was
close to zero and in NE Scotland was seven-fold lower
than areas with relatively low effort (Tonkin, et al., 2013,).
These results support the premise that trapping is having
the desired effect of reducing grey squirrel abundance.
The reactive pattern of trapping in South Scotland in
response to detecting squirrelpox, makes the data from this
region problematic for assessing the impact of trapping on
grey squirrel abundance. Added to this, despite the scale of
trapping effort in South Scotland, it was apparent that the
virus was still spreading (White & Lurz, 2014; Tonkin, et
al., 2015). Hence SSRS sought the help of researchers to
assess if containment of the virus was a realistic objective.
White & Lurz (2014) used a spatially explicit population
model to explore the spread of the disease under a range
of control scenarios and levels of effort. Simulated control
Bryce & Tonkin: Grey squirrels in Scotland
Fig. 3 Results of spring survey tetrads with both species, either species or none detected for a) North-east Scotland b)
Tayside and (c) South Scotland for the years indicated (adapted from Shikorshidi & Tonkin 2018).
183
Island invasives: scaling up to meet the challenge. Ch 1C Rodents: Lessons
Table 1 Grey squirrel control achieved in all SSRS regions by control officers and land managers supported by
SRDP funding.
Grey Squirrel Officer Control
SRDP supported land manger control
Year
Total no of trap days
No of greys captured
2009
6,610
471
2010
18,615
1,637
2011
34,150
2,191
2012
40,783
2,797
2013
42,991
1,690
2014
32,889
1,819
2015
27,833
2,013
2016
9,776
490
Total
213,647
13,108
Average grey squirrel capture rate/100TN
6.14
was parameterised to assess the impact of current control
measures; approximating the number of grey squirrels
removed and adjusting the intensity of control in the model
to mirror these levels by varying the area over which
control was applied (White & Lurz, 2014). An alternative
control scenario involving control along key dispersal
routes was also assessed. The projections highlighted that
current levels of control would not prevent the spread of
the disease across Southern Scotland. Targeted control
could help slow the spread of the virus but was unlikely
to halt its spread in areas where grey squirrels are already
established. However, the modelling also indicated that
co-ordinated grey squirrel control should allow local red
squirrel populations to persist and their density can recover
after disease outbreaks in conifer dominated landscapes
(White & Lurz, 2014; White, et al., 2016).
Annual (presence/absence) monitoring red and grey
squirrels
The programme of annual presence/ absence monitoring
indicates that red squirrel distributions have remained
stable and that there have been some reductions in the
range of grey squirrels in north Scotland and conversely
some expansion in south Scotland (Fig. 3, Shirkhorshidi &
Tonkin, 2018). The 2017 results of the tetrads in the north of
Scotland as a whole (NE Scotland and Central Lowlands)
show a significant change in favour of red squirrels when
compared with 2012 (RI=-0.17, P=0.02). Contributing to
this is a significant decrease in grey squirrel occupancy
across the north and an increase in red squirrel distribution
in the north-east, particularly in areas close to the City of
Aberdeen; meanwhile red squirrel occupancy has been
stable across the Highland Line (Shirkhorshidi & Tonkin,
2018). Although not significant, the overall changes in the
south of Scotland have been in favour of grey squirrels
(2013–2017 RI = 0.19). This reflects an increase in
grey squirrel occupancy (largely outside SSRS areas of
operation), whilst red squirrel occupancy appears to have
been maintained (Shirkhorshidi & Tonkin, 2018). Whilst
noting that squirrel populations experience fluctuations
between years relating to seed crops, we interpret the
overall trends in occupancy as an indication that SSRS’s
actions are helping to meet the project aims.
Public sightings of squirrels
Although they do not represent a systematic sample,
public sightings help to harness public support and provide
an early warning of range expansion of both red and grey
squirrels. Sightings are mapped on the SSRS website
184
Total no of trap days No of greys captured
7,610
71
41,365
510
76,906
2,817
237,738
3,630
256,456
3,730
222,031
2,713
172,222
3,293
89,316
1,029
1,103,644
17,793
1.61
(SSRS, 2018). For example, public reports have helped
illustrate where red squirrels have returned to areas where
they had not been recorded in the last 20 or 30 years
following grey squirrel control, such as Aberdeen city
parks (SSRS, 2017).
There has also been some standardisation and analysis
of public sightings of squirrels between 1991 and 2010
across different regions of the UK (Gurnell, et al., 2014a).
The data suggest red squirrel occupancy is declining in
all regions over this period (at different rates), with the
exception of Central Lowlands (east) of Scotland which
fluctuates showing little overall change. However, an
upward turn in red squirrel occupancy in the last two or
three years is noted across all regions, especially South
Scotland. Gurnell, et al. (2014a) indicate it is too early
to speculate if the apparent upturn in the fortunes of red
squirrels is as a result of grey squirrel control or other
factors.
DISCUSSION AND FURTHER DEVELOPMENTS
We have described the evidence showing that sustained
grey squirrel trapping can reduce grey squirrel abundance
and occupancy at a landscape scale (Tayside and NE
Scotland). Trapping data from Wales around the same time,
indicates that sustained trapping can bring about reductions
in grey squirrel populations at a landscape scale (Schuchert,
et al., 2014). However, it has also been demonstrated that
recolonisation can occur following intensive grey squirrel
control after between one and three months (Lawton &
Rochford, 2007; Schuchert, et al., 2014). Hence, SSRS
are involved in further work to better quantify the level of
control that might need to be sustained and, in particular, to
put in place more sustainable delivery models.
The collaboration with researchers that started in South
Scotland is now focussed on addressing the question of
how much control may be required along the Highland
Line to prevent grey squirrels from extending their range
to the north. A spatially explicit population model (White,
et al., 2017) has examined the impact of three levels of
trapping intensity on grey squirrel populations in the
Central Lowlands. Projections include the presence of
squirrelpox virus (as a worst-case scenario) even though
it has not yet been detected in this region. The potential
density of red and grey squirrels in each 1 km × 1 km
patch is derived based on average squirrel densities for
the mixture of habitats encountered. Control is applied in
targeted zones (typically 10 km ×10 km) and, mirroring
trapping practice, can occur from 1 April to 30 September
Bryce & Tonkin: Grey squirrels in Scotland
(183 days), which is split into three 61 day (2 month)
control periods. Trapping is applied in a responsive way
to grid squares in which grey squirrels are present and
in grid squares in a 2 km buffer zone in each of the three
control periods. The model was run for three levels of trap
intensity (TD = 0.3 – low; 0.5 – medium; 0.75 – high). This
equates to 0.3 × 183 = 55 trap days per year (in a 1 km ×1
km grid square) in the low intensity scenario, 92 trap days
per year (medium) and 137 trap days per year per (high),
respectively.
In the low intensity scenario, trapping represents a kind
of harvesting; there are abundant greys to catch and greys
persist indefinitely. At medium intensity, control appears
to be largely effective at preventing the northwards spread
along the Highland Line. The model also highlights key
dispersal routes where high intensity trapping is likely
to be required (White, et al., 2017). Taking the Tayside
Demonstration area (a key dispersal route) as an example
the average, annual control effort predicted to be required
(regions 7, 8, 9 in White, et al., 2017) is c. 18,000 trap
days under the high intensity scenario. Given the modelled
control area includes a slightly larger area than the Tayside
demonstration area in order to prevent recolonisation, the
levels of control suggested by the model are of the same
order as control on the ground between 2012 and 2014
(Table 2) suggesting that this level of control effort needs
to be maintained (White, et al., 2017).
Having successfully reduced the range, abundance
and occupancy of grey squirrel populations in NE
Scotland (Tonkin, et al., 2013; Shirkhorshidi & Tonkin,
2018), eradication of this isolated population now seems
like a realistic prospect. However, some of the locations
remaining untrapped are more challenging (smaller,
fragmented and increasingly urban habitats). In 2014,
SSRS set up an additional layer of monitoring to establish
grey squirrel occupancy across the entire wooded network
in the region. Feeder-box squirrel hair traps (n=223) have
been distributed through all the suitable grey squirrel
habitat patches in urban Aberdeen and the surrounding
area. These data will allow analysis of grey squirrel
occupancy (MacKenzie, et al., 2006) and better projections
for the time and effort required to eradicate this isolated
population. The monitoring will be complemented by
rapid-response grey squirrel control.
SSRS’s approach in South Scotland has adapted
following the continued spread of squirrelpox virus and
the model outcomes reported in White & Lurz (2014). The
modelling indicates that co-ordinated grey squirrel control
can help to protect red squirrel populations from the threat
of squirrelpox virus in conifer dominated landscapes,
Table 2 Combined grey squirrel control effort (annual
trap days) achieved by control officers and landowners
supported by SRDP funding in the four demonstration
areas 2009–2016.
NE
Scotland
2009
2010
2011
2012
2013
2014
2015
2016
2,465
5,946
7,878
10,554
6,178
6,614
7,500
4,840
Central Scotland
Argyll &
Tayside
Trossachs
NA
NA
3,389
48
8,201
360
14,677
6,803
16,158
6,721
15,004
6,482
9,780
6,590
7,990
6,294
*Landowner data not available
South
Scotland
4,987
14,678
17,912
16,934
20,079
15,206
5,114
1,654*
where red squirrels typically occur at low densities, but
importantly higher than those of grey squirrels. SSRS’s
control efforts have now shifted from the ‘frontline’ of
squirrelpox detection to protecting identified priority areas
for red squirrel conservation in South Scotland.
Sustaining the action
SSRS was initially funded for three years and eight
months (2008–12), which was then extended by a further
four years. The lead partner, SWT, secured a mixture of
public and charitable funds to meet a project budget of
just over £3 million covering the period 2008–16. In 2016,
SWT secured a Heritage Lottery Fund Award of £2.46
million for the next 5 years (until 2022) towards a total
project cost of £4.4 million. Hence, the costs have been
roughly £0.5 million per annum to date. The piecemeal
nature of project funding creates a challenge for sustaining
co-ordinated grey squirrel control. Under the new SSRS
phase, costs are anticipated to rise to c. £0.88 million per
annum reflecting the additional activities aimed at ensuring
the long-term sustainability of the control network and
with a view to substantially reducing costs thereafter.
By September 2013, 197 landowners were in receipt
of five-year SRDP funding at a cost of £4.5 million over
the five-year period covered by the contracts. Although
trapping by control officers is on average nearly four times
more efficient (more captures per 100 trap days, Table 1)
than SRDP-supported landowner grey squirrel control,
landowner control provides five times more trapping effort
than is provided by control officers (Table 1). Hence,
being able to access public funding support has been
hugely important. However, public funding is not without
its challenges including: ease of access to the scheme
for applicants; and ensuring trapping data are available
to SSRS. Added to this there are uncertainties about the
future of support upon leaving the European Union.
Quantifying the control effort needed to deliver
SSRS’s objectives has been challenging. However, SSRS’s
monitoring and associated modelling has supported that the
levels achieved seem ‘about right’. However, this equates
to a substantial network of grey squirrel control that needs
to be sustained.
Reflecting the successes to date and the challenges
ahead, the next five-year phase is called Saving Scotland’s
Red Squirrels – Developing Community Action. This
project’s actions are geared towards long-term sustainability
and how SSRS’s work can be embedded in routine land
management and community action, with a move away
from reliance on project staff. Project funding at this level
of investment is increasingly hard to find, hence there is an
expectation that red squirrel conservation will increasingly
rely on public delivery.
SSRS – Developing Community Action now aims to
eradicate grey squirrels from NE Scotland within 10 years.
In South Scotland the aims have been refined and focus on
building the skills and resources available to local people
and land managers working to control grey squirrels in
identified priority areas. As part of this, a Community Hub
information management system is being developed for
staff and volunteers, which will better capture and integrate
data from all sources and will allow improved feedback.
For each priority area, an annual trapping programme is
being developed that is capable of continuing to protect the
red squirrel population. As yet it remains to be determined
if the necessary levels of control can be sustained by these
means. However, there is a shift in the focus of SSRS work
from demonstrating the efficacy of control on to how can
it be delivered.
Largely based on the evaluation of work co-ordinated
by SSRS, the national policy position now recognises
coordinated grey squirrel control as an integral part of the
185
Island invasives: scaling up to meet the challenge. Ch 1C Rodents: Lessons
long-term approach to achieving the strategy aims (Scottish
Squirrel Group, 2015).
The challenge of protecting Scotland’s red squirrels
remains significant given the scale and the ongoing nature
of the work. However, the prospects for alternative/ or
complementary approaches are also improving. Immunocontraceptives and squirrelpox vaccines are actively being
explored with the support of parallel initiatives under the
‘UK Squirrel Accord’ but are likely to be some years in
development. New research into the role of pine marten on
the dynamics of red and grey squirrels also offers promising
insights in that as pine marten populations recover their
range and densities, grey squirrel populations appear to be
suppressed in the presence of this novel (to them) native
predator, thereby reducing the levels of management
control required to promote red squirrel persistence
(Sheehy, et al., 2018).
ACKNOWLEDGEMENTS
The Saving Scotland’s Red Squirrels partnership
includes the Scottish Wildlife Trust, Scottish Natural
Heritage, Forestry Commission Scotland, Royal Society
for the Protection of Birds (Scotland), Scottish Land
& Estates and the Red Squirrel Survival Trust. The
partnership is grateful to all organisations and individuals
that have made funding contributions. We are particularly
grateful to all the landowners and managers, gamekeepers,
rangers, researchers and volunteers who have so willingly
assisted the project.
REFERENCES
Ashbrook Research Consultancy Ltd (2010). Assessment of Public
Attitudes to Grey Squirrel Control in Aberdeen. Scottish Natural
Heritage, Commissioned Report No. 384. Battleby, Perth: Scottish
Natural Heritage.
Bertolino, S. and Genovesi, P. (2003). ‘Spread and attempted eradication
of the grey squirrel (Sciurus carolinensis) in Italy, and consequences for
the red squirrel (Sciurus vulgaris) in Eurasia’. Biological Conservation
109(3): 351–358.
Bryce, J., Johnson, P.J. and Macdonald, D.W. (2001). ‘Can niche use
in red and grey squirrels offer clues for their apparent coexistence?’
Journal of Applied Ecology 39: 875–887.
Central Science Laboratory (2009). Review of Methods of Humane
Destruction of Grey Squirrels (Sciurus carolinensis). Scottish Natural
Heritage Commissioned Report No, 317. Battleby, Perth: Scottish
Natural Heritage.
Gurnell, J. (1987). The Natural History of Squirrels. London: Christopher
Helm.
Gurnell, J. and Pepper, H. (1993). ‘A critical look at conserving the
British red squirrel (Sciurus vulgaris)’. Mammal Review 23: 125–136.
Gurnell, J., Lurz, P.W.W. and Bertolino, W. (2014a). ‘The changing
patterns in the distribution of red and grey squirrels in the north of
England and Scotland between 1991 and 2010 based on volunteer
surveys’. Hystrix, the Italian Journal of Mammalogy 25 (2): 83–89.
Gurnell, J., Lurz, P.W.W. and Wauters, L.A. (2014b). ‘Years of Interactions
and Conflict in Europe: Competition Between Eurasian Red Squirrels
and North American Grey Squirrels’. In: C.M. Shuttleworth, P.W.W.
Lurz and M.W. Hayward (eds.) Red Squirrels: Ecology Conservation
and Management in Europe, pp. 19–37. Llangefni: European Squirrel
Initiative.
Lawton, C. and Rochford, J. (2007). ‘The recovery of grey squirrel
(Sciurus carolinensis) populations after intensive control programmes’.
Biology and Environment: Proceedings of the Royal Irish Academy Vol.
107B, No. 1: 19–29.
Lloyd, H.G. (1983). ‘Past and present distributions of red and Grey
squirrels’. Mammal Review 13: 69–80.
MacKenzie, D.I., Nichols, J.D., Royle, A.J., Pollock, K.H., Bailey,
L.L. and Hines, J.E (2006). Occupancy Estimation and Modlling:
Inferring Patterns and Dynamics of Species occurrence. Burlington,
Massachusetts: Elsevier Academic Press.
McInnes, C.J., Coulter, L., Dagleish, M.P., Fiegna, C., Gilray, J.,
Willoughby, K., Cole, M., Milne, E., Meredith, A., Everest, DJ., and
MacMaster, A. (2009). ‘First cases of squirrelpox in red squirrels
(Sciurus vulgaris) in Scotland’. Veterinary Record 164: 528–531.
Middleton, A.D. (1930). ‘The ecology of the American grey squirrel
(Sciurus carolinensis Gmelin) in the British Isles’. Proceedings of the
Zoological Society of London 100: 809–843.
Middleton, A.D. (1931). The Grey Squirrel. London: Sidgwick & Jackson.
Pepper, H., and Patterson, G. (1998). Red Squirrel Conservation. Forestry
Commission Practice Note 5. Edinburgh: Forestry Commission. <www.
forestry.gov.uk/pdf/fcpn5.pdf/$file/fcpn5.pdf>.
186
Reynolds, J. (1998). ‘Details of the geographic replacement of the red
squirrel (Sciurus vulgaris) by the grey squirrel (Sciurus carolinensis) in
Eastern England’. Journal of Animal Ecology 54: 149–162.
Rushton, S.P., Lurz, P.W.W. Gurnell, J., Nettleton, P., Bruemmer, C. and
Shirley, M.D.F. (2006). ‘Disease threats posed by alien species: the
role of a poxvirus in the decline of the native red squirrel in Britain’.
Epidemiology and Infection 134: 521–533.
Sainsbury, A. W., Nettleton, P., Gilray, J. and Gurnell, J. (2000). ‘Grey
squirrels have a high seroprevalence to a parapoxvirus associated with
deaths in red squirrels’. Animal Conservation 3: 229–233.
Schuchert, P., Shuttleworth, C.M., McInnes, C.J., Everest, D.J. and
Rushton, S.P. (2014). ‘Landscape scale impacts of culling upon a
European grey squirrel population: Can trapping reduce population size
and decrease the threat of squirrelpox virus infection for the native red
squirrel?’. Biological Invasions 16: 2381–2391.
Scottish Government Statistics (2014) ‘Number and value of Rural
Development Contract: Rural Priorities Options funding under
the 2007–13 SRDP’. [Statistics web resource]. <www.gov.scot/
Resource/0044/00449780.pdf>. Accessed 20 February 2018.
Scottish Natural Heritage (2010). Protecting Scotland’s Red Squirrels.
The Consultation Response. Battleby, Perth: Scottish Natural Heritage.
Scottish Squirrel Group (2004). Scottish Strategy for Red Squirrel
Conservation. Battleby, Perth: Scottish Natural Heritage.
Scottish Squirrel Group (2015). Scottish Strategy for Red Squirrel
Conservation. Battleby, Perth: Scottish Natural Heritage. https://www.
nature.scot/scottish-strategy-red-squirrel-conservation-june-2015.
Sheail, J. (1999). ‘The grey squirrel (Sciurus carolinensis) – A UK
historical perspective on a vertebrate pest species’. Journal of
Environmental Management 55: 145–156.
Sheehy, E., Sutherland, C., O’Reilly, C. and Lambin, X. (2018). ‘The
enemy of my enemy is my friend: native pine marten recovery reverses
the decline of the red squirrel by suppressing grey squirrel populations’.
Proceedings of the Royal Society B 285: 20172603. <://dx.doi.
org/10.1098/rspb.2017.2603>. Accessed 9 April 2018.
Shorten, M. (1962). ‘Squirrels, their biology and control’. Ministry of
Agriculture Fisheries and Food Bulletin 184: 1–44.
Shirkhorshidi, M. and Tonkin, M. (2018) Results of Spring 2017 Squirrel
Surveys. Saving Scotland’s Red Squirrels report. Edinburgh: Scottish
Wildlife
Trust.
<www.scottishsquirrels.org.uk/docs/087_506__
ssrssquirrelsurvey2017_1517589214.pdf>.
Signorile, A.L. (2013). ‘Genetic Determinants of the Expansion of
Eastern Grey Squirrel Populations Across Europe’. PhD thesis. London:
Imperial College London and Institute of Zoology, Zoological Society.
Skelcher, G. (1997). ‘The Ecological Replacement of Red by Grey
Squirrels’. In: J. Gurnell and P.W.W. Lurz (eds.) The conservation of
Red Squirrels, Sciurus vulgaris L., pp. 67–78. London: Peoples Trust
for Endangered Species.
SSRS (2017). ‘New sightings signal red squirrels’ return to Aberdeen
city parks and gardens’. [website] (published 15/11/2017). <www.
scottishsquirrels.org.uk/news/article/new-sightings-signal-redsquirrelsrsquo-return-to-aberdeen-city-parks-and-gardens/>.
SSRS (2018).
‘Map of squirrel sightings’ [website] <www.
scottishsquirrels.org.uk/squirrel-sightings/squirrel-sighting-map.>
Accessed 20 February 2018.
Staines, B. (1986). ‘The spread of grey squirrels (Sciurus carolinensis)
into north-east Scotland’. Scottish Forestry 40: 190–196.
Thomas, K., Tompkins, D.M., Sainsbury, A.W. Wood, A.R., Dalziel, R.,
Nettleton, P.F and McInnes, C.J. (2003). ‘A novel poxvirus lethal to
red squirrels (Sciurus vulgaris)’. Journal of General Virology 84: 3337–
3341.
Tonkin, M., Balmforth, L. and Ramoo, K. (2013). Evaluation of Grey
Squirrel Control in the Saving Scotland’s Red Squirrels Partnership
Project (Phase Two) – Second Report 2009–2013. Unpublished Report
to SSRS Steering Group Partners. Edinburgh: Scottish Wildlife Trust.
Tonkin, M., Garritt, J., Bryce, J. and Cole, M. (2015). ‘Species
Management in Scotland 2007–2012: Red and Grey Squirrels’. In:
M.J. Gaywood, P.J. Boon, D.B.A. Thompson and I.M. Strachan (eds.)
The Species Action Framework Handbook, pp. 82–93. Battleby, Perth:
Scottish Natural Heritage.
Usher, M.B., Crawford, T.J. and Banwell, J.L. (1992). ‘An American
invasion of Great Britain: The case of the native and alien squirrel
(Sciurus) species’. Conservation Biology 6: 108–115.
Wauters, L.A., Gurnell, J., Martinoli, A. and Tosi, G. (2002). ‘Interspecific
competition between native Eurasian red squirrels and alien grey
squirrels: does resource partitioning occur?’. Behavioural Ecology and
Sociobiology 52: 332–341.
White, A. and Lurz, P.W.W. (2014). A Modelling Assessment of Control
Strategies to Prevent/Reduce Squirrelpox Spread. Scottish Natural
Heritage Commissioned Report 627, Battleby, Perth: Scottish Natural
Heritage.
White, A., Lurz, P. W. W., Bryce, J., Tonkin, M., Ramoo, K., Bamforth,
L., Jarrott, A. and Boots, M. (2016). ‘Modelling disease spread in real
landscapes: Squirrelpox spread in Southern Scotland as a case study’.
Hystrix, the Italian Journal of Mammalogy 27 (1). DOI: 10.4404/
hystrix-27.1-11657.
White, A, Lurz, P.W.W. and Boots, M. (2017). Grey Squirrel Control
Along the Highland Line: A Model Analysis. NERC Innovation grant
report (Ref: NE/M021319/1). <www.macs.hw.ac.uk/~awhite/SWT_
SNH_HighlandLine_FinalReport.pdf>.
A. Carter, R. van Dam, S. Barr and D. Peters
Carter, A.; R. van Dam, S. Barr and D. Peters. Testing auto-dispensing lure pumps for incursion
control of rats with reduced effort on a small, re-invadable island in New Zealand
Testing auto-dispensing lure pumps for incursion control of rats with
reduced effort on a small, re-invadable island in New Zealand
A. Carter1, R. van Dam2, S. Barr2 and D. Peters3
Department of Ecology, Evolution & Organismal Biology, Iowa State University, 2200 Osborne Dr., Ames, Iowa
50011, USA. 2Goodnature Limited, 8 Horner St., Wellington 6021, New Zealand. 3Department of Conservation Te Papa
Atawhai, 18–32 Manners St., Wellington 6011, New Zealand. <dpeters@doc.govt.nz>.
1
Abstract In locations with a high potential for re-invasion, such as inshore islands, sustained control of invasive species
is as important as the initial knock-down for the long-term recovery of native populations. However, ongoing trap
maintenance and lure replenishment are barriers to minimising the time and financial costs of long-term suppression,
even when automatic traps are used. Control of invasive mammal species is a high priority for the more than 200 islands
within Rakiura National Park in southern New Zealand, many of which support nationally and internationally threatened
endemic species and ecosystems. We previously used automatic, toxicant-free traps to control rats on Native Island, a
62 ha inshore island within the National Park, where tracking indices were 73% in mid-2013. After 18 months, tracking
indices remained below 10%, and site visits were reduced to twice per year, following introduction of novel auto-lure
pumps. Tracking indices for rats remained low after six months, then increased to 37% in May 2017. That increase, as
well as small fluctuations in measured activity levels throughout the study, could indicate continued incursion from the
mainland, highlighting the importance of continued suppression. Additional work is needed to determine the limitations
of the automatic lure dispensers and optimise their use for long-term suppression of pest mammals in ecosystems that are
highly vulnerable to re-invasion.
Keywords: conservation, invasion biology, invasive mammals, island biosecurity, Norway rat (Rattus norvegicus),
Pacific rat (Rattus exulans), ship rat (Rattus rattus)
INTRODUCTION
Introduced mammals are one of the most significant
threats to island ecosystems (Towns, et al., 2006;
Bellingham, et al., 2010; Harper & Bunbury, 2015). In
particular, rats (Rattus spp.) and other rodents have become
major predators of endemic island species, causing several
local extinctions (Courchamp, et al., 2003; Towns, et al.,
2006; Bellingham, et al., 2010). Thus, they are a main target
of eradication operations (Howald, et al., 2007; Glen, et
al., 2013; Holmes, et al., 2015). However, while numerous
offshore rat eradications have been undertaken successfully
since the 1980s, eradication is more difficult in locations
that are close enough to a non-controlled pest population
to facilitate rapid, and inevitable, re-invasion (Russell,
et al., 2008; Nathan, et al., 2015). At highly re-invadable
sites, such as near-shore islands, a single operation can
theoretically eliminate a population of invasive rats.
However, that ‘eradication’ is only temporary. Sustained
control is required in order to prevent re-establishment
(Simberloff, 2011), which can occur rapidly and with only
a few invaders (Russell, et al., 2008; Nathan, et al., 2015).
Most successful eradication operations on New
Zealand islands – both of rats and of other pest mammals
– have been undertaken using site-wide toxicant
applications (Blackie, et al., 2013; Keitt, et al., 2015).
However, toxicant-based methods are not as effective
for sustained control in highly re-invadable sites as they
are on relatively isolated, offshore islands. Importantly,
a re-invading population of mammals has to achieve a
minimum density in order for repeated toxicant use to be
considered a cost-effective means of control (Warburton &
Thomson, 2002), but that density threshold is higher than
the maximum density under which many native species can
successfully recover (Gillies, et al., 2003; Norbury, et al.,
2015). Thus, conservation-motivated, long-term mammal
suppression in re-invadable sites requires the availability
of sustained-use, cost-effective methods. Throughout this
paper, we use the terms ‘suppression’ and ‘maintenance
control’ interchangeably to refer to any control method
used to prevent the re-establishment of a population of pest
mammals in an island due to incursion. However, the same
principles can be applied within any controlled area that is
at risk of being invaded, or re-invaded, from an adjacent,
un-controlled population.
Unlike mammal control operations that rely on sitewide application of toxicants, traps can be left in situ and
used for incursion control. However, current best-practice
methods of trapping require continual re-baiting and, if
a trap is triggered, re-setting of the trapping mechanism
to remain effective (DOC, 2006). In addition, traps may
be less effective at controlling low-density populations
than they are at eradication of established, high-density
populations (Thorsen, et al., 2000; Chappell, 2004). As
a result, effective island biosecurity still requires regular
surveillance and the availability of funding to undertake
contingency response in the event of an incursion
(Russell, et al., 2008). A relatively new technique for
long-term control of invasive mammal populations is the
use of automatic, or self-resetting, trapping and toxicant
application mechanisms (reviewed in Campbell, et al.,
2015). Like single-use trapping methods, self-resetting
mechanisms – both traps and toxicant-delivery devices –
can be designed with relatively high specificity, reducing
the rate of non-target kills, relative to that realised
following site-wide toxicant applications (Campbell, et al.,
2015). Unlike single-use traps, automatic mechanisms can
remove multiple pests before requiring maintenance and/
or re-baiting (Blackie, et al., 2011; Blackie, et al., 2013;
Murphy, et al., 2014; Carter, et al., 2016).
Automatic toxicant-delivery devices have been
designed for stoats (Mustela ermina), possums (Trichosurus
vulpecula) (Blackie, et al., 2016) and rats (Blackie, et al.,
2013; Murphy, et al., 2014). Automatic, toxicant-free traps
and corresponding long-life lures have been developed
by Goodnature® Ltd for possums, rats, and stoats (Carter,
et al., 2016; Carter & Peters, 2016), with the advantage
that devices that do not rely on poisons may be more
acceptable for control of invasive mammals in locations
that support populations of native mammals (Campbell, et
al., 2015). Self-resetting traps have been used to control
sympatric populations of Norway rats (Rattus norvegicus),
ship rats (R. rattus), and Australian brushtail possums
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 187–190. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
187
Island invasives: scaling up to meet the challenge. Ch 1C Rodents: Lessons
on a single, near-shore island (Carter, et al., 2016) and
to control ship rats and mice (Mus musculus) within an
unprotected mainland site (Carter & Peters, 2016) in New
Zealand. One pest control operation in Hawaii also found
that automatic traps were more beneficial for decreasing
predation of native species by rats than single-action snaptraps (Franklin, 2013).
The long-term financial costs of using automatic traps
for control of invasive mammal populations are comparable
to those of using basic Victor® snap-traps, especially when
work is undertaken primarily by contractors, and slightly
lower than the costs of using DOC-200 traps, heavy-duty,
single-action tunnel traps commonly used for maintenance
control (Carter, et al., 2016; Carter & Peters, 2016). The
use of self-resetting traps greatly reduces the frequency at
which site visits must be undertaken, relative to traditional
methods of trapping that require regular rebaiting and
resetting to maintain effectiveness (e.g., Franklin, 2013).
However, the rate at which even long-life lures must be
replenished in self-resetting traps – approximately monthly
– is still higher than the rate at which pests are killed,
following initial suppression of the population (Carter, et
al., 2016). As a result, the costs of long-term suppression
of pest mammals – in terms of both equipment and personhours – are increased significantly by the investment in
on-the-ground trap maintenance, even when self-resetting
traps are used (Franklin, 2013; Glen, et al., 2013; Carter, et
al., 2016; Carter & Peters, 2016).
The continued effectiveness of self-resetting traps relies
largely on maintaining attractiveness of the highly viscous
lure, which is contained within a plastic bottle housed
inside the trap. When a targeted mammal population is
relatively dense and the lure is consumed regularly, the
force of gravity is sufficient to keep ‘fresh’ lure available.
Once a population of invasive mammals has been knocked
down, human intervention is required to ensure that uneaten lure does not become mouldy and unpalatable after
being exposed to air. Thus, the mechanism of lure delivery
itself remains a barrier to minimising costs of maintenance
control. Here, we tested the use of auto-dispensing lure
pumps for retaining lure freshness and maintaining low
levels of rats on a previously-controlled inshore island,
while significantly reducing the person-hours required for
undertaking site visits.
with novel auto-lure pumps. The auto-lure pump is a
soft-sided lure bottle that uses hydrogen gas expansion to
deliver 55 g of non-toxic lure over a period of six months
(Fig. 1). The CO2 cartridges and auto-lure pumps were
replaced every six months.
During the initial control operation only, we used
tracking tunnels (Pest Control Research [PCR] Ltd.,
Christchurch, New Zealand) with inked tracking cards
(Black Trakka®, Gotcha Traps, Auckland, New Zealand)
to monitor mammal activity within the trapping grid on
Native Island and at a control site, located 3.5 km away on
Stewart Island (Gillies & Williams, 2013). We estimated
rat activity using tracking indices, with detection corrected
for interference with the tracking cards by possums, where
required (Gillies & Williams, 2013). Tracking tunnels
were installed at 50 m intervals on Native Island in six
lines of five tunnels each, and at the control site in three
lines of five tunnels and two lines of ten tunnels. During
each monitoring period, tracking tunnels were baited with
peanut butter and tracking cards retrieved after 24 hours
(Carter, et al., 2016). Following installation of the autolure pumps, rat activity was monitored at two subsequent
intervals of approximately six months, at the Native Island
site only.
RESULTS
During the initial control operation, tracking indices
for rats on Native Island decreased from 73% to 7%
within nine months of initiation of trapping and remained
perpetually at or below 10% during the monitoring
MATERIALS AND METHODS
In November 2013, we installed 143 CO2-powered,
automatic rat traps (A24; Goodnature® Ltd, Wellington,
New Zealand; https://www.goodnature.co.nz) on a 100 m
× 50 m grid on Native Island (46°54’54″ E 168°09’25″ S)
(Carter, et al., 2016), a mostly forested, 62 ha Scenic Reserve
within Rakiura National Park in southern New Zealand
(DOC, 2012). Because Native Island sits approximately
100 m from the coast of the main island of Stewart Island
(also known as Rakiura), incursion by multiple species
of pest mammals from the mainland following a control
operation is inevitable (Atkinson, 1986). The presence
of Norway rats, ship rats and brushtail possums has been
confirmed on Native Island (DOC, 2012), and all three
species were observed during establishment of the trapping
network. In addition, Pacific rats (kiore, Rattus exulans) are
present on the nearby mainland and may pose an additional
incursion risk.
Each trap was initially baited with a bottle of non-toxic,
peanut-based lure and checked approximately monthly,
with lure bottles and CO2 cartridges replenished every six
months (Carter, et al., 2016). Due to lack of resources, the
traps were not maintained for the eight months between
September 2015 and May 2016. In May 2016, we replaced
all CO2 cartridges and replaced the standard lure bottles
188
Fig. 1 Diagram of an (a) auto-lure pump and (b) deconstructed interior of an A24 self-resetting rat trap.
Activation of the trigger by a rat as it accesses the lure
causes rapid deployment of the CO2-powered piston,
which strikes the skull and results in near-instantaneous
death. The trap automatically resets after each strike,
up to 24 times. Gradual expansion of hydrogen gas
inside the soft-sided, auto-lure pump slowly delivers lure
through the bottle opening over a period of six months.
Image courtesy of Goodnature® Ltd (Wellington, New
Zealand).
Carter, et al.: Auto-dispensing lure pumps
period, while tracking indices at the control site remained
comparatively high (Fig. 2; see Carter, et al., 2016 for
complete results). On the first monitoring visit following
installation of the auto-lure pumps, tracking indices on
Native Island were 7% but increased to 37% during the
most recent site visit in May 2017 (Fig. 2). Rat activity was
not monitored at the control site after the initial trapping
operation. However, tracking indices within a separate, untrapped area on Stewart Island were 40% in March 2017
(SIRCET, 2017). Between the first and second monitoring
visits, air temperatures were between -1.5°C and 0.5°C of
monthly regional (Southland) averages, varying between
8°C and14°C during the study months (Macara, 2013), and
rainfall levels were at or below normal levels (Fig. 2).
DISCUSSION
This project was the first in situ test of auto-dispensing
lure bottle technology, following a previous knockdown.
One of the primary motivations for developing timesaving technologies for invasive mammal control – lack
of sufficient available person-hours for maintaining traps
and monitoring for incursions – was both the impetus for
and the main limitation of this study. Because rat activity
levels were not monitored for the year prior to installation
of the auto-lure pumps, nor when they were installed,
we cannot say definitively that they were as effective as
standard lure bottles at maintaining low levels of rats. That
is, the activity levels recorded in October 2016 could be
indicative of no incursions, with switching of standard
Fig. 2 Summary of monitoring data from tracking tunnels
on Native Island (black) and the control site on Stewart
Island (grey), with the introduction of auto-lure pumps
indicated by the dashed vertical line. Except for the
period from May 2015 – May 2016, the spacing of x-axis
labels is proportional to the amount of time between
monitoring dates. The percentage of tracking cards
with interference by possums was high at the control
site on Stewart Island throughout the initial trapping
operation, so true activity of rats may be higher than
indicated by the plot, especially in February and May of
2015. No data are available for the period May 2015 –
October 2016, and no monitoring was undertaken at the
control site after May 2015. Climate data are shown with
dashed lines. The rainfall axis shows the approximate
percentage of local rainfall in each month, relative to
‘normal’ conditions (i.e., a value of 100% is equal to
normal). The temperature axis shows the deviation of
local air temperatures from average conditions, with
a value of 0 equal to the respective monthly mean.
Weather information was obtained from NIWA ‘Current
climate’ monthly summaries (https://www.niwa.co.nz/
climate/nzcu/). Plot adapted from Carter, et al. (2016).
bait bottles for auto-lure pumps having no effect on the
consistently low activity levels observed since at least
August 2014. However, given the proximity of the study
site to uncontrolled populations of multiple rat species, as
well as fluctuating activity levels throughout the original
control operation and slight increase observed in May
2015, that activity levels were still below 10% a year later
is unlikely. More likely is that rat activity levels increased
to some extent prior to installation of the auto-lure pumps
and that the pumps effectively reduced activity levels
during the first five months of their operation.
During the second, but not the first, monitoring visit
to Native Island, the lure was noticeably mouldy and may
have been less attractive to rats. Mould growth may be
related to environmental conditions at the study site, which
would suggest that the rate of gas expansion inside the
auto-lure pump may be insufficient to keep the lure fresh
in certain conditions. Climate has been implicated in the
failure of mammal control operations across methods, with
stationary bait stations being most similar to trapping. Bait
station-based eradication failures have been associated
with higher mean annual temperatures and increased
variation in inter-annual precipitation in tropical locations,
which become more important with increasing island size
(Holmes, et al., 2015). High temperatures, in particular,
are a significant predictor of failure across toxicant-based
methods of rat eradication (Holmes, et al., 2015).
The importance of climate to the success of mammal
control has been examined primarily in relation to the
timing of toxicant application, particularly in the tropics,
where more consistent food availability increases the
difficulty of targeting rodents using attractant-based
toxicants (Holmes, et al., 2015; Russell & Holmes, 2015).
Air temperatures at our study location did not vary much
from average monthly conditions, and more months were
relatively ‘dry’ than ‘wet,’ compared with regional norms
(Fig. 2). However, further research should be undertaken
to determine whether abiotic environmental conditions
constrain the efficacy of auto-lure pumps. If so, either
(1) increasing the rate of gas expansion inside the autolure bottle or (2) increasing the rate of site visits during
particular seasons or in climates normally conducive to
mould growth may be required.
Assuming the number of successful control operations
in incursion-vulnerable sites increases, so too will the
costs of controlling invasive mammals. In highly reinvadable sites, true eradication is an impractical aim
(Simberloff, 2011). Indeed, if mammal densities can be
maintained at levels low enough to facilitate the recovery
of native populations, then eradication becomes less
immediately imperative. Thus, cost-effective suppression
of pest mammals is a realistic goal for conservation of
endemic island biodiversity. Technologies that minimise
the time and financial investments required for long-term
control will be key to maximising the area within which
populations of invasive mammals can be controlled. More
work is needed to optimise the use of auto-lure pumps and
quantify their limitations. However, effective automatic
lure delivery devices would be a transformative addition
to the pest-control toolbox and should continue to be
rigorously developed and tested.
ACKNOWLEDGEMENTS
Thanks to field volunteers from the East Taranaki
Environment Trust; Nick Beckers, Mark Caskey, Joanna
Greig, Bob Schumacher and Karen Schumacher; and Dean
Caskey from the Taranaki Regional Council. We also thank
staff from the New Zealand Department of Conservation
Te Papa Atawhai, especially Klaartje van Schie, and two
landowners for permission to install traps in the privately189
Island invasives: scaling up to meet the challenge. Ch 1C Rodents: Lessons
owned areas of Native Island. We would also like to
acknowledge the contributions of current and former
Goodnature staff. Funding for this project was provided by
the Department of Conservation.
REFERENCES
Atkinson, I.A.E. (1986). ‘Rodents on New Zealand’s Offshore Islands:
Distribution, Effects and Precautions Against Further Spread’. In: A.E.
Wright and R.E. Beever (eds.) The Offshore Islands of Northern New
Zealand, pp.13–40. Information Series 14. Wellington: New Zealand
Department of Lands and Survey.
Bellingham, P.J., Towns, D.R., Cameron, E.K., Davis, J.J., Wardle, D.A.,
Wilmshurt, J.M. and Mulder, C.P.H. (2010). ‘New Zealand island
restoration: Seabirds, predators, and the importance of history’. New
Zealand Journal of Ecology 34: 115–136.
Blackie, H.M., MacKay, J.W.B., Allen, W.J., Smith, D.H.V., Barrett,
B., Whyte, B.I., Murphy, E.C., Ross, J., Shapiro, L., Ogilvie, S., Sam,
S., MacMorran, D., Inder, S. and Eason, C.T. (2013). ‘Innovative
developments for long-term mammalian pest control’. Pest Management
Science 70: 345–351.
Blackie, H., MacKay, J., Barrett, B., Inder, S., MacMorran, D., Bothwell,
J., Clout, M. and Eason, C. (2016). A novel device for controlling
brushtail possums (Trichosurus vulpecula). New Zealand Journal of
Ecology 40: 60–64.
Blackie, H., Woodhead, I., Diegel, H., MacMorran, D., Shapiro, L.,
Murphy, E. and Eason, C. (2011). ‘Integrating ecology and technology
to create innovative pest control devices’. In: J. Jacob and A. Esther
(eds.) Proceedings of the 8th European Vertebrate Pest Management
Conference, pp. 152–153. Berlin: Julius Kühn-Institut.
Campbell, K.J., Beek, J., Eason, C.T., Glen, A.S., Godwin, J., Gould, F.,
Holmes, N.D., Howald, G.R., Madden, F.M., Ponder, J.B., Threadgill,
D.W., Wegmann, A.S. and Baxter, G.S. (2015). ‘The next generation
of rodent eradications: innovative technologies and tools to improve
species specificity and increase their feasibility on islands’. Biological
Conservation 185: 47–58.
Carter, A.C., Barr, S., Bond, C., Paske, G., Peters, D. and van Dam,
R. (2016), ‘Controlling sympatric pest mammal populations in New
Zealand with self-resetting toxicant-free traps: A promising tool for
invasive species management’. Biological Invasions 18: 1723–1736.
Holmes, N.D., Griffiths, R., Pott, M., Alifano, A., Will, D., Wegmann, A.S.
and Russell, J.C. (2015). ‘Factors associated with rodent eradication
failure’. Biological Conservation 185: 8–16.
Howald, G., Donlan, C.J., Galván, J.P., Russell, J.C., Parkes, J.,
Samaniego, A., Wang, Y., Veitch, D., Genovesi, P., Pascal, M., Saunders,
A. and Tershy, B. (2007). ‘Invasive rodent eradication on islands’.
Conservation Biology 21: 1258–1268.
Keitt, B., Griffiths, R., Boudjelas, S., Broome, K., Cranwell, S., Millett,
J., Pitt, W. and Samaniego-Herrera, A. (2015). ‘Best practice guidelines
for rat eradication on tropical islands’. Biological Conservation 185:
17–26.
Macara, G.R. (2013). ‘The Climate and Weather of Southland’. NIWA
Science and Technology Series 63. Wellington: National Institute of
Water and Atmospeheric Research.
Murphy, E., Sjoberg, T., Barun, A., Aylett, P., MacMorran, D. and Eason,
C. (2014). ‘Development of Re-setting Toxin Delivery Devices and
Long-life Lures for Rats’. In: R.M. Timm and J.M. O’Brien (eds.)
Proceedings of the 26th Vertebrate Pest Conference, pp. 396–399.
Davis: University of California.
Nathan, H.W., Clout, M.N., MacKay, J.W.B., Murphy, E.C. and Russell,
J.C. (2015). ‘Experimental island invasion of house mice’. Population
Ecology 57(2): 363–371.
Norbury, G.L., Pech, R.P., Byrom, A.E. and Innes, J. (2015). ‘Density
impact functions for terrestrial vertebrate pests and indigenous biota:
guidelines for conservation managers’. Biological Conservation 191:
409–420.
Russell, J.C. and Holmes, N.D. (2015). ‘Tropical island conservation: rat
eradication for species recovery’. Biological Conservation 185: 1–7.
Russell, J.C., Towns, D.R., Beaven, B.M., MacKay, J.W.B. and Clout,
M.N. (2008), ‘Testing island biosecurity systems for invasive rats’.
Wildlife Research 35: 215–221.
Simberloff, D. (2011). ‘Today Tiritiri Matangi, tomorrow the World! Are
we aiming too low in invasives control?’ In: C.R. Veitch, M.N. Clout
and D.R. Towns (eds.) Island invasives: eradication and management,
pp. 4–12. Occassional Paper SSC no. 42. Gland, Switzerland: IUCN
and Auckland, New Zealand: CBB.
SIRCET. (2017) SIRCET Rat Monitoring Results, 25th May 2017,
Unpublished report. Stewart Island, NZ: Stewart Island / Rakiura
Community and Environment Trust.
Carter, A.C. and Peters, D. (2016). ‘Self-resetting Traps Provide Sustained,
Landscape-scale Control of a Rat Plague in New Zealand’. In: R. M.
Timm and R. A. Baldwin (eds.) Proceedings of the 27th Vertebrate Pest
Conference, pp. 308–313. Davis, CA: University of California.
Thorsen, M., Shorten, R., Lucking, R. and Lucking, V. (2000). ‘Norway
rats (Rattus norvegicus) on Frégate Island, Seychelles: The invasion,
subsequent eradication attempts and implications for the island’s fauna’.
Biological Conservation 96: 133–138.
Chappell, R. (2004). Motutapere Island Rodent Invasion 2004. Internal
report, file ISL 0013. Coromandel, New Zealand: Department of
Conservation.
Towns, D.R., Atkinson, I.A.E. and Daugherty, C.H. (2006). ‘Have
the harmful effects of introduced rats on islands been exaggerated?’.
Biological Invasions 8: 863–891.
Courchamp, F., Chapuis, J.-L. and Pascal, M. (2003). ‘Mammal invaders
on islands: impact, control and control impact’. Biological Reviews 78:
347–383.
Warburton, B. and Thomson, C. (2002). ‘Comparison of three methods
for maintaining possums at low density’. Science for Conservation. 189.
Wellington, New Zealand: Department of Conservation.
DOC [Department of Conservation, New Zealand] (2006). Island
Biosecurity Best Practice Manual. Wellington, New Zealand:
Department of Conservation.
DOC [Department of Conservation] (2012). Stewart Island / Rakiura
Conservation Management Strategy and Rakiura National Park
Management Plan 2011–2021. Invercargill, New Zealand: Department
of Conservation.
Franklin, K. (2013). Informational Report on the Use of Goodnature A24
Rat Traps in Hawaii. Unpublished report. Oahu, HI: Pacific Cooperative
Studies Unit, Research Corp. of the University of Hawaii, Oahu Army
Natural Resources Program.
Gillies, C.A., Leach, M.R., Coad, N.B., Theobald, S.W., Campbell, J.,
Herbert, T., Graham, P.J. and Pierce, R.J. (2003). ‘Six years of intensive
pest mammal control at Trounson Kauri Park, a Department of
Conservation ‘‘mainland island’’, June 1996–July 2002’. New Zealand
Journal of Zoology 30: 399–420.
Gillies, C.A. and Williams, D. (2013). DOC tracking tunnel guide v2.5.2:
using tracking tunnels to monitor rodents and mustelids. Unpublished
report. Hamilton, New Zealand: Science and Capability Group.
Glen, A.S., Atkinson, R., Campbell, K.J., Hagen, E., Holmes, N.D.,
Keitt, B.S., Parkes, J.P., Saunders, A., Sawyer, J. and Torres, H. (2013),
‘Eradicating multiple invasive species on inhabited islands: The next
big step in island restoration?’. Biological Invasions 15: 2589–2603.
Harper, G.A. and Bunbury, N. (2015). ‘Invasive rats on tropical islands:
Their population biology and impacts on native species’. Global
Ecology and Conservation 3: 607–627.
190
J.L. Herrera-Giraldo, C.E. Figuerola-Hernández, N.D. Holmes, K. Swinnerton, E.N. Bermúdez-Carambot, J.F. González-Maya and D.A. Gómez-Hoyos
Herrera-Giraldo, J.L.; C.E. Figuerola-Hernández, N.D. Holmes, K. Swinnerton, E.N. Bermúdez-Carambot, J.F. González-Maya and D.A. Gómez-Hoyos.
Survival analysis of two endemic lizard species before, during and after a rat eradication attempt on Desecheo Island, Puerto Rico
Survival analysis of two endemic lizard species before, during and
after a rat eradication attempt on Desecheo Island, Puerto Rico
J.L. Herrera-Giraldo1, C.E. Figuerola-Hernández1, N.D. Holmes1, K. Swinnerton1,2, E.N. Bermúdez-Carambot3,
J.F. González-Maya4 and D.A. Gómez-Hoyos4
Island Conservation, 2100 Delaware Avenue, Suite 1, Santa Cruz, CA 95060, USA.
<jose.herrera@islandconservation.org>. 2Current affiliation: The Island Endemics Foundation, P.O. Box 1908,
Boquerón, Puerto Rico 00622. 3US Fish and Wildlife Service -Vieques National Wildlife Refuge, Vieques, Puerto Rico.
4
ProCAT International/The Sierra to Sea Institute, Las Alturas, Puntarenas, Costa Rica.
1
Abstract Rodent eradications are a key island restoration activity to counteract extinction and endangerment to native
species. Despite the widespread use of brodifacoum as a rodenticide for island restoration, there has been little examination
of its potential negative effects on native reptiles. Here we examined the survival of two endemic insular lizard populations
before, during and after a brodifacoum-based rodent eradication using a mark-recapture study. We found no evidence of an
effect from baiting in Anolis desechensis and evidence of a change in recapture rates after baiting for Ameiva desechensis.
Effects of baiting on survival rates were not measurable due to a small sample size. Results suggest that brodifacoum
did not result in population-level impacts during the three-week study period after brodifacoum exposure. For invasive
species eradications using toxicants, potential risks to non-target species should be assessed against the expected benefits
to native biota from the removal of threats posed by invasive mammals. We recommend continued studies that directly
examine non-target risk to native reptile populations derived from toxicant baiting programs, particularly on tropical
islands that are home for high numbers of endemic reptiles.
Keywords: brodifacoum, mark-recapture, non-target species, reptiles, rodent eradication
INTRODUCTION
Islands represent approximately 5% of the land area
of the Earth, yet 61% of extinctions have been insular
species, and 37% of species listed by the IUCN as critically
endangered are confined to islands (Tershy, et al., 2015).
Invasive species are a major driver of species extinctions on
islands and remain a significant risk to threatened species
(Bellard, et al., 2016; Doherty, et al., 2016). Invasive
rats have been introduced to approximately 80% of the
archipelagos of the world, and have wide-ranging negative
impacts on native flora and fauna (Towns, et al., 2006).
Techniques to eradicate invasive rodents from islands are
available and the practice is increasing in scope, scale,
and application (Howald, et al., 2007; Keitt, et al., 2011),
with restoration benefits being accrued when eradication is
achieved (Jones, et al., 2016). To date there have been over
650 eradication attempts of rats (Rattus spp.) on more than
500 islands worldwide (Russell & Holmes, 2015).
Successful rodent eradication from islands larger than
5 ha primarily relies on the use of anticoagulant rodenticide
(Howald, et al., 2007). Second generation anticoagulants
are the most commonly used toxicant in invasive rodent
eradication programmes (Holmes, et al., 2015). When
using toxicants for rodent eradication on islands, the risk
to non-target native species is typically assessed. Efforts to
reduce this risk during eradication operations commonly
include application of bait when susceptible species are
absent, temporary captive-holding of species during
potential periods of exposure, and alternative delivery
methods to reduce bait access (Howald, et al., 2007). While
reptiles have been known to consume cereal-based rodent
baits (Merton, 1987; Marshall & Jewell, 2007), they have
typically been considered at lower risk (Hoare & Hare,
2006), in part because of decreased susceptibility due to
differences in blood chemistry and physiology compared
to mammals and birds (Merton, 1987; Hoare & Hare,
2006). Although evidence of population level impact to
reptiles is sparse, observations from an increasing number
of rodenticide-based eradications, plus targeted studies,
have suggested the risk is low (Harper, et al., 2011).
Nevertheless, additional studies are required to improve
general knowledge of the risk of rodenticides to reptiles
during rodent eradication operations.
During 2012, an eradication of Rattus rattus using
rodenticide bait was attempted on Desecheo Island located
approximately 21 km off the north-west coast of Puerto
Rico. Black rats were introduced in the early 1900s and
are considered an important threat on Desecheo, including
impacts on native reptiles from direct predation and habitat
modification via seed and seedling predation and soil
nutrient changes (U.S. Fish and Wildlife Service, 2016).
An additional potential threat from rats to native reptiles
could include competition for space and food resources,
consistent with rat impacts on reptiles elsewhere (Shiels,
et al., 2014; Harper & Bunbury, 2015). Two years prior to
the eradication operation, exposure of bait to the endemic
Desecheo ameiva (Ameiva desechensis) and Desecheo anole
(Anolis desechensis) was assessed through a placebo nontoxic bait biomarker study. The study found no evidence
of ameivas (n=18 marked, n=5 recaptured) interacting
with bait, but 21% of anoles recaptured were exposed
(n=97 marked, n=20 recaptured) (Herrera & BermúdezCarambot, 2010). However, because these species occur
only on Desecheo, and thus had high conservation value,
the fate of both lizard species was followed during the
application of toxic bait during the eradication operation.
Here we report the results of a mark-recapture study to
monitor the short-term survival of the ameivas and the
anoles before, during and after the 2012 rodent eradication
operation on Desecheo Island.
MATERIALS AND METHODS
Study area
Desecheo Island is a 117 ha hilly island located
approximately 21 km off the north-west coast of the
Commonwealth of Puerto Rico (18o 23’ N, 67o 29’ W; Fig.
1). It was declared a U.S. National Wildlife Refuge (NWR)
in 1976 and is currently administered and managed by the
U.S. Fish and Wildlife Service. Sub-tropical dry forest (i.e.
woodland) is present primarily on the leeward slopes and
valleys, and is dominated by the semi-deciduous almácigo
tree (Bursera simaruba). The windward slopes and ridges
also harbour cacti, shrubs and open grasslands. The annual
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 191–195. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
191
Island invasives: scaling up to meet the challenge. Ch 1C Rodents: Lessons
the island, but is most common in forested areas (e.g.
valleys) and their margins (Evans et al., 1991). Average
size for adult males is 57 mm (SVL) and for females 45
mm (SVL). Field surveys in 2009 and 2010 estimated the
island population at 52,111 individuals (95% CI 31,464–
72,758) (McKown et al., 2010).
Reptile monitoring
Fig. 1 Location of Desecheo Island and sampling sites
for Anolis desechensis and Ameiva desechensis impact
assessment during black rat eradication operations.
During the eradication, we implemented a reptile
monitoring program between February and April 2012. We
used a standard mark-recapture methodology (Jolly, 1965;
Seber, 1965) over three discrete sampling periods of six
days each, which coincided with bait application stages
during the eradication. The first period began 21 days
prior to the first bait application, the second between the
first and second bait application, and the third began three
days after the second bait application. The sampling sites
were randomly located in five different locations within the
woodland habitats in the Long and West Valleys (Fig. 1).
Ameivas were sampled within one 100 × 10 m plot and
rainfall average is 828 mm (range 750–1039 mm; Morrison
& Menzel, 1972) with a seasonal dry period between
January and March, followed by a rainy season between
July and November. The island supports five singleisland endemic species (three lizards and two arachnids)
as well as one of the three remaining populations of the
threatened higo chumbo cactus (Harrisia portoricensis).
Previous anthropogenic activities on the island included
livestock grazing, military operations (e.g. bombing and
gunnery range) and the introduction of invasive mammals:
black rats (Rattus rattus), goats (Capra hircus), feral cats
(Felis catus), and rhesus macaques (Macaca mulatta). The
extirpation of nesting seabirds from Desecheo Island has
been linked to the presence of these invasive mammals
(U.S. Fish and Wildlife Service, 2016). The island is
currently closed to the public due to the existence of
unexploded ordnance.
Rat eradication
Aerial bait broadcast for rodent eradication was carried
out on Desecheo between March 13 and 23, 2012. The bait
used for the eradication was “Brodifacoum Conservation25D” manufactured by Bell Laboratories in Madison,
Wisconsin, USA. The bait was a 2 g extruded pellet, dyed
green, and contained 25 ppm of the toxin brodifacoum. The
bait broadcast was completed in two aerial applications
separated by 10 days and with a ground application rate
of 17 kg/ha for the first application and 9.1 kg/ha for
the second application. There is no weather station on
Desecheo Island and data were obtained from weather
stations located in Rincon (13 miles from Desecheo) and
Isabela (29 miles) and the Standard Precipitation Index
(SPI) produced by Caribbean Regional Climate Center.
January and March are usually a dry period but data from
two weather stations and comparisons with 2008 and 2010
vegetation cover indicate that in 2012 Desecheo received
greater than average rainfall.
Study species
The Desecheo ameiva (Fig. 2a) is a common lizard
species found in coastal areas, including shoreline margins,
in habitats of maximum solar exposure but frequently
near some vegetation cover or shade (Evans, et al., 1991).
Adult females tend to be smaller (SVL <90 mm) than
males (average SVL 97 mm). Field surveys conducted in
2009 and 2010 estimated the island population at 7,469
individuals (95% CI 1,800–13,137) (McKown, et al.,
2010). The Desecheo anole (Fig. 2b) is present throughout
192
Fig. 2 Ameiva desechensis (a) and Anolis desechensis (b).
Herrera-Giraldo, et al.: Survival of two lizard species
two 50 × 10 m plots. Anoles were sampled within two 100
× 10 m plots. Each plot was surveyed by two observers,
each responsible for sampling one side (5 m) of a central
transect through the plot. Each sampling day accounted for
8 hours of intensive searches for both species, and included
the detection and capture of each observed individual.
Individuals were captured using a pole and noose and by
hand capture. Each anole was marked on the hind limb
with a unique visible alphanumeric implant tag and each
ameiva was marked with a unique combination of coloured
glass beads sewed to the base of the tail (Fig 2a), and a
unique combination of clipped toes (Censky, pers. comm.
and modified from Fisher & Muth, 1989). Each individual
was released at their capture location.
Statistical analyses
Survival of individuals was estimated using a markrecapture model based on multiple capture histories
within each sampling period (Cooch & White, 2015).
We estimated the probability of recapture based on time
and apparent survival to assess any potential impacts on
either species as a result of the rodent bait application.
We used MARK 5.0 (White & Burnham, 1999) to model
factors influencing variation in survival. The CormackJolly-Seber (CJS) model based on live animal recaptures
in an open population (Lebreton, et al., 1992) was used
to estimate the apparent survival (phi or ø). Models were
constructed based on the recapture rates (p) and apparent
survival (ø) remaining constant (.) or changing in time (t),
and according to the bait dispersal events – before, during,
and after (asp). The best performing model was selected
using the Akaike Information Criteria (AIC) through the
proportion test with Akaike weights (AICw; Burnham &
Anderson, 2002). The assumptions of the CJS model were
tested using TEST 2 and TEST 3 in the U-CARE program
version 2.3 M 7.5 (Choquet, et al., 2005). To evaluate the
fit of the set of models to the data, a Global TEST was
conducted to calculate the variance inflation factor (ĉ).
Fig. 3 Apparent survival percentage of Anolis desechensis
and Ameiva desechensis during a black rat eradication
on Desecheo Island (Error bars: 95% confidence
intervals).
RESULTS
A total of 452 anoles and 57 ameivas were captured
and marked across 18 days of field sampling and 144
person-hours of sampling effort in the five study sites
(Table 1). Although ameivas were detected less frequently
across the study sites, they had a higher rate of recapture
(35 recaptures, 61.4%) than anoles (92 recaptures, 20.4%;
Table 1).
The best supported model for anoles explained the
probability of recapture according to time and with
apparent survival remaining constant (Table 2). For
ameivas, the best supported model was the one in which
the recapture probability varied across the sampling
periods (i.e. bait application) and when apparent survival
remained constant (Table 2). Both models indicated no
changes in apparent survival along the three periods (asp)
of bait applications. TEST 2 and TEST 3 showed no
differences in the probability of recaptures and survival for
the marked individuals (p>0.61). Global TEST indicated
a sub-dispersion in the data (ĉ<1), thus no effect on the
Fig. 4 Recapture probability for (a) Anolis desechensis
and (b) Ameiva desechensis before, during and after
bait dispersal for a black rat eradication on Desecheo
Island (Error bars: 95% confidence intervals). We retain
individual survey events in Figure 3b as these were found
to be associated with recapture probability.
variance, therefore this parameter was not modified in the
models (Cooch & White, 2015).
Apparent survival for both lizard species during the
study period was estimated to be time-invariant and close
to 100% (anoles: ø = 0.99, 95% CI = 0.91–0.99; ameivas: ø
= 1, 95% CI = 1–1; Fig. 3). However, the precise apparent
Table 1 Anolis desechensis and Ameiva desechensis previously unmarked and accumulated recaptures
(in parenthesis) according to sampling site and bait application stage during the black rat eradication
on Desecheo Island.
Bait application stage
Before
During
After
Anolis desechensis
Total
Site1
Site 2
215(13)
89(2)
126(11)
95(43)
49(6)
46(37)
142(92)
75(24)
67(68)
Ameiva desechensis
Total
Site1
Site2
Site3
27(7)
7(0)
12(3)
8(4)
16(14)
8(3)
5(7)
3(4)
14(35)
7(13)
3(13)
4(9)
193
Island invasives: scaling up to meet the challenge. Ch 1C Rodents: Lessons
Table 2 Comparison of models to estimate the apparent survival (ø) and probability of recapture (p), according
to the bait application stage (asp: before, during and after) and time (t) for Anolis desechensis and Ameiva
desechensis during black rat eradication operations at Desecheo Island.
Model
Anolis desechensis
ø(.) p(t)
ø(asp) p(t)
ø(.) p(.)
ø(asp) p(.)
ø(asp) p(asp)
ø(.) p(asp)
ø(t) p(.)
ø(t) p(asp)
ø(t) p(t)
Ameiva desechensis
ø(.) p(asp)
ø(.) p(.)
ø(asp)p(asp
ø(asp) p(.)
ø(.) p(t)
ø(asp) p(t)
ø(t) p(.)
ø(t) p(asp)
ø (t) p(t)
AICc
ΔAICc
AICc weights
k
Deviance
888.04
891.79
893.33
895.79
896.42
897.23
915.12
915.68
918.92
0
3.76
5.29
7.76
8.39
9.19
27.09
27.64
30.89
0.788
0.120
0.059
0.016
0.012
0.008
0
0
0
18
20
2
4
6
4
18
20
33
249.89
249.33
288.51
286.92
283.46
288.36
276.98
273.22
247.56
268.91
270.44
273.47
274.67
287.42
294.05
311.42
312.15
348.35
0
1.52
4.55
5.76
18.51
25.13
42.50
43.23
79.44
0.615
0.287
0.063
0.034
<0.001
0
0
0
0
1
0
0
0
0
0
0
0
0
163.39
169.26
163.39
169.15
144.47
144.47
168.46
162.57
143.76
survival estimate for ameivas was not realistic due to the
small sample size (Fig. 3). An effect of time across bait
dispersal over the recapture probability was found in the
ameiva, with a tendency to decrease during and after bait
dispersal (Fig. 4a). In contrast, for the recapture probability
of the anoles there was no pattern associated with bait
dispersal, but this variation was related to survey events
(Fig. 4b). No mortality was observed for either species.
DISCUSSION
We estimated the survival and recapture rates of two
native reptile species during a black rat eradication on
Desecheo Island, Puerto Rico. During our study, we found
no significant change in apparent survival rates across the
sampling periods in anoles or ameivas, indicating that
the application of rodenticide bait did not result in any
detectable mortality or negative effect on both populations.
Furthermore, the recapture probabilities for anoles varied
through time (between survey events), but were not
dependent on bait application, suggesting that while anoles
were exposed to rodent bait (23% of individuals), exposure
did not impact survivorship within the sampling period.
For ameivas, the placebo-bait biomarker study found no
direct or indirect exposure of ameivas to rodent bait. For
the current study, the precise apparent survival estimate for
ameivas was influenced by the small sample size and was
not considered statistically valid. The recapture probability
estimate for the species decreased during bait application
and then increased following the bait application, which
may have been an artefact of increased human activity
during the operation affecting movement of these animals.
Behavioural ecology, diet, and foraging habitat of
lizards are important considerations in understanding
potential pathways of exposure to rodenticides. Although
we did not observe anoles or ameivas feeding directly on the
194
placebo biomarker or toxic bait, other studies have shown
direct consumption of bait by different reptile species
(Merton, 1987; Merton, et al., 2002). Bait availability
monitoring showed bait disappeared three days after the
second bait application, thus removing a pathway of direct
exposure (consumption) for ameivas and anoles. However,
we anticipate that anoles were exposed to bait via indirect
pathways through consumption of invertebrates feeding
on bait. Few anole species are dietary specialists and most
species, including the Desecheo anole, consume a wide
variety of insects and fruit (Meier & Noble, 1991). The
ameiva, a larger species than the anole, primarily forages
on the ground where it could be easily exposed to bait
through secondary pathways (e.g. ground-foraging beetles
and ants that feed on bait).
Delayed response to toxicant impacts on reptiles has
been previously reported. Telfair’s skinks (Leiolopisma
telfairii) on Round Island, Mauritius, showed an apparent
increased mortality three to six weeks after a brodifacoum
bait application (Merton, 1987) and Harper et al. (2011)
estimated 4.5% mortality of the Galápagos marine
iguana up to two months following rat eradication using
brodifacoum. While toxicant as the cause of death was
not confirmed during these events, a cautious approach
suggests it be considered a risk. While our study was
undertaken for approximately three weeks (22 days) after
bait was dispersed, the impacts of the rodenticide could not
be assessed beyond this timeframe.
This study focused on the survivorship of two reptile
species because of the high conservation value of these
single-island endemics. Rodenticide application risk
assessments should also consider the role of lizards as
prey items, and thus as potential toxin pathways to other
native species. Food web models that include rodenticide
introduction can inform risk assessments, including
Herrera-Giraldo, et al.: Survival of two lizard species
potential pathways and levels of exposure. Residue
analyses can help confirm these assessments. Ultimately,
risk assessments for rodent eradication operations using
toxicants must evaluate the cost and benefit impacts of
these efforts (i.e. negative impacts from using toxicants
versus positive impacts from removing rats). Whereas
reports of individual reptile mortality during rodenticidebased eradications are evident (Merton, 1987; Harper, et
al., 2011) a greater body of evidence suggests that reptile
populations benefit following rodent eradication (Jones,
et al., 2016). Studies such as ours provide another case
study to evaluate the value of island restoration efforts on
reptiles. The combination of studies such as these can help
managers make informed decisions about the potential
negative impacts of rodenticides used during eradication
operations versus the expected positive impact to native
biota from the permanent removal of threats posed by
invasive species.
ACKNOWLEDGEMENTS
This study was supported in part by the United States
Fish and Wildlife Service (USFWS), the National Fish
and Wildlife Foundation, the David and Lucile Packard
Foundation, and private donors to Island Conservation.
The views and conclusions contained in this document
are those of the authors and should not be interpreted
as representing the opinions or policies of the National
Fish and Wildlife Foundation. Mention of trade names or
commercial products does not constitute their endorsement
by the National Fish and Wildlife Foundation. The USFWS
Caribbean Islands National Wildlife Refuge Complex
provided logistical support and technical assistance during
the study period. Thanks to the editor and reviewers for
providing insightful comments and suggestions that
significantly improved this manuscript. We want to
express our deepest gratitude to the field assistants from
Island Conservation, USFWS and local biologists for their
dedication and commitment that made these conservation
efforts on Desecheo possible.
REFERENCES
Bellard, C., Cassey, P. and Blackburn, T.M. (2016). ‘Alien species as a
driver of recent extinctions’. Biology Letters 12: 20150623.
Burnham, K.P. and Anderson, D.R. (2002). Model Selection and Multi
Model Inference: A Practical Information-theoretic Approach. New
York: Springer-Verlag.
Choquet, R, Reboulet, A.M., Lebreton, J.D., Gimenez, O. and Pradel,
R. (2005). U-care 2.2 User´s Manual. CEFE, Montpellier, France.
<https://ftp.cefe.cnrs.fr/biom/Soft-CR/>.
Cooch, E. and White, G. (2015). Program MARK: A Gentle Introduction.
Colorado State University, Fort Collins. < http://www.fwspubs.org/
doi/suppl/10.3996/122012-JFWM-110R1/suppl_file/10.3996_122012jfwm-110r1.s8.pdf?code=ufws-site>.
Doherty, T.S., Glen, A.S., Nimmo, D.G., Ritchie, E.G and Dickman, C.R.
(2016). ‘Invasive predators and global biodiversity loss’. Proceedings
of the National Academy of Sciences’ 113: 11261–11265.
Evans, M., Herbert, H.J. and Rohnke, K. (1991). ‘Observations on
the Status of the Herpetofauna of Desecheo Island National Wildlife
Refuge, Puerto Rico’. In: J.A. Moreno (ed.) Estatus y Distribución de
los Reptiles y Anfibios de la Región de Puerto Rico, pp. 34–36. San
Juan: Departamento de Recursos Naturales de Puerto Rico.
Fisher, M. and Muth, A. (1989). ‘A technique for permanently marking
lizards’. Herpetological Review 20: 45–46.
Harper, G.A. and Bunbury, N. (2015). ‘Invasive rats on tropical islands:
Their population biology and impacts on native species’. Global
Ecology and Conservation 3: 607–627.
Harper, G.A., Zabala, J. and Carrion, V. (2011). ‘Monitoring of a
population of Galapagos land iguanas (Conolophus subcristatus) during
a rat eradication using brodifacoum’. In: C.R. Veitch, M.N. Clout and
D.R. Towns (eds.) Island invasives: eradication and management, pp.
309–312. Occassional Paper SSC no. 42. Gland, Switzerland: IUCN
and Auckland, New Zealand: CBB.
Herrera, J.L. and Bermúdez-Carambot, E.N. (2010). Survey of the
Desecheo Dwarf Sphaero, Sphaerodactylus levinsi (Squamata:
Gekkonidae) and a Mark-recapture Study with Anolis desechensis and
Ameiva desechensis, Desecheo Island, Puerto Rico. Internal report to
Island Conservation. Unpublished.
Hoare, J.M. and Hare, K.M. (2006). ‘The impact of brodifacoum on nontarget wildlife: Gaps in knowledge’. New Zealand Journal of Ecology
30: 157–167.
Holmes, N.D., Griffiths, R., Pott, M., Alifano, A., Will, D., Wegmann, A.S.
and Russell, J.C. (2015). ‘Factors associated with rodent eradication
failure’. Biological Conservation 185: 8–16.
Howald, G., Donlan, C.J, Galvan, J.P., Russell, J.C., Parkes, J.,
Samaniego, A., Wang, Y., Veitch, D., Genovesi, P., Pascal, M., Saunders,
A. and Tershy, B. (2007). ‘Invasive rodent eradication on islands’.
Conservation Biology 21: 1258–1268.
Jolly, G.M. (1965). ‘Explicit estimates from capture-recapture data with
both death and immigration-stochastic model’. Biometrika 52: 225–247.
Jones, H.P., Holmes, N.D., Butchart, S.H.M., Tershy, B.R., Kappes,
P.J., Corkery, I., Aguirre-Muñoz, A., Armstrong, D.P., Bonnaud, E.,
Burbidge, A.A., Campbell, K., Courchamp, F., Cowan, P.E., Cuthbert,
R.J., Ebbert, S., Genovesi, P., ‘Howald, G.R., Keitt, B.S., Kress, S.W.,
Miskelly, C.M., Oppel, S., Poncet, S., Rauzon, M.J., Rocamora, G.,
Russell, J.C., Samaniego-Herrera, A., Seddon, P. J., Spatz, D.R., Towns,
D.R. and Croll, D.A. (2016). ‘Invasive mammal eradication on islands
results in substantial conservation gains’. Proceedings of the National
Academy of Sciences 113: 4033–4038.
Keitt, B., Campbell, K., Saunders, A., Clout, M., Wang, Y., Heinz,
R., Newton, K and Tershy, B. (2011). ‘The Global Islands Invasive
Wertebrate Eradication Database: A Tool to Improve and Facilitate
Restoration of Island Ecosystems’. In: C.R. Veitch, M.N. Clout and
D.R. Towns (eds.) Island invasives: eradication and management, pp.
74–77. Occassional Paper SSC no. 42. Gland, Switzerland: IUCN and
Auckland, New Zealand: CBB.
Lebreton, J.D., Burnham, K.P., Clobert, J. and Anderson, D.R. (1992).
‘Modelling survival and testing biological hypotheses using marked
animals: A unified approach with case studies’. Ecological Monographs
62: 67–118.
Marshall, J.E. and Jewell, T. (2007). ‘Consumption of non-toxic baits
by grand (Oligosoma grande) and Otago (O. otagense) skinks’. DOC
Research and Development Series 272: 5–11.
McKown, M., Swinnerton, K. and Herrera-Giraldo, J.L. (2010). Desecheo
Reptile Monitoring Data. Internal report to Island Conservation.
Unpublished.
Meier, A.J. and Noble, R.E. (1991). ‘Notes on the natural history of Anolis
desechensis’. Florida Field Naturalist 19: 17–18.
Merton, D. (1987). ‘Eradication of rabbits from Round Island, Mauritius:
A conservation success story’. Dodo 24: 19–43.
Merton, D., Climo, G., Laboudallon, V., Robert, S. and Mander, C. (2002).
‘Alien mammal eradication and quarantine on inhabited islands in the
Seychelles’. In: C.R. Veitch and M.N. Clout (eds.) Turning the tide:
the eradication of invasive species, pp. 182–198. Occasional Paper SSC
no. 28. IUCN SSC Invasive Species Specialist Group. IUCN, Gland,
Switzerland and Cambridge, UK.
Morrison, J.A and Menzel Jr, E.W. (1972). ‘Adaptation of a freeranging rhesus monkey group to division and transplantation’. Wildlife
Monographs 31: 1–79.
Russell, J.C. and Holmes, N.D. (2015). ‘Tropical island conservation: Rat
eradication for species recovery’. Biological Conservation 185: 1–7.
Seber, G.A.F. (1965). ‘A note on the multiple recapture census’.
Biometrika 52: 249–259.
Shiels, A.B., Pitt, W.C., Sugihara, R.T. and Witmer, G.W. (2014). ‘Biology
and impacts of Pacific Island invasive species. 11. Rattus rattus, the
black rat (Rodentia: Muridae)’. Pacific Science 68: 145–184.
Tershy, B.R., Shen, K.W., Newton, K.M., Holmes, N.D. and Croll, D.A.
(2015). ‘The importance of islands for the protection of biological and
linguistic diversity’. BioScience: 10.1093/biosci/biv031.
Towns, D.R., Atkinson, I.A.E. and Daugherty, C.H. (2006). ‘Have
the harmful effects of introduced rats on islands been exaggerated?’
Biological Invasions 8: 863–891.
U.S. Fish and Wildlife Service (2016). Final Environmental Assessment:
Desecheo National Wildlife Refuge Rat Eradication Project. U.S.
Department of the Interior Fish and Wildlife Service, San Juan. <https://
www.fws.gov/caribbean/refuges/desecheo/>.
White, G.C. and Burnham, K.P. (1999). ‘Program MARK: Survival
estimation from populations of marked animals’. Bird Study 46: 120–
138.
195
P. Lago, J.S. Santiago Cabello and K. Varnham
Lago, P.; J.S. Santiago Cabello and K. Varnham. Long term rodent control in Rdum tal-Madonna yelkouan shearwater colony
Long term rodent control in Rdum tal-Madonna yelkouan
shearwater colony
P. Lago¹, J.S. Santiago Cabello¹ and K. Varnham²
1BirdLife Malta,57/28, Marina Court, Abate Rigord Street, Ta’ Xbiex (Malta) XBX 1120.
<paulolagobarreiro@gmail.com>. 2RSPB. The Lodge, Potton Road, Sandy. Bedfordshire. SG19 2DL, UK.
Abstract Rodent predation on eggs and chicks is one of the main threats to procellariiform species in the Mediterranean,
where the black rat (Rattus rattus) and brown rat (R. norvegicus) have been present on many islands for centuries. The
yelkouan shearwater (Puffinus yelkouan) is an endemic Mediterranean seabird species classified as vulnerable. Malta
holds up to 10% of the global population; the largest colony, Rdum tal-Madonna (RM), protected as a Natura 2000 site,
hosts around 500 breeding pairs. This colony has been monitored since its discovery in 1969. A very low reproductive
success due to rat predation was noticed in the late 1990s to early 2000s. In 2007 a seasonal rodent control programme
was established during the breeding season of yelkouan shearwater to reduce rat predation on eggs and chicks. Rodent
control took place between 2007 and 2010 and was reviewed and continued from 2012 to 2017. Breeding success since
2007 has been higher than 80%. In two other colonies with rat presence and where rodent control did not take place, the
breeding success in 2016 and 2017 was substantially lower than in the colony with the rodent control programme. The
European storm-petrel (Hydrobates pelagicus melitensis) only breeds in rat-free areas like remote sea caves or islets
around the Maltese islands. In 2014, the first breeding attempt by European storm-petrel was recorded on the Maltese
mainland at RM with a chick fledging successfully for the first time in 2016. The ongoing LIFE Arcipelagu Garnija
project is assessing rat predation in all Maltese yelkouan shearwater colonies in order to establish predator control in the
most important yelkouan shearwater breeding sites in 2018.
Keywords: breeding success, chicks, eggs, littering, rats, seabirds
INTRODUCTION
Malta is a southern European archipelago in the
Mediterranean Sea with three main islands: Malta, Gozo
and Comino; and other important islets: Filfla, Saint Paul,
Fungus Rock and Cominotto. Each island and islet harbours
important colonies of seabirds. The archipelago lies 80 km
from the south of Sicily (Italy), 284 km from the east of
Tunisia and 333 km from the north of Libya. The islands
cover over 315 km². Malta hosts internationally important
breeding populations of procellariiforms: yelkouan
shearwater (Puffinus yelkouan) (estimated 1370–2000
pairs, constituting up to 10% of the global population)
(Metzger, et al., 2015), Scopoli’s shearwater (Calonectris
diomedea) (estimated 4,500 pairs, up to 5 % of the global
population) and European storm-petrel (Hydrobates
pelagicus melitensis) (estimated 5,000–8,000 pairs, around
50% of the Mediterranean population) (Sultana, et al.,
2011).
The invasion of ecosystems by introduced species is
one of the most significant sources of ecosystem change
(Howald, et al., 2007) and biodiversity loss on islands
(Martin, et al., 2000; Courchamp, et al., 2003). Introduction
of alien rodents has been shown to have devastating effects
on insular ecosystems and some rodent species can be
important predators of nesting seabirds (Traveset, et al.,
2009), especially procellariiforms (Imber, 1978). Rodent
predation on eggs and chicks is one of the main dangers to
this group of seabirds across the world (Booth, et al., 1996;
Hobson, et al., 1999; Gaze, 2000; Imber, et al., 2000). Rats
are associated with extinctions or declines of burrowing
seabirds (Seto & Conant, 1996; Towns, et al., 2006). Rats
have a severe impact on breeding success and are a major
cause of seabird mortality in the world (Jones, et al., 2008;
Pascal, et al., 2008).
Rats were introduced into the Mediterranean over
2000 years ago and have been present on many islands
for centuries (Atkinson, 1985; Audoin-Rouzeau & Vigne,
1994; Martin, et al., 2000). Black rat (Rattus rattus) is the
most devastating predator of seabirds in the Mediterranean
(Igual, et al., 2006) and the main reason for breeding failure
on some islands, for example Corsica (Thibault, 1995).
Therefore, the persistence of native long-lived seabirds
in the Mediterranean basin, despite the long-standing
introduction of black rat on most islands, constitutes an
amazing conservation paradox (Ruffino, et al., 2009).
Yelkouan shearwater is an endemic Mediterranean
seabird belonging to the family Procellariidae. It is a longlived species that lays a single egg each season in deep
burrows. It has been classified as vulnerable since 2012
according to the IUCN (BirdLife International, 2016).
The Maltese population of yelkouan shearwater has
declined in recent years, mainly due to predation by rats,
loss of breeding habitat, illegal hunting, fishing bycatch,
disturbance and light and sound pollution (Sultana, et al.,
2011).
The main colony in Malta situated in Rdum talMadonna (RM) holds around 500 breeding pairs (2 or 3%
of the global population). It is a Natura 2000 site – part
of the European network of protected areas. This colony
is situated along 1 km of coralline limestone sea cliff. It
has been monitored since its discovery in 1969 and it was
noticed that the breeding success in the late 1990s to early
2000s was very low, largely due to rat predation, with
very few chicks fledging (Sultana, et al., 2011). The best
response to such a situation is almost always to control
the alien population, either by frequently reducing their
numbers, or better still, by eradicating the whole population
(Courchamp, et al., 2003).
As the colony is located on the Maltese main island,
eradication of rats was not feasible because it is not possible
to isolate the area from rat populations found across the
rest of the island. The population of rats benefits from the
persistent availability of food close to the colony. Litter
from recreational users in the area making barbecues and
camping is compounded by the inefficient and inadequate
waste disposal and collection system. Actions to increase
awareness about littering between site users and authorities
were carried out but no substantial improvement in the
situation was observed.
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
196
up to meet the challenge, pp. 196–199. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Lago, et al.: Long term rodent control
In 2007, a seasonal rodent control programme was
established at the site to reduce rat predation. The control
programme has now been active for 11 years from 2007
to 2017. In this paper we present the results of the rodent
control programme on the breeding success of the yelkouan
shearwater colony. We discuss the results and lessons learnt
and their applicability to other locations.
MATERIALS AND METHODS
The colony site is surrounded by the ocean on three
sides, making it an ideal site for rodent control. The
methodology chosen for rodent control was seasonal
control using rodenticide. The most frequent rodenticide
distribution method used on small islands around the
world is bait stations (Howald, et al., 2007) and other
projects have shown that using a permanent bait-station
system is an efficient methodology to control rats (Orueta,
et al., 2005; Pascal, et al., 2008). Around 90 bait stations
(PROTEXX TM) were distributed over 25 ha of RM on the
top of the cliff plateau and the lower part of the cliffs where
yelkouan shearwaters breed (Fig.1). The bait stations on
top of the cliffs create a buffer to prevent rats accessing the
colony. Bait stations were placed around areas of high rat
presence, for example those areas subject to littering from
campers. Rodent control took place around nesting sites
between February and July during the yelkouan shearwater
breeding season, when eggs and chicks are most vulnerable.
The bait stations were baited one to three times per month,
depending on rodent activity. Each bait station contained
two blocks of anticoagulant rodenticide. Between 2007
and 2015 the rodenticide used was brodifacoum 0.005%
and from 2016 it was bromadiolone 0.005% to reduce the
risk of secondary poisoning. The bait blocks were threaded
on to metal skewers that were clipped in place, to prevent
them falling out of the stations even if they were shaken
violently.
Every time the bait stations were checked, data were
collected on the amount of rat sign (droppings and rat teeth
marks in the wax bait blocks), non-target species sign like
mice, shrews and insects taking the bait, and the number of
bait blocks replaced. The area baited was checked for signs
of dead rats and primary or secondary poisoning of nontarget species. Rat presence was calculated as the number
of bait stations with rat teeth marks on the bait divided by
the total number of bait stations.
Table 1 Breeding success (% of chicks fledged per eggs
laid) of yelkouan shearwater at Rdum tal-Madonna
between 2007 and 2017.
Year
2007
2008
2009
2012
2013
2014
2016
2017
No. of nests
6
12
11
16
32
25
24
38
Breeding success
83%
92%
91%
94%
88%
88%
88%
84%
Table 2 Breeding success (% of chicks fledged per eggs
laid) of yelkouan shearwater in 2016 and 2017 in Rdum
tal-Madonna (rodent control) and St. Paul’s Island and
Majjistral (no rodent control).
Colony
RM
St. Paul’s Is
Majjistral
RM
St. Paul’s Is
Majjistral
Year
2016
2016
2016
2017
2017
2017
No. of nests
24
9
12
38
9
11
Breeding success
88%
67%
33%
84%
11%
55%
RESULTS
Rodent control took place between 2007 and 2010,
after which it was reviewed and then continued from 2012
to 2017. After the first season of rodent control in 2007, the
occurrence of eggs and chicks depredated by rats dropped
dramatically and there have been few recorded signs of rat
predation during the subsequent 11 years. Breeding success
has been very high since rodent control started (Table 1),
with a mean of 88 % (averaged over the eight years for
which data are available).
In 2016 and 2017, the breeding success (chicks fledged
per eggs laid) in RM (88% and 84%, respectively) was
much higher than in two other colonies where rats were
known to be present but no rat control took place, St. Paul’s
Island (67% and 11%, respectively) and Majjistral Park
(33% and 55%, respectively) (Table 2).
In RM, rat activity varies throughout the yelkouan
shearwater breeding season. Rats are regularly present
from February until July. Rat activity is reduced after the
first month of rat control in February, the peak of activity
is in May and then a small upturn in June (Fig. 3). Rodent
activity over the period 2012–2017 (data available for four
years) shows a decrease in rat presence in recent years. No
signs of secondary poisoning have been found in the study
period.
Fig. 1 Map of Rdum tal-Madonna colony in Malta where
the Yelkouan shearwater colony is situated (rectangle)
and the location of the bait stations (dots).
After the first season of rat control in 2007, European
storm-petrels were regularly seen in RM (Borg, et al.,
2010). In 2014 the first breeding attempt was recorded
and in 2016 and 2017 chicks fledged successfully. The
data collected during 2014–2017 suggest European Stormpetrel is establishing a breeding colony in RM.
197
Island invasives: scaling up to meet the challenge. Ch 1C Rodents: Lessons
Fig. 2 Proportion (%; with mean and standard deviation)
of bait stations with rodent bite marks throughout the
yelkouan shearwater breeding season from February to
July (for years 2012, 2015, 2016, 2017 combined).
The rodent control programme showed its effectiveness
at increasing the breeding success of yelkouan shearwater
and allowed the establishment of a new European stormpetrel population. The main yelkouan shearwater colony
locations are situated on the main islands of Malta and Gozo
making the eradication of rats not possible. Rat eradication
could only be feasible in the islands of Comino, Cominotto
and Saint Paul that hold smaller colonies. Ongoing rodent
control programmes are therefore needed in the main
colonies to secure yelkouan shearwater populations in
the archipelago and to improve their situation. Building
on the lessons learnt and the success of the rodent control
programme in RM, the current EU-Life Arċipelagu Garnija
project LIFE14 NAT/MT/991 is assessing predation by
rats in all Maltese yelkouan shearwater colonies in order
to establish predator control in the most important sites in
2018 and secure the main colonies across Malta. During
the study period, no evidence of secondary poisoning was
found but, in any case, from 2016 the bait was changed
from brodifacoum to bromadiolone that has less risk
of secondary poisoning. Less toxic bait, such as first
generation anticoagulants, are not available in Malta. In
order to reduce the amount of anticoagulant used in the
new rat control programmes, the current project is testing
methodologies to replace or combine anticoagulant baiting
with auto-reset mechanical traps and carrying out activities
to increase awareness about littering among site users.
ACKNOWLEDGEMENTS
Fig.3 The percentage of bait stations with rodent bite
marks by year; 2012, 2015, 2016 and 2017.
DISCUSSION
Seasonal rat control in seabird colonies where
eradication is not feasible is an effective way to reduce rat
predation and increase reproductive success (Imber, et al.,
2000; Martin, et al., 2000; Jouventin, et al., 2003; Orueta,
et al., 2005; Igual, et al., 2006; Pascal, et al., 2008). In
many cases, the removal of the alien invasive species is
followed by a fast and often great recovery of the damaged
local populations (Courchamp, et al., 2003), even allowing
new colonies of other species to become established, as
has been seen at RM (Malta). However, only intensive and
constant long-term poisoning will control rats satisfactorily
(Jouventin, et al., 2003).
The increase in the reproductive success observed
during recent years in the yelkouan shearwater colony in
RM is correlated with the lower rat activity as a result of
rodent control programme. Rat activity varies throughout
the yelkouan shearwater breeding season. The peak
of activity in May is related to the increase in ambient
temperature and also to the start of camping activity in
the area. The presence of campers increases littering (i.e.
supplying food for rodents) which is the likely reason for
the increase in the rat population around the colony. The
general decrease of rat presence in 2016 and 2017 may
be related to the very dry weather in these two years, but
possibly also to increased public awareness about littering.
On 30 April 2017, an intensive clean-up by more than 100
volunteers was organised in the area.
198
We thank B. Metzger, N. Barbara, J. J. Borg, A.
Raine, H. Raine, G. Meier, and all staff and volunteers
working on rat control over the last 11 years. The rodent
control programme was funded throughout the years by
the EU-Life Yelkouan Shearwater Project LIFE06 NAT/
MT/000097 from 2007 till 2010, MEPA and HSBC. It is
currently part of the EU-Life funded Arċipelagu Garnija
project LIFE14 NAT/MT/991, co-funded by the Maltese
Ministry for the Environment, Sustainable Development
and Climate Change (MESDC). All the activities were
carried out under MEPA/ERA permits.
REFERENCES
Atkinson, I.A.E. (1985). ‘The spread of commensal species of Rattus to
oceanic islands and their effects on island avifaunas’. In: P.J. Moors (ed.)
Conservation of Island Birds, pp. 35–81. ICBP Technical Publication 3.
Audoin-Rouzeau F. and Vigne J.D. (1994). ‘La colonisation de l’Europe
par le rat noir Rattus rattus’. Revue de Paléobiologia13: 125–145.
BirdLife International (2016). Puffinus yelkouan. The IUCN Red List
of Threatened Species 2016: e.T22698230A93672084. <http://dx.doi.
org/10.2305/IUCN.UK.2016-3.RLTS.T22698230A93672084.en>.
Downloaded on 13 February 2018.
Booth, A.M., Minot, E.O., Fordham, R.A. and Innes, J.G. (1996). ‘Kiore
(Rattus exulans) predation on the eggs of the little shearwater (Puffinus
assimilis haurakiensis)’. Notornis 43: 147–153.
Borg, J.J., Raine, A. and Raine, H. (2010). ‘Records in Malta of European
storm-petrels Hydrobates pelagicus melitensis during the breeding
season away from known breeding sites’. Il-Merill 32: 11–14.
Courchamp, F., Chapuis, J.L. and Pascal, M. (2003). ‘Mammal invaders
on islands: Impact, control and control impact’. Biological Reviews
78(3): 347–383.
Gaze, P. (2000). ‘The response of a colony of sooty shearwater (Puffinus
griseus) and flesh-footed shearwater (P. carneipes) to the cessation of
harvesting and the eradication of Norway rats (Rattus norvegicus)’.
New Zealand Journal of Zoology 27: 375–379.
Hobson, K.A., Drever, M.C. and Kaiser, G.W. (1999). ‘Norway rats
as predators of burrow-nesting seabirds: insights from stable isotope
analyses’. The Journal of Wildlife Management 63: 14–25.
Lago, et al.: Long term rodent control
Howald, G., Donlan, C., Galván, J.P., Russell, J.C., Parkes, J.,
Samaniego, A.,Wang, Y., Veitch, D., Genovesi, P., Pascal, M., Tershy,
B. and Saunders, A. (2007). ‘Invasive rodent eradication on islands’.
Conservation Biology 21(5): 1258–1268.
Igual, J.M., Forero, M.G., Gomez, T., Orueta, J.F. and Oro, D. (2006). ‘Rat
control and breeding performance in Cory’s shearwater (Calonectris
diomedea): Effects of poisoning effort and habitat features’. Animal
Conservation 9(1): 59–65.
Imber, M.J. (1978). ‘The effects of rats on breeding success of petrels’.
In: P.R. Dingwall, I.A.E. Atkinson and C. Hay (eds.) The Ecology
and Control of Rodents in New Zealand Nature Reserves, pp. 67–71.
Wellington: Department of Lands and Survey Information Series.
Imber, M., Harrison, M. and Harrison, J. (2000). ‘Interactions between
petrels, rats and rabbits on Whale Island, and effects of rat and rabbit
eradication’. New Zealand Journal of Ecology 24: 153–160.
Jones, H.P., Tershy, B.R., Zavaleta, E.S., Croll, D.A., Keitt, B.S.,
Finkelstein, M.E. and Howald, G.R. (2008). ‘Severity of the effects of
invasive rats on seabirds: a global review’. Conservation Biology 22(1):
16–26.
Jouventin, P., Bried J. and Micol, T. (2003). ‘Insular bird populations can
be saved from rats: a long-term experimental study of white-chinned
petrels Procellaria aequinoctialis on Île de la Possession (Crozet
archipelago)’. Polar Biology 26: 371–378.
Martin, J.L., Thibault, J.C. and Bretagnolle, V. (2000). ‘Black rats,
island characteristics, and colonial nesting birds in the Mediterranean:
consequences of an ancient introduction’. Conservation Biology 14:
1452–1466.
Metzger, B., Oppel, S., Carroll, M., Meirinho, A., Dias, M., Barbara,
N. and Lago, P. (2015). Malta Marine IBA Inventory Report. EU
LIFE+Malta Seabird Project LIFE10 NAT/MT/090. BirdLife Malta.
[online] available at: <https://birdlifemalta.org/conservation/maltaseabirdproject/> Accessed 15 March 2018.
Orueta, J.F., Aranda, Y., Gómez, T., Tapia, G.G. and Sanchez-Marmol, L.
(2005). ‘Successful eradication of invasive rodents from a small island
through pulsed baiting inside covered stations’. Biological invasions
7(1): 141–147.
Pascal, M., Lorvelec, O., Bretagnolle, V. and Culioli, J.M. (2008).
‘Improving the breeding success of a colonial seabird: a costbenefit comparison of the eradication and control of its rat predator’.
Endangered Species Research4(3): 267–276.
Ruffino, L., Bourgeois, K., Vidal, E., Duhem, C., Paracuellos, M.,
Escribano, F., Sposimo, P., Baccetti N., Pascal, M. and Oro, D. (2009).
‘Invasive rats and seabirds after 2,000 years of an unwanted coexistence
on Mediterranean islands’. Biological invasions 11(7): 1631–1651.
Seto, N.W.H. and Conant, S. (1996). ‘The effects of rat (Rattus rattus)
predation on the reproductive success of the Bonin petrel (Pterodroma
hypoleuca) on Midway Atoll’. Colonial Waterbirds 19 (2):171–185.
Sultana, J., Borg, J.J., Gauci, C. and Falzon, V. (2011). The breeding birds
of Malta. Malta: Birdlife Malta.
Thibault, J.C. (1995). ‘Effect of predation by the black rat Rattus rattus
on the breeding success of Cory’s Shearwater Calonectris diomedea in
Corsica’. Marine Ornithology 23: 1–10.
Towns, D.R., Atkinson, I.A. and Daugherty, C.H. (2006). ‘Have the
harmful effects of introduced rats on islands been exaggerated?’.
Biological Invasions 8: 863–891.
Traveset, A., Nogales, M., Alcover, J.A., Delgado, J.D., López-Darias,
M., Godoy, D., Igual, J.M. and Bover, P. (2009). ‘A review on the effects
of alien rodents in the Balearic (Western Mediterranean Sea) and Canary
Islands (Eastern Atlantic Ocean)’. Biological Invasions11: 1653–1670.
199
S. Oppel, G.T.W. McClelland, J.L. Lavers, T. Churchyard, A. Donaldson, N. Duffield, S. Havery, J. Kelly, T. Proud, J.C. Russell and A.L. Bond
Oppel, S.; G.T.W. McClelland, J.L. Lavers, T. Churchyard, A. Donaldson, N. Duffield, S. Havery, J. Kelly, T. Proud, J.C. Russell and A.L. Bond.
Seasonal variation in movements and survival of invasive Pacific rats on sub-tropical Henderson Island: implications for eradication
Seasonal variation in movements and survival of invasive Pacific rats
on sub-tropical Henderson Island: implications for eradication
S. Oppel1, G.T.W. McClelland1, J.L. Lavers1,2, T. Churchyard1, A. Donaldson1, N. Duffield1, S. Havery1, J. Kelly1,
T. Proud1, J.C. Russell3 and A.L. Bond 1
RSPB Centre for Conservation Science, Royal Society for the Protection of Birds, The Lodge, Sandy, Bedfordshire,
SG19 2DL United Kingdom. <steffen.oppel@rspb.org.uk>. 2Current affiliation: Institute for Marine and Antarctic
Studies, University of Tasmania, 20 Castray Esplanade, Battery Point, Tasmania, 7004 Australia. 3School of Biological
Sciences, University of Auckland, Private Bag 92019, Auckland 1142, New Zealand.
1
Abstract Invasive rodents are successful colonists of many ecosystems around the world, and can have very flexible
foraging behaviours that lead to differences in spatial ranges and seasonal demography among individuals and islands.
Understanding such spatial and temporal information is critical to plan rodent eradication operations, and a detailed
examination of an island’s rat population can expand our knowledge about possible variation in behaviour and demography
of invasive rats in general. Here we investigated the movements and survival of Pacific rats (Rattus exulans) over five
months on sub-tropical Henderson Island in the South Pacific Ocean four years after a failed eradication operation. We
estimated movement distances, home range sizes and monthly survival using a spatially-explicit Cormack-Jolly-Seber
model and examined how movement and survival varied over time. We captured and marked 810 rats and found a median
maximum distance between capture locations of 39 ± 25 m (0–107 m) in a coastal coconut grove and 61 ± 127 m (0–1,023
m) on the inland coral plateau. Estimated home range radii of Pacific rats on the coral plateau varied between ‘territorial’
(median: 134 m; 95% credible interval 106–165 m) and ‘roaming’ rats (median: 778 m; 290–1,633 m). The proportion of
rats belonging to the ‘roaming’ movement type varied from 1% in early June to 23% in October. There was no evidence to
suggest that rats on Henderson in 2015 had home ranges that would limit their ability to encounter bait, making it unlikely
that limited movement contributed to the eradication failure if the pattern we found in 2015 is consistent across years.
We found a temporal pattern in monthly survival probability, with monthly survival probabilities of 0.352 (0.081–0.737)
in late July and 0.950 (0.846–0.987) in late August. If seasonal variation in survival probability is indicative of resource
limitations and consistent across years, an eradication operation in late July would likely have the greatest probability of
success.
Keywords: home range, introduced species, island restoration, Pitcairn Islands, Rattus exulans
INTRODUCTION
Eradications are a powerful and frequently used
management option to counter the native biodiversity
loss caused by invasive species on islands (Jones, et al.,
2016). Planning for an eradication requires a fundamental
understanding of the ecology and movement characteristics
of the target invasive species (Zavaleta, 2002; Keitt, et al.,
2015). Among the most widespread invasive species on
islands are three species of rat (Rattus rattus, R. norvegicus,
R. exulans), which now occur on >80% of the world’s island
groups (Atkinson, 1985; Jones, 2010). Rat eradications
have been successfully completed on hundreds of islands
(Howald, et al., 2007), but eradications on tropical islands,
where a lack of seasonal fluctuation in resource abundance
allows rodents to reproduce throughout the year, still
have a lower success probability than eradications on
temperate islands (Holmes, et al., 2015; Keitt, et al., 2015).
Detailed information on rat movements and demography
from tropical islands should therefore benefit eradication
planning on tropical islands (Keitt, et al., 2015).
Rodent eradications on islands larger than 100 ha are
generally conducted by aerially distributing cereal-based
toxic bait pellets across the island, and are only successful
if every individual rodent has access to sufficient bait
within its home range to consume a lethal dose of toxin
(Cromarty, et al., 2002; Howald, et al., 2007; Broome, et al.,
2014; Holmes, et al., 2015). Hence, a better understanding
of the size of home ranges can inform the density at
which bait pellets need to be dispersed on the ground.
Movements of invasive rodents on islands vary by habitat,
population density, food availability, individuals’ age and
sex (Bramley, 2014a; Ringler, et al., 2014; Harper, et al.,
2015), but more information on the size of movements
and their variation over time of year could contribute to
eradication planning on islands.
Besides ensuring each individual has access to a
sufficient quantity of bait, rodent eradications are also
more likely to succeed if they are timed to coincide with
a predictable period of rodent stress (e.g. mortality). On
temperate islands, mortality occurs during a predictable
seasonal shortage in resource availability during autumn
or winter, and therefore provides a natural time window
for an eradication operation when rodents are more likely
to switch to palatable poison baits (Howald, et al., 2007;
Russell & Ruffino, 2012). On tropical islands, with lessdefined seasonality and irregular periods of resource
limitation, there is still very little information on how the
survival of rodents varies within a year (but see Tamarin
& Malecha, 1971). Additional information on seasonal
variation in survival of rodents on tropical islands can
inform when an eradication operation would have the
highest probability of success and therefore aid the
planning of an eradication operation (Howald, et al., 2007;
Holmes, et al., 2015; Keitt, et al., 2015).
Here we use data from a large spatial capture-recapture
programme and conventional radio-tracking to investigate
the movements of invasive Pacific rats (R. exulans) on an
uninhabited sub-tropical island (Henderson) in the South
Pacific. An eradication operation on this island in 2011
failed to kill all individuals. Among the reasons that can
cause eradication failure, insufficient bait toxicity could
be excluded due to follow-up experiments (Amos, et al.,
2016). However, two further potential causes, namely that
not all rats had access to bait and that the eradication was
poorly timed and coincided with high survival, have not
been investigated so far. Our study was designed to provide
knowledge to better understand the 2011 eradication
failure and improve the probability of success of a future
eradication attempt. We estimate movement distances and
home range sizes using mark-recapture and radio-tracking
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
200
up to meet the challenge, pp. 200–208. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Oppel, et al.: Movement & survival of Pacific rats
data and evaluate if the smallest rodent home ranges would
contain a sufficient quantity of bait pellets based on bait
distribution rates used during the eradication attempt in
2011. We further estimate survival of rats over a fivemonth period, examine temporal variation in their monthly
survival probability, and assess whether the timing of the
failed operation in 2011 was appropriate.
METHODS
Study area
Henderson Island (24º22′ S, 128º20′ W) is a flat, raised
coral atoll of 4,309 ha in the sub-tropical Pacific Ocean
with two distinct habitats – a central plateau roughly 25
m above sea level (4,290 ha), and a sandy beach area with
a vegetated margin (hereafter referred to as ‘embayment
forest’, 14 ha). Henderson Island has a sub-tropical climate
with erratic rainfall patterns, and there are no permanent
freshwater bodies on the island (Spencer, 1995; Weigelt,
et al., 2013). The plateau substrate is fossilised coral
with uniform, dense native vegetation consisting mostly
of Pandanus tectorius, Xylosma suaveolens and Psydrax
odorata (Waldren, et al., 1995). The beach and embayment
forest areas have a sandy substrate with a mixed shrubby
vegetation and small stands of introduced coconut (Cocos
nucifera) (Paulay & Spencer, 1989; Waldren, et al., 1995).
Pacific rats were introduced to Henderson Island
by Polynesians several hundred years ago (Steadman
& Olson, 1985), and currently have adverse effects on
native biodiversity on Henderson Island (Brooke, et
al., 2010; Dawson, et al., 2015). In late August 2011, an
operation using the aerial distribution of cereal-based
pellets containing 20 μg/g of the toxin brodifacoum was
carried out to eradicate all Pacific rats from Henderson
Island. Although the baiting operation met best practice
standards, had no spatial gaps in bait distribution, used
bait pellets containing a sufficient amount of toxin (Torr
& Brown, 2012), and used bait application densities well
beyond those needed to overcome estimated hermit-crab
consumption (Cuthbert, et al., 2012), the eradication
operation was unsuccessful and 60-80 individual rats
were predicted to have survived (Amos, et al., 2016). Rat
populations recovered within 2–4 years (Bond, et al., 2019)
and were at an unknown stage of expansion or fluctuation
during 2013 and 2015.
Rat live trapping
To obtain a robust estimate of rat survival probability,
and to document rat movements over five months, we
implemented a spatial capture-mark-recapture programme
in 2015. Rats were live-trapped on the plateau from 28
May to 16 October 2015 during seven primary sessions of
10 trapping nights each, with a window of 8–15 days with
no trapping between primary sessions. This time frame was
chosen because food availability for rats was assumed to
be lower during the ‘winter’ months on Henderson than at
other times of the year (Spencer, 1995; Brooke & Towns,
2008). In the embayment forest, rats were live-trapped
between 1 August and 19 September 2015 during three
primary sessions of 6–10 trapping nights each.
On the plateau we established a trap network placed
along 3 km of cleared path (Fig. 1). Traps were arranged
at distances from 3–20 m at 343 locations, with a different
subset of trap locations used during each primary
session due to gradual progression of trail clearance. In
the embayment forest, we established a grid of 63 traps
arranged in an oblique rectangular configuration (Fig. 1)
with traps spaced 10 m apart. Traps were placed on the
ground, marked with a unique number, and locations were
recorded to within 5 m using a hand-held GPS device.
Fig. 1 Map of the trapping network used on Henderson
Island in 2015. Black triangle is the research camp, the
grey line is the upper margin of the beach, + indicate
trap locations on the coral plateau, and black dots
indicate trap locations in the embayment forest.
We used two different live trap types, a small metal box
(7.6 × 8.9 × 22.9 cm, LFA Folding Trap, H. B. Sherman
Traps Inc., Tallahassee, Florida, USA), and a metal cage
(13 × 16 × 27 cm, Metal Rat Cage Trap, Key Industries,
Auckland, New Zealand). Before the first primary session
in each habitat, traps were deployed, but not opened,
for approximately five days to allow rats to overcome
neophobia (Yackel Adams, et al., 2011; Russell, et al.,
2015). For each night in each ten-day trapping period, traps
were baited with a small (1 × 1 cm) cube of fresh coconut
between 1600–1800 h local time (UTC-8), and checked the
following morning between 0800–1000 h.
Each captured rat was fitted with a uniquely numbered
ear tag (size 1005-1; National Band & Tag Company,
Newport, Kentucky, USA), and the rat was released
next to the trap. We recorded the trap location for each
capture, whether female rats showed signs of lactation or
pregnancy, and whether traps were available to capture
rats or had been de-activated (e.g., by crabs). Upon their
first capture, rats were sexed by examination of external
genitalia, weighed using a spring balance (± 1 g; Pesola
AG, Schindellegi, Switzerland), and their body and tail
lengths were measured to the nearest 1 mm using a metal
ruler (Cunningham & Moors, 1996).
In November 2015, after the mark-recapture effort, we
also conducted lethal snap-trapping along a subset of the
locations of the live-trap locations on the plateau. This
lethal trapping was designed to provide definitive age and
201
Island invasives: scaling up to meet the challenge. Ch 1C Rodents: Lessons
sex classifications and body measurements for as many
tagged rats as possible.
Radio-tracking
To provide an alternative estimate of movement range
not dependent on the recapture of a rat, we radio-tracked
rats that were captured on the plateau in July 2013 using
the same small metal box traps as mentioned above. We
fitted radio-collars (pipAg393, 2.6 g, Biotrack, Wareham,
UK) attached to plastic collars with rubber tubing to each
rat. After fitting the collar, rats were placed back in the trap
and monitored for five minutes; adjustments were made to
the collar if necessary before the rat was released at the site
of capture. The capture location, sex, reproductive status
(males with or without descended testes; females with or
without a perforated vagina) and mass were recorded for
all radio-tracked rats as described above.
After release, rats were located at least twice daily
during daylight hours using a three-element Yagi antenna
and Telonics TR-4 receivers with each radio-collar
separated by frequency. Locations were either recorded by
homing using a hand-held GPS device with an accuracy of
<5 m or estimated through bisection by using distance and
bearing from two observation points with an accuracy of
ca. 25 m (Kenward, 2001).
Calculation of movement distances
We first calculated the straight-line distance between
trap locations for subsequent captures of individual
rats. These distances are a conservative estimate of rats’
movement distances, because they assume an unrealistic
direct line of travel from one trap to the next. We summed
all distances between subsequent captures and divided the
total travel distance calculated for each individual by the
number of captures to provide an overall estimate of mean
distance moved between two capture events that is not
dependent on the number of captures (Püttker, et al., 2012).
We also calculated the observed range length, defined as
the maximum distance between any two capture locations
for a given individual (Stickel, 1954; Lindsey, et al., 1973).
We present results as median ± standard deviation and
range.
Analysis of home range size and survival
To estimate rat survival while taking movements
into account, we used a spatially-explicit CormackJolly-Seber (CJS) model adapted from similar models
(Gardner, et al., 2010; Raabe, et al., 2013; Royle, et al.,
2016). We considered each primary session as a capture
occasion and reduced binomial capture data from trapping
nights to counts of each individual at each trap location
during a given primary session because robust-design
formulations of the spatial CJS model (Ergon & Gardner,
2014) did not converge. We removed all rats that were
captured only once from the analysis, because these
transients do not provide any information on movement or
survival probability (Pradel, et al., 1997), and we draw no
inferences from estimated capture probabilities. We also
implemented a non-spatial CJS survival model following
Russell & Ruffino (2012), to compare to the spatial
model. This model yielded similar mean estimates and
temporal variation in survival, suggesting the spatial model
results are valid, but with much greater precision by not
incorporating the large variance in rat movements (ESM
Fig. S1). Understanding and incorporating rat movements
is critical for distinguishing survival from movement in
apparent survival models (Gilroy, et al., 2012; Schaub
& Royle, 2014), especially for inferring potential factors
in eradication failure, and we therefore present only the
results of the spatial CJS model.
202
Our spatial CJS model assumed that rat home ranges
were circular, but that the estimated centre of a rat’s home
range could vary spatially based on an individual-specific
correlated random walk parameter (Royle, et al., 2016),
which effectively allowed rats to shift their activity centre
over time. We also assumed that capture probability of rats
at a given trap followed a negative exponential function
based on the distance of the rat’s home range centre to the
trap (Ergon & Gardner, 2014; Royle, et al., 2016), and
that the shape of this capture probability function varied
over time and among individuals. Because exploratory
analysis of rat movements indicated that neither individual
nor environmental covariates could adequately capture
the variation in rat movement, we assumed that the shape
of the capture probability function originated from two
different statistical distributions: one distribution reflected
‘territorial’ rats and was specified as a normal distribution
with a mean of σ = 30, which corresponds to a typical home
range radius for insular rats (Bramley, 2014b; Ringler, et al.,
2014; Harper, et al., 2015). The other distribution reflected
‘roaming’ rats with a uniform distribution between σ = 60
– 400, allowing a movement radius of 1,000 m, which has
been recorded for Pacific rats in other studies (Wirtz, 1972;
Lindsey, et al., 1973). For each individual rat, we allowed
the model to select the home range radius parameter
belonging to either the ‘territorial’ or ‘roaming’ movement
type, and we report the proportion of males and females
that were estimated to belong to each type.
We estimated rats’ survival probability between
primary sessions and assumed that survival varied over
time. Because the interval among primary sessions was not
constant, we calculated the interval as the time difference
between the mid-point of subsequent primary sessions
(range: 17–25 days) and converted survival probabilities to
monthly survival probabilities to allow a direct comparison
among different primary sessions. In a CJS model the
probabilities of capture and survival are confounded for
the last trapping occasion; to allow inference on survival
probability up to our last live-trapping occasion in October
2015, we included data from a final additional session
of kill trapping in November 2015 in the model (sensu
Nathan, et al., 2015), and allowed for a different capture
probability for that trapping period. Because rat survival
may vary by sex and may depend on food availability
(Russell & Ruffino, 2012; Ringler, et al., 2014), we
included individual sex and the Normalised Difference
Vegetation Index (NDVI) as covariates affecting survival
probability. NDVI is a measure of vegetation ‘greenness’
derived from remote sensing imagery and can serve as
a useful proxy for rat food availability (Pettorelli, et al.,
2011; Pettorelli, et al., 2014). We downloaded NDVI
for Henderson Island at a 250 m resolution from NASA
Earth Data (https://daacmodis.ornl.gov/cgi-bin/MODIS/
GLBVIZ_1_Glb/modis_subset_order_global_col5.pl),
and averaged NDVI over 32 days centred on the mid-point
of each survival period to reflect the food availability for
rats during the period over which survival was estimated.
We used diffuse priors for covariate effects on survival, but
used informative priors for daily survival probabilities that
were based on previous studies (Tamarin & Malecha, 1971;
Moller & Craig, 1987; Roberts & Craig, 1990). Timespecific priors for daily survival probability were drawn
from a random uniform distribution between 0.9 and 1.
We fitted the robust design CJS model in JAGS v 3.4.0
(Plummer, 2012) using the ‘jagsUI’ package (Kellner,
2016) called from R 3.2.5 (R Core Team, 2016). We
ran three Markov chains each with 30,000 iterations,
discarded the first 7,000 iterations as adaptation and burnin, and tested for convergence using the Gelman-Rubin
diagnostic (Brooks & Gelman, 1998) as well as visual
representations of all parameters of interest. We report
posterior mean estimates and 95% credible intervals for
Oppel, et al.: Movement & survival of Pacific rats
survival probability and the spatial shift of home range
centres among primary capture sessions. Code to repeat
the analysis can be downloaded from: https://github.com/
steffenoppel/henderson/blob/master/Oppel_etal_SECR_
ANALYSIS_and_DATA.zip.
To estimate a ‘home range radius’ from the shape of
the spatial detection function, we assumed a circular
exponential distribution for individual home ranges, and
calculated an approximation of the home range radius
that would encompass 95% of an individual’s territory
using the function ‘circular.r’ in R package secr 2.10.2
(Efford, 2016). We converted this estimate of home range
radius to an estimate of home range size using standard
geometry (A = r2). This estimate of space use, although
not equivalent to a home range estimate obtained from
telemetry, allowed us to compare the space use inferred
from our spatial trapping approach to a similar metric
estimated from radio-tracking to compare the conclusions
from each approach.
To provide a comparable estimate of home range size
from radio-tracking data, we first calculated the minimum
convex polygon (MCP) for each tracked animal and then
calculated the 95% kernel utilization distribution using the
‘kernelUD’ function of the ‘adehabitatHR’ package in R
(Calenge, 2006) for all rats with >10 position fixes after
capture. We parameterized our kernel density estimation
model using a grid size of 1000, and a smoothing parameter
of h = 10 m to avoid overestimation of home ranges due to
large kernels around single locations.
Adequacy of cereal bait distribution during
eradication attempt
To assess how many bait pellets would have been
available to rats, we calculated the approximate number
of bait pellets that would have been available in minimum
home range sizes of rats during the eradication operation
in 2011 based on mean bait application rates. In 2011, bait
was distributed at 40–60 kg/ha in the embayment forest
and 10 kg/ha on the plateau during the first of two bait
applications. Given that a bait pellet weighs ca. 1.8 g, there
were between 22,000 and 33,000 pellets/ha available in
the embayment forest, and 5,500 pellets/ha on the plateau.
For each of the home range estimates from radio-tracking
and spatial re-capture, we multiplied the estimated size of
the minimum home range area by the density of pellets to
infer how many bait pellets would have been accessible to
individual rats.
RESULTS
Rat movement
We captured and marked a total of 810 rats, of which
580 were recaptured at least once, yielding a total of 4,920
capture events at 396 unique trap locations. On the plateau,
we captured 727 individuals of which 524 were recaptured
at least once; in the embayment forest we captured 86
individuals of which 56 were recaptured at least once; only
three individuals were captured in both habitats.
The median movement distance between subsequent
captures was 17 ± 19 m (range: 0–153 m) in the embayment
forest and 23 ± 70 m (0–970 m) on the plateau (Table
1). The median maximum distance between subsequent
capture locations averaged across all individuals was 31
± 23 m in the embayment forest and 54 ± 105 m on the
plateau. The observed range length was 39 ± 25 m (0–107
m) in the embayment forest and 61 ± 127 m (0–1,023 m)
on the plateau. The total minimum movement distance of
individuals summed across all their capture events was 83
± 100 m (range: 0–387 m) in the embayment forest and 140
± 617 m (0–8,022 m) on the plateau; however, due to the
unequal trapping effort in both time and space these basic
movement distances are not directly comparable between
the two habitats. Males showed generally longer and more
variable movements than females in both habitats, but this
effect was more pronounced on the plateau where much
longer movements could be recorded by the larger trap
network (Table 1). There was very little difference among
females that were recorded with or without signs of current
reproduction (Table 1). Of the rats recaptured at least
once, 8.4% were only captured in one trap location. With
the exception of one lactating female which was captured
nine times in the same trap location, all rats that were
captured >5 times moved between at least two different
trap locations.
Table 1 Median and standard deviation (sd) straight-line movement distances (m) and observed range lengths of Pacific
rats between live capture events during a spatial mark–recapture study on Henderson Island in May–October 2015.
Note that the trapping effort in the two habitats had a different spatial and temporal extent (see Fig. 1 for spatial extent
of trap locations). ‘breed’ females were classified if they had obvious signs of lactation or pregnancy.
Embayment forest
Parameter
males
median
n individuals
sd
non-breed
females
median
sd
Coral plateau
breed
females
median
sd
non-breed
females
males
median
sd
median
sd
breed females
median
sd
32
13
20
262
201
171
n captures
171
49
77
2010
1195
608
mean distance
between subsequent
captures (m)
17.5
17.9
20.7
19.6
13.6
21.5
27.2
80.4
20.9
53.0
21.6
57.0
maximum distance
between subsequent
captures (m)
36.7
23.3
31.7
23.2
18.4
21.9
60.7
117.6
49.8
91.1
46.3
85.8
observed range length
(m)
43.3
25.2
35.4
23.2
18.4
21.9
70.4
144.9
55.5
106.2
46.3
93.0
total minimum
distance travelled (m)
93.6
105.1
108.4 102.2
23.7
56.7
172.0
793.2
115.7
367.4
85.0
204.6
203
Island invasives: scaling up to meet the challenge. Ch 1C Rodents: Lessons
Seasonal variation in survival and space use
Based on the capture and recapture of 540 individual
rats on the plateau (including recapture in snap traps
in November), we found seasonal variation in monthly
survival probability (Fig. 2), but no evidence that survival
was influenced by sex (β = -0.15; 95% credible interval
-0.43–0.12) or NDVI (β = 0.44; -0.87–1.73). In June and
early July, the median monthly survival probabilities of
Pacific rats on the plateau were 0.794 (0.306–0.967) and
0.781 (0.471–0.933), respectively, but dropped to 0.353
(0.081–0.737) and 0.636 (0.488–0.763) in late July and
early August, respectively (Fig. 2). Remaining survivors
had very high survival in late August (0.950; 0.846–0.986)
and September (Fig. 2), despite persisting low NDVI (Fig.
S2). Similar estimates were obtained from 60 individual
rats in the embayment forest, with median monthly survival
probabilities of 0.361 (0.054–0.907) in early August and
0.864 (0.466–0.995) in September.
The survival estimates had very low precision due to
the potential for confounding emigration, because during
the times of lower mean survival probability, a larger
number of rats appeared to exhibit longer movements. Rat
movements were captured by two frequency distributions
(Fig. 3), with the majority of rats (79.1%) belonging to a
‘territorial’ type that exhibited home range radii between
100 and 200 m, and a smaller proportion (19.9% of males,
22.0% of females) belonging to a ‘roaming’ type with highly
variable and occasionally very long-distance movements
(Fig. 3). The proportion of captured rats belonging to the
roaming type increased from 0.8% in June to 13.8% in late
July (Table 2). In the embayment forest, we estimated only
marginally smaller home range radii as on the plateau in
early August (Table 2).
Besides large movements around a central point in their
territory, our model also indicated that, for rats that were
captured in two subsequent primary sessions, the central
point of their activity shifted by a median of 50 m (5–290
m) between early and late August, and by a median of 92 m
(4–378 m) between September and October (Fig. 4).
Fig. 2 Mean (95% credible interval) monthly survival
probability of Pacific rats on Henderson Island between
seven primary trapping sessions over five months in
2015 estimated from a spatial Cormack-Jolly-Seber
model. Note that survival probability is scaled over a
30-day period due to unequal time intervals between
primary trapping sessions.
median 50% utilization distribution (the core home range)
was 0.095 ± 0.08 ha (range 0.05–0.30 ha), and the 95%
utilization distribution (UD) was 0.55 ± 0.37 ha (range
0.21–1.58 ha). The minimum convex polygon home range
was more variable with a median of 0.36 ± 0.86 ha (range
0.003–2.99 ha). Rats used vegetation in the canopy or subcanopy during less than 20% of re-locations. There was no
relationship between the number of days a rat was tracked
(range: 7–54 days) and the size of its home range (MCP:
p = 0.11; 95% UD: p = 0.31). Thus, the estimates derived
from radio-tracking suggested much smaller rat home
range areas than those derived from spatially-explicit
mark-recapture models, which ranged from 2.88 to 931.6
ha for territorial rats on the plateau, and from 0.11 to 53.6
ha in the embayment forest, assuming that these rats used
a circular home range.
Home range sizes estimated from telemetry
Adequacy of cereal bait distribution during the 2011
eradication attempt
In 2013, we successfully tracked 19 rats (9 females,
10 males) between 1 July and 24 August with body mass
ranging from 29 to 107 g (median: 71 g, SD: 32 g). The
The lowest confidence limit for an estimated home
range for any season based on our spatial capture data was
2.88 ha on the plateau and 0.11 ha in the embayment forest.
Table 2 Home range radius (m) of two different behavioural types of Pacific rats on the coral plateau and
in the embayment forest of Henderson Island between June and October 2015, estimated from a spatial
mark–recapture model. Median estimated home range radius and lower (lcl) and upper (ucl) 95% credible
limits are given in m. ‘prop’ indicates the proportion of captured rats in a 10-day trapping session that
belonged to one of the behavioural types. Roaming rats could not be detected in the embayment forest.
prop
Plateau
June
early July
late July
early Aug
late Aug
Sept
Oct
Embayment early Aug
forest
late Aug
Sept
204
0.99
0.91
0.86
0.89
0.93
0.88
0.77
Residential rats
median
lcl
135
132
133
135
138
137
132
96
137
142
107
103
104
107
110
110
102
37
36
34
ucl
prop
162
161
162
165
167
171
171
228
377
382
0.01
0.09
0.14
0.11
0.07
0.12
0.23
Roaming rats
median
lcl
399
776
866
1,038
1,229
688
724
290
223
279
307
619
150
293
ucl
584
1,659
1,725
1,767
1,774
1,579
1,568
Oppel, et al.: Movement & survival of Pacific rats
Home ranges of this size would result in 15,988 toxic bait
pellets being available within a rat’s home range on the
plateau, and 2,456 in the embayment forest. Based on
radio-tracking, where the smallest 95% UD was 0.21 ha,
1,175 pellets would have been available in a rat’s home
range on the plateau, and 4,700 pellets in the embayment
forest.
DISCUSSION
We demonstrated that invasive Pacific rats on
Henderson Island exhibited substantial individual and
temporal variation in their movement and survival over
a five-month period. We found no evidence to suggest
that rats had home ranges that would have limited their
ability to encounter bait if bait was distributed with a
density similar to the 2011 eradication attempt. Indeed,
the movements and home range estimates that we obtained
were considerably higher than those of any other published
study on the same species (Table 3), including populations
that have been eradicated (Bramley, 2014b). The timing of
the failed eradication operation in mid/late August 2011
also appears to have been at a time of the year where we
recorded naturally low survival in 2015, and the seasonal
timing of the operation was likely appropriate if conditions
in 2011 followed a similar phenology as in 2015 (Fig. S2).
Monthly survival probability of Pacific rats varies
between 0.40 and 0.72 (Tamarin & Malecha, 1971;
Moller & Craig, 1987; Bunn & Craig, 1989), with an
expected life span around 8–10 months (Harrison, 1956;
Fig. 3 Histogram of the number of individual Pacific rats
having a home range of a radius estimated from a
spatial Cormack-Jolly-Seber based on mark–recapture
data from the coral plateau on Henderson Island during
seven primary trapping sessions in 2015. Vertical lines
indicate the population mean (solid) and 95% credible
interval (dashed) home range radius.
Fig. 4 Frequency of displacement distances of activity
centres of male and female Pacific rats on the interior
coral plateau of Henderson Island between seven
primary trapping sessions over five months in 2015
estimated from a spatial Cormack-Jolly-Seber model.
205
Island invasives: scaling up to meet the challenge. Ch 1C Rodents: Lessons
Bourliere, 1959). We estimated broadly similar median
monthly survival probabilities of 0.36–0.90 on Henderson
Island. However, previous estimates were mostly based
on raw recapture rates and did not account for recapture
probabilities, and our slightly higher estimates of survival
in June, early July, and late August may be due to our model
accounting for low recapture probability. A higher survival
probability of Pacific rats on Henderson Island might also
be expected given the absence of larger competitors (R.
rattus or R. norvegicus).
There was temporal fluctuation in survival probability
of Pacific rats in other tropical (Tamarin & Malecha, 1971)
and temperate island populations (Moller & Craig, 1987;
Bunn & Craig, 1989), and we found similar short-term
variability in survival on Henderson Island. We currently do
not understand what may have caused the temporal decline
in survival probabilities in July and August, and whether
such a reduction occurs predictably every year in response
to regular environmental events. As a sub-tropical island,
Henderson Island experiences only moderate fluctuations
in temperature and day length, which are unlikely to lead
to the same predictable population fluctuations as observed
on temperate islands (Russell & Holmes, 2015). The
changes in both survival and movement within our fivemonth study period on Henderson may have reflected a
period of resource shortage from late July to September
that may have induced higher mortality and emigration
as a larger proportion of rats belonged to the ‘roaming’
movement type. Assuming that the reduced survival that
we observed in 2015 was caused by resource limitation
(e.g. Russell & Ruffino, 2012), and that fluctuations in
resource availability and survival are similar among years
(Fig. S2), an operational timing in July or early August
may maximise the chances of eradication success.
Our spatial mark-recapture data on the plateau,
where traps were up to 1.5 km apart, revealed many
long movements by rats. These movements matched
or exceeded the previously estimated maximum travel
distance of 1,097 m or home range estimate of 3 ha for
Pacific rats (Lindsey, et al., 1973; Nass, 1977; Lindsey, et
al., 1999; Clapperton, 2006; Scheffler, et al., 2012), and
were similar to movements typically found in the much
larger Norway rat (R. norvegicus) (Clapperton, 2006;
Bramley, 2014b). Despite some long movements that we
recorded, the extrapolated ‘home range areas’ from our
spatial capture data are possibly biased high, because
these extrapolations are based on the assumption that rats
occupy a circular home range, which may not be the case
(Nass, 1977; Lindsey, et al., 1999; Clapperton, 2006). In
particular, our trails may have affected rat movement by
providing highly nutritious and palatable coconut bait
in traps that is otherwise not available on the plateau.
However, our trails were characterised by an absence
of vegetation between 30 to 250 cm above ground, and
probably did not materially affect the movement ability of
rats on the ground. Nonetheless, the maximum estimates
of home range area that we provide must be considered
with caution, as the areas actually exploited by rats may
be significantly smaller than the assumed circular radius
range.
Based on our estimates of movement behaviour from
radio-tracking in 2013 and spatial mark-recapture in 2015,
individual rats would have theoretically encountered
hundreds to thousands of bait pellets in their typical home
range, which would likely be sufficient for them to ingest
a lethal dose even if crab consumption gradually reduced
bait density over time (Cuthbert, et al., 2012). We therefore
consider it unlikely that the eradication failed because
individual rats did not have access to a sufficient quantity
of toxic bait, but uncertainty remains with respect to certain
life stages (e.g. nursing female rats and freshly weaned
pups): the number of rats surviving the 2011 operation
was very small, constituting <0.2% of the estimated rat
population (Amos, et al., 2016). An eradication operation
may fail if only a very small number of rats exhibit no
movement and would therefore not encounter a sufficient
quantity of bait. Of the 810 rats that we captured in
2015, 28% were never recaptured, and of those that were
Table 3 Summary of home range size (ha) estimates of Pacific rats (Rattus exulans) on subtropical and tropical islands
derived from either radio tracking (TR) or spatial capture–mark–recapture (CMR); type of estimate refers to minimum
convex polygon (MCP) or spatially-explicit capture recapture (SECR) and indicates what measure of uncertainty
(standard deviation, SD; range) is provided with the estimate.
Tracking
method
TR
Sex
n
Home range (ha)
Type of estimate
Reference
F
28
0.06 (0.01–0.18)
(Nass, 1977)
CMR
F
40
0.08 (0.01–0.48)
TR
M
6
0.14 ± 0.04
CMR
M
19-40
0.17 (0.01–0.73)
TR
TR
M
F
29
5
0.18 (0.01-1.21)
0.18 ± 0.05
MCP (range)
Mean minimum
(range)
MCP (mean ± SD)
Mean minimum
(range)
MCP (range)
MCP (mean ± SD)
TR
F+M
19
0.32 ± 0.38
MCP (mean ± SD)
This study
CMR
F+M
541
0.11 –931.6
SECR (range)
This study
Hilo, HI, USA
TR
F+M
26
1.73
Hilo, HI, USA
Kahanahaiki, HI,
USA
TR
F+M
3
3
Circle with radius
mean distance from (Lindsey, et al., 1973)
burrow
MCP (mean ± SD) (Lindsey, et al., 1999)
TR
Unk
1
1.8
95% kernel
Location
Hilo, HI, USA
Green Island, Kure
Atoll, HI, USA
Kapiti Island, NZ
Green Island, Kure
Atoll, HI, USA
Hilo, HI, USA
Kapiti Island, NZ
Henderson Island,
Pitcairn Islands
Henderson Island,
Pitcairn Islands
206
(Wirtz, 1972)
(Bramley, 2014b)
(Wirtz, 1972)
(Nass, 1977)
(Bramley, 2014b)
(Shiels, 2010)
Oppel, et al.: Movement & survival of Pacific rats
recaptured at least once, 8% were only captured in a single
location. Because we did not record any movement for
a greater proportion of rats than the estimated surviving
population in 2011, it is theoretically possible that there are
some individuals that move very little or move very little
for a short period of time during which bait is available
on the ground. Unfortunately, the probability of detecting
a non-moving phenotype that exists with a prevalence of
<0.2% in the population is virtually zero for any practically
feasible sample size.
Bramley, G.N. (2014b). ‘Home ranges and interactions of kiore (Rattus
exulans) and Norway rats (R. norvegicus) on Kapiti Island, New
Zealand’. New Zealand Journal of Ecology 38: 328–334.
In summary, the rat eradication attempt on Henderson
Island in 2011 failed to kill all individuals, and our work
provides new knowledge to evaluate the potential causes
of this failure. An eradication failure can occur if (i) not
all individuals had access to sufficient bait; (ii) not all
individuals died despite consuming bait; or (iii) not all
individuals consumed a lethal dose of bait despite having
access (Holmes, et al., 2015). We have shown that the timing
of the operation was appropriate and that it is unlikely that
rats did not have access to sufficient bait. Previous work
confirmed that rats remain susceptible to brodifacoum,
suggesting that toxicological resistance is an unlikely
cause of the 2011 eradication failure (Amos, et al., 2016).
A combination of factors leading to high alternative food
availability and a small number of rats preferring natural
food sources and disregarding bait may have resulted in
the failure of the eradication attempt in 2011, and further
research is required to examine whether that risk can be
reduced for a new eradication attempt.
Brooks, S.P. and Gelman, A. (1998). ‘General methods for monitoring
convergence of iterative simulations’. Journal of Computational and
Graphical Statistics 7: 434–455.
ACKNOWLEDGEMENTS
Brooke, M. de L. and Towns, D. (2008). A Feasibility Study for the
Eeradication of Kiore Rattus exulans from Henderson Island. Royal
Society for the Protection of Birds, Sandy, UK.
Brooke, M. de L., O’Connell, T.C., Wingate, D., Madeiros, J., Hilton,
G.M. and Ratcliffe, N. (2010). ‘Potential for rat predation to cause
decline of the globally threatened Henderson petrel Pterodroma atrata:
Evidence from the field, stable isotopes and population modelling’.
Endangered Species Research 11: 47–59.
Broome, K., Cox, A., Golding, C., Cromarty, P., Bell, P. and McClelland,
P. (2014). Rat Eradication Using Aerial Baiting. Current Agreed Best
Practice Used in New Zealand. Internal document DOC-29396. New
Zealand Department of Conservation, Wellington, New Zealand.
Bunn, T.J. and Craig, J.L. (1989). ‘Population cycles of Rattus exulans:
Population changes, diet, and food availability’. New Zealand Journal
of Zoology 16: 409–418.
Calenge, C. (2006). ‘The package “adehabitat” for the R software: A
tool for the analysis of space and habitat use by animals’. Ecological
Modelling 197: 516–519.
Clapperton, B.K. (2006). A Review of the Current Knowledge of Rodent
Behaviour in Relation to Control Devices. Wellington, NZ: Department
of Conservation.
Cromarty, P.L., Broome, K.G., Cox, A., Empson, R.A., Hutchinson,
W.M. and McFadden, I. (2002). ‘Eradication planning for invasive
alien animal species on islands – the approach developed by the New
Zealand Department of Conservation’. In: C.R. Veitch and M.N. Clout
(eds.) Turning the tide: the eradication of invasive species, pp. 85–91.
Occasional Paper SSC no. 28. IUCN SSC Invasive Species Specialist
Group. IUCN, Gland, Switzerland and Cambridge, UK.
We thank the Government of the Pitcairn Islands
for permission to work on Henderson Island; M. de L.
Brooke, L. MacKinnon, A. Forrest, A. Skinner, N. Torr,
S. O’Keefe, and P. Warren for assistance in the field;
and J. Vickery, J. Hall, A. Schofield, and C. Stringer for
general support. The crews of the Braveheart, Claymore
II, Teba, and Xplore, provided transportation to and from
Henderson Island. We thank A. Hughes for performing
the triangulation of rat telemetry relocations. The David
and Lucile Packard Foundation, Darwin Plus: Overseas
Territories Environment and Climate Fund, British Birds,
generous donors, and the RSPB, the UK partner in Birdlife
International, helped to fund our research. Scientific and
ethical approval was granted by the Government of the
Pitcairn Islands, and the RSPB Council (paper 2/13/62 and
protocol EAC 2015/01). Comments from M. de L. Brooke,
and two anonymous reviewers improved this manuscript.
Cunningham, D.M. and Moors, P.J. (1996). Guide to the Identification
and Collection of New Zealand Rodents, 2nd edition, Wellington, NZ:
Department of Conservation.
REFERENCES
Gilroy, J.J., Virzi, T., Boulton, R.L. and Lockwood, J.L. (2012). ‘A new
approach to the “apparent survival” problem: Estimating true survival
rates from mark–recapture studies’. Ecology 93: 1509–1516.
Amos, W., Nichols, H.J., Churchyard, T. and Brooke, M. de L. (2016).
‘Rat eradication comes within a whisker! A case study from the South
Pacific’. Royal Society Open Science 3: 160110.
Atkinson, I.A.E. (1985). ‘The Spread of Commensal Species of Rattus to
Oceanic Islands and their Effects on Island Avifaunas.’. In: P.J. Moors
(ed.) Conservation of Island Birds. pp. 35–81. Cambridge, United
Kingdom: International Council for Bird Preservation.
Bond, A.L., Churchyard, T., Cuthbert, R.J., Donaldson, A., Duffield,
N., Forrest, A., Havery, S., Lavers, J.L., MacKinnon, L., McClelland,
G.T.W., Kelly, J., Proud, T., Skinner, A., Torr, N., Vickery, J.A. and
Oppel, S. (2019). ‘Recovery of introduced Pacific rats (Rattus exulans)
following a failed eradication attempt on Subtropical Henderson Island,
South Pacific Ocean’. In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C.
Russell and C.J. West (eds.) Island invasives: scaling up to meet
the challenge, pp. 167–174. Occasional Paper SSC no. 62. Gland,
Switzerland: IUCN.
Bourliere, F. (1959). ‘Lifespans of Mammalian and Bird Populations
in Nature’. In: G.E.W. Wolstenholme and M. O’Connor (eds.) The
Lifespan of Animals, pp. 90–102. London: J & A Churchill.
Cuthbert, R.J., Brooke, M. de L. and Torr, N. (2012). ‘Overcoming
hermit-crab interference during rodent-baiting operations: A case study
from Henderson Island, South Pacific’. Wildlife Research 39: 70–77.
Dawson, J., Oppel, S., Cuthbert, R.J., Holmes, N., Bird, J.P., Butchart,
S.H.M., Spatz, D.R. and Tershy, B. (2015). ‘Prioritizing islands for the
eradication of invasive vertebrates in the United Kingdom overseas
territories’. Conservation Biology 29: 143–153.
Efford, M. (2016). ‘secr – Spatially Explicit Capture–recapture in R. R
Package Version 2.10.4’. <https://CRAN.R-project.org/package=secr>.
Ergon, T. and Gardner, B. (2014). ‘Separating mortality and emigration:
Modelling space use, dispersal and survival with robust-design spatial
capture–recapture data’. Methods in Ecology and Evolution 5: 1327–
1336.
Gardner, B., Reppucci, J., Lucherini, M. and Royle, J.A. (2010). ‘Spatially
explicit inference for open populations: Estimating demographic
parameters from camera-trap studies’. Ecology 91: 3376–3383.
Harper, G.A., van Dinther, M., Russell, J.C. and Bunbury, N. (2015).
‘The response of black rats (Rattus rattus) to evergreen and seasonally
arid habitats: Informing eradication planning on a tropical island’.
Biological Conservation 185: 66–74.
Harrison, J. (1956). ‘Survival rates of Malayan rats’. Bulletin of the
Raffles Museum 27: 5–26.
Holmes, N.D., Griffiths, R., Pott, M., Alifano, A., Will, D., Wegmann,
A. and Russell, J.C. (2015). ‘Factors associated with rodent eradication
failure’. Biological Conservation 185: 8–16.
Howald, G., Donlan, C.J., Galvan, J.P., Russell, J.C., Parkes, J.,
Samaniego, A., Wang, Y.W., Veitch, D., Genovesi, P., Pascal, M.,
Saunders, A. and Tershy, B. (2007). ‘Invasive rodent eradication on
islands’. Conservation Biology 21: 1258–1268.
Jones, H.P. (2010). ‘Seabird islands take mere decades to recover
following rat eradication’. Ecological Applications 20: 2075–2080.
Bramley, G.N. (2014a). ‘Habitat use by kiore (Rattus exulans) and
Norway rats (R. norvegicus) on Kapiti Island, New Zealand’. New
Zealand Journal of Ecology 38: 64–75.
207
Island invasives: scaling up to meet the challenge. Ch 1C Rodents: Lessons
Jones, H.P., Holmes, N.D., Butchart, S.H.M., Tershy, B.R., Kappes,
P.J., Corkery, I., Aguirre-Muñoz, A., Armstrong, D.P., Bonnaud, E.,
Burbidge, A.A., Campbell, K., Courchamp, F., Cowan, P.E., Cuthbert,
R.J., Ebbert, S., Genovesi, P., Howald, G.R., Keitt, B.S., Kress, S.W.,
Miskelly, C.M., Oppel, S., Poncet, S., Rauzon, M.J., Rocamora, G.,
Russell, J.C., Samaniego-Herrera, A., Seddon, P.J., Spatz, D.R., Towns,
D.R. and Croll, D.A. (2016). ‘Invasive mammal eradication on islands
results in substantial conservation gains’. Proceedings of the National
Academy of Sciences 113: 4033–4038.
Keitt, B., Griffiths, R., Boudjelas, S., Broome, K., Cranwell, S., Millett,
J., Pitt, W. and Samaniego-Herrera, A. (2015). ‘Best practice guidelines
for rat eradication on tropical islands’. Biological Conservation 185:
17–26.
Kellner, K.F. (2016). ‘jagsUI: A Wrapper Around rjags to Streamline
JAGS Analyses. R Package Version 1.4.2‘. <https://CRAN.R-project.
org/package=jagsUI>.
Kenward, R. (2001). A Manual for Wildlife Radio Tagging, London:
Academic Press.
Lindsey, G.D., Nass, R.D., Hood, G.A. and Hirata, D.N. (1973).
‘Movement patterns of Polynesian rats (Rattus exulans) in sugarcane’.
Pacific Science 27: 239–246.
Lindsey, G.D., Mosher, S.M., Fancy, S.G. and Smucker, T.D. (1999).
‘Population structure and movements of introduced rats in an Hawaiian
rainforest’. Pacific Conservation Biology 5: 94–102.
Moller, H. and Craig, J.L. (1987). ‘The population ecology of Rattus
exulans on Tiritiri Matangi Island, and a model of comparative
population dynamics in New Zealand’. New Zealand Journal of
Zoology 14: 305–328.
Nass, R.D. (1977). ‘Movements and home ranges of Polynesian rats in
Hawaiian sugarcane’. Pacific Science 31: 135–142.
Nathan, H.W., Clout, M.N., MacKay, J.W.B., Murphy, E.C. and Russell,
J.C. (2015). ‘Experimental island invasion of house mice’. Population
Ecology 57: 363–371.
Paulay, G. and Spencer, T. (1989). ‘Vegetation of Henderson Island’. Atoll
Research Bulletin 328: 1–20.
Pettorelli, N., Ryan, S., Mueller, T., Bunnefeld, N., Jędrzejewska,
B., Lima, M. and Kausrud, K. (2011). ‚The normalized difference
vegetation index (NDVI): unforeseen successes in animal ecology’.
Climate Research 46: 15–27.
Pettorelli, N., Laurance, W.F., O’Brien, T.G., Wegmann, M., Nagendra, H.
and Turner, W. (2014). ‘Satellite remote sensing for applied ecologists:
opportunities and challenges’. Journal of Applied Ecology 51: 839–848.
Plummer, M. (2012). ‚JAGS Version 3.3.0‘. <http://sourceforge.net/
projects/mcmc-jags/files/Manuals/>.
Pradel, R., Hines, J.E., Lebreton, J.D. and Nichols, J.D. (1997). ‘Capture–
recapture survival models taking account of transients’. Biometrics 53:
60–72.
Püttker, T., de Barros, C.d.S., Martins, T.K., Sommer, S. and Pardini, R.
(2012). ‘Suitability of distance metrics as indexes of home-range size in
tropical rodent species’. Journal of Mammalogy 93: 115–123.
R Core Team (2016). R: A Language and Environment for Statistical
Computing. Vienna, Austria: R Foundation for Statistical Computing.
Raabe, J.K., Gardner, B. and Hightower, J.E. (2013). ‘A spatial capture–
recapture model to estimate fish survival and location from linear
continuous monitoring arrays’. Canadian Journal of Fisheries and
Aquatic Sciences 71: 120–130.
Ringler, D., Russell, J., Jaeger, A., Pinet, P., Bastien, M. and Le Corre, M.
(2014). ‘Invasive rat space use on tropical islands: Implications for bait
broadcast’. Basic and Applied Ecology 15: 179–186.
Roberts, M. and Craig, J. (1990). ‘The demography of kiore, Rattus
exulans in three habitats’. New Zealand Journal of Zoology 17: 43–53.
Royle, J.A., Fuller, A.K. and Sutherland, C. (2016). ‘Spatial capturerecapture models allowing Markovian transience or dispersal’.
Population Ecology 58: 53–62.
Russell, J.C. and Ruffino, L. (2012). ‘The influence of spatio-temporal
resource fluctuations on insular rat population dynamics’. Proceedings
of the Royal Society B: Biological Sciences 279: 767–774.
Russell, J.C., Caut, S., Anderson, S.H. and Lee, M. (2015). ‘Invasive rat
interactions and over-invasion on a coral atoll’. Biological Conservation
185: 59–65.
Russell, J.C. and Holmes, N.D. (2015). ‘Tropical island conservation: rat
eradication for species recovery’. Biological Conservation 185: 1–7.
208
Schaub, M. and Royle, J.A. (2014). ‘Estimating true instead of apparent
survival using spatial Cormack–Jolly–Seber models’. Methods in
Ecology and Evolution 5: 1316–1326.
Scheffler, P.Y., Foote, D., Forbes Perry, C., Schlappa, K. and Stone,
C.P. (2012). Population Dynamics of Introduced Rodents in Hawai’i
Volcanoes National Park 1986–1990. Honolulu, HI: Pacific Cooperative
Studies Unit, University of Hawai’i.
Shiels, A.B. (2010). ‘Ecology and Impacts of Introduced Rodents (Rattus
spp. and Mus musculus) in the Hawaiian Islands’. PhD thesis. Honolulu,
HI: University of Hawai’i.
Spencer, T. (1995). ‘The Pitcairn Islands, south Pacific Ocean: Plate
tectonic and climatic contexts’. Biological Journal of the Linnean
Society 56: 13–42.
Steadman, D.W. and Olson, S.L. (1985). ‘Bird remains from an
archaeological site on Henderson Island, South Pacific: Man-caused
extinctions on an “uninhabited” island’. Proceedings of the National
Academy of Sciences 82: 6191–6195.
Stickel, L.F. (1954). ‘A comparison of certain methods of measuring
ranges of small mammals’. Journal of Mammalogy 35: 1–15.
Tamarin, R.H. and Malecha, S.R. (1971). ‘The population biology of
Hawaiian rodents: Demographic parameters’. Ecology 52: 383–394.
Torr, N. and Brown, D. (2012). Henderson Island restoration project
post-operational report. Sandy, UK: Royal Society for the Protection
of Birds.
Waldren, S., Florence, J. and Chepstow-Lusty, A.J. (1995). ‘A comparison
of the vegetation communities from the islands of the Pitcairn group’.
Biological Journal of the Linnean Society 56: 121–144.
Weigelt, P., Jetz, W. and Kreft, H. (2013). ‘Bioclimatic and physical
characterization of the world’s islands’. Proceedings of the National
Academy of Sciences 110: 15307–15312.
Wirtz, W.O.I. (1972). ‘Population ecology of the Polynesian rat, Rattus
exulans, on Kure Atoll, Hawaii’. Pacific Science 26: 433–464.
Yackel Adams, A.A., Stanford, J.W., Wiewel, A.S. and Rodda, G.H.
(2011). ‘Modelling detectability of kiore (Rattus exulans) on Aguiguan,
Mariana Islands, to inform possible eradication and monitoring efforts’.
New Zealand Journal of Ecology 35: 145–152.
Zavaleta, E.S. (2002). ‘It’s often better to eradicate, but can we eradicate
better?’. In: C.R. Veitch and M.N. Clout (eds.) Turning the tide: the
eradication of invasive species, pp. 393–404. Occasional Paper SSC
no. 28. IUCN SSC Invasive Species Specialist Group. IUCN, Gland,
Switzerland and Cambridge, UK.
A. Samaniego-Herrera, S. Boudjelas, G.A. Harper, and J.C. Russell
Samaniego-Herrera, A.; S. Boudjelas, G.A. Harper and J.C. Russell. Assessing the critical role
that land crabs play in tropical island rodent eradications and ecological restoration
Assessing the critical role that land crabs play in tropical island
rodent eradications and ecological restoration
A. Samaniego-Herrera1,2, S. Boudjelas2,3, G.A. Harper4 and J.C. Russell1
School of Biological Sciences, University of Auckland, Private Bag 92019, Auckland 1142, New Zealand.
<ara.samaniego.mx@gmail.com>. 2Pacific Invasives Initiative, University of Auckland, Private Bag 92019, Auckland
1142, New Zealand. 3Centre for Biodiversity and Biosecurity, University of Auckland, Private Bag 92019, Auckland
1142, New Zealand. 4Biodiversity Restoration Specialists, PO Box 65, Murchison, New Zealand.
1
Abstract Invasive rodent eradications are one of the most effective conservation interventions to restore island ecosystems.
However, achievements in the tropics are lagging behind those in temperate regions. Land crab interference in bait uptake
has been identified as one of the main causes of rodent eradication failure on tropical islands, but the issue of effective
mitigation of bait loss due to land crab consumption is poorly understood. For example, there are over 100 species of
land crab and each may behave differently. We reviewed the available literature to answer: (1) which crab species are the
most problematic? (2) what mitigation measures have been effective? and (3) how do invasive rodents impact land crab
communities? We analysed a systematic dataset from six tropical islands to test two hypotheses: (a) bait uptake is highest
when burrowing (Brachyura) land crabs are present; and (b) small land crabs (including juveniles of the larger species)
are highly vulnerable to rodent predation. We found that large species (e.g. genera Cardisoma, Johngarthia and Birgus)
are the most problematic during rodent eradications. Effective mitigation measures to prevent bait loss include using
higher bait application rates and conducting eradications during the driest months. Land crab communities tend to go
through significant changes after rodent removal. From our analyses, we confirmed pre-eradication data are valuable for
eradication planning, as seasonality and type of crab can influence outcomes. Post-eradication data confirmed small crab
species (<60 mm) are highly vulnerable to rodent predation. More effort should be invested into monitoring land crabs
in tropical latitudes, particularly to determine any biogeographic or taxon trends in land crab interference. Land crabs are
key for the restoration of the islands, as they shape ecosystems through their role as ecosystem engineers, hence they are
excellent indicators of ecosystem recovery. Our results will contribute to the better planning of future rodent eradications
on tropical islands where land crabs are significant bait competitors.
Keywords: Birgus, Cardisoma, Coenobita, Gecarcinus, impacts, monitoring, Mus, Rattus
INTRODUCTION
Islands are some of the most important repositories for
biodiversity, with 15–20% of terrestrial biodiversity held
on only 5.3% of the world’s land area (Weigelt, et al., 2013).
Tropical islands are particularly important due to their high
levels of endemism (Myers, et al., 2000). Island species
are also highly vulnerable to anthropogenic impacts, of
which invasive alien species (IAS) introductions are often
the most severe (Russell, et al., 2017), causing 86% of
island endemic species extinctions (Bellard, et al., 2016).
Moreover, IAS also interrupt ecosystem functioning
through predation of, and competition with, other biotic
components (Athens, et al., 2002; Towns, et al., 2006;
Hilton & Cuthbert, 2010).
Over the past 50 years, eradication of IAS has been
increasing (Towns, et al., 2013; Jones, et al., 2016), with
the removal of invasive rodents proving highly effective in
targeted species recovery and island ecosystem restoration
(Le Corre, et al., 2015; Croll, et al., 2016). Over 90%
of rat eradication attempts have been successful, with
increasingly larger islands being effectively targeted
(Holmes, et al., 2015). However, the rate of eradication
failure on tropical islands has been 2–2.5 times higher
than on temperate islands (Russell & Holmes, 2015).
This discrepancy is due to several contributing factors
(Holmes, et al., 2015). Probably the most significant are
the benign climate facilitating rodent reproduction (Harper
& Bunbury, 2015), and bait competition from abundant
land crabs (Wegmann, 2008; Griffiths, et al., 2011).
Land crabs comprise over a hundred species in three
broad groups, burrowing crabs, hermit crabs and coconut
crabs, although the latter single species (Birgus latro) is
technically a hermit crab. As the largest invertebrates on
islands, particularly coral atolls, land crabs are often the
apex land predator (Burggren & McMahon, 1988), and can
attain high population densities and occupy the niches of
vertebrates on small oceanic islands. As such, they act as
allogenic ecosystem engineers (Green, et al., 2008; Paulay
& Starmer, 2011) through their significant influence on
forest structure, plant species composition, soil formation
and nutrient transfer and cycling (Green, et al., 1999;
Sherman, 2002; Gutiérrez & Jones, 2006; Gutiérrez, et
al., 2006; Sherman, 2006; Green, et al., 2008; Lindquist,
et al., 2009). As keystone consumers (Paine, 1966),
the removal of or reduction in crab abundance through
the introduction of IAS can trigger a trophic cascade of
effects, leading to ‘meltdown’ in island ecosystems in the
worst cases (O’Dowd, et al., 2003; Pitman, et al., 2005;
Nigro, et al., 2017). Moreover, as smaller crab species
in particular are vulnerable to predation by rodents (St
Clair, 2011; Samaniego-Herrera & Bedolla-Guzmán,
2012) and invasive rodents are found on >80% of island
groups (Atkinson, 1985), an improved understanding of
the interaction between rodents and land crabs is urgently
required. However, land crabs have rarely been monitored
before and after rodent eradications (but see Nigro, et al.,
2017), and basic tools such as inventories are lacking for
most tropical islands where rodent eradications are being
planned.
The Pacific Invasives Initiative (PII) commissioned the
first review on land crab interference in rodent eradications
about 10 years ago (Wegmann, 2008) and many lessons
have been learnt since. To improve the justification and
implementation of rodent eradications on tropical islands,
we conducted literature reviews on two main topics: the
role of land crabs in invasive rodent eradications and the
vulnerability of land crabs to rodent invasion. A case study
from six tropical islands is presented, demonstrating the
utility of monitoring land crabs both pre- and post-rodent
eradications. Based on our previous observations of land
crabs across islands, we expected (a) bait uptake to be
highest on the islands where large burrowing species were
abundant, and (b) population abundance of small burrowing
species to increase over time after rodent eradications.
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 209–222. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
209
Island invasives: scaling up to meet the challenge. Ch 1C Rodents: Lessons
METHODS
Land crabs and rodent eradications
Rodent impacts on land crabs
Following Burggren & McMahon (1988), we consider
land crabs to be crabs that show significant behavioural,
morphological, physiological, or biochemical adaptations
permitting extended activity out of water. This includes a
few families of the diverse infraorders Anomura (hermit
crabs) and Brachyura (burrowing crabs), yet there are over
a hundred species that can be considered land crabs. Land
crab distribution ranges from tropical to subtropical areas,
hence the scope of this paper focuses on islands located
between ~25° north and south of the equator. We also focus
on the two most common rodent eradication methods:
aerial and hand broadcast of bait directly onto the ground
(Howald, et al., 2007; DIISE, 2016).
Some of the information on the impacts of rodents
on land crabs was collated from the project documents
mentioned above. In addition, we also searched the Web
of Science, Scopus and Google Scholar for published
literature using keywords: [island OR atoll OR cay OR
archipelago] AND [rodent OR rat OR rattus OR mus] AND
[“land crab” OR invertebrate]. We collated information on
impacts through review of the resulting publications and
relevant references listed in these.
The islands included in the review are a subset from
the Database of Island Invasive Species Eradications
(DIISE, 2016). These were selected based on the following
criteria: 1) location: between latitudes ~25° north and
south of the equator, 2) target IAS taxa: Muridae, 3) whole
island eradications, 4) toxicant used: 2nd generation
anticoagulant, 5) main bait delivery method: hand or
aerial broadcast, 6) quality of data: good or satisfactory,
the latter were updated to good, and 7) eradication status:
known or ‘to be confirmed’, the latter were updated to
failed or successful. Islands without land crabs such as
the Galapagos Islands and Western Australia islands were
excluded.
For each island, we collated the following additional
data: bait rates used during the rodent eradications,
island type (savanna, tropical seasonal forest or tropical
rainforest), presence/absence status and abundance for
each land crab group (hermit, coconut, burrowing),
land crab group identified as the main bait competitor
and timing of the eradication (dry or wet season). This
information was collated through review of project
documents (i.e. feasibility studies, operational plans, postoperation reports and scientific papers). We also sought
inputs from project managers when we required further
clarification/confirmation or information was missing from
the documents available. Given the scarcity of scholarly
information on land crabs, and the lack of a single source
with the basic biology and current taxonomy for all land
crabs (as most crab species are marine), we conducted an
additional literature review to compile such information.
A 2-way ANOVA test for unbalanced designs was used
to evaluate the variations in bait rates in relation to island
type and main bait competitors. Data were log-transformed
to achieve normality. All analyses were performed in R 3.4.
Fig. 1 Dominant vegetation on the islands where the case
study took place.
210
Case study: Mexican tropical islands
Study sites
As part of a wider restoration programme led by
Grupo de Ecología y Conservación de Islas (GECI)
(Samaniego-Herrera, et al., 2011), bait uptake and land
crab monitoring was conducted on six Mexican tropical
islands. The islands, three in the Gulf of Mexico, one in
the Mexican Pacific, and two in the Caribbean Sea fall
into the three categories of tropical island ecosystems
described by Russell and Holmes (2015): savanna, tropical
seasonal forest and tropical rainforest, respectively (Table
1; Fig. 1). The aims of the monitoring were to inform
the specific rodent eradication plans by assessing the
potential interference of each land crab community, and to
compare such communities before and after the removal
of the invasive rodents. Invasive rodents (Table 1) were
successfully eradicated from all islands either by hand or
aerial broadcast of bait (Samaniego-Herrera, et al., 2014;
Samaniego-Herrera, et al., 2018), following international
best practices (Keitt, et al., 2015).
Bait uptake
Two types of bait were used: placebo bait for preeradication assessments and toxic bait for the actual rodent
eradications. The toxic bait consisted of 2 g cereal bait
pellets containing 25 ppm brodifacoum (second generation
anticoagulant), manufactured by Bell Labs. The placebo
bait, also from Bell Labs, was identical but non-toxic. Total
bait uptake (i.e. by the target and non-target species) was
measured before and during each eradication operation.
Pre-eradication, bait uptake was monitored to help decide
application rates for the eradication. During eradications,
the monitoring took place to (a) validate the intended bait
rate, by estimating bait density on the ground immediately
after the bait drops, (b) assess the daily uptake rate, by
repeating measurements every 24 hours, and (c) investigate
the relationship of bait uptake rate, rodent abundance, and
land crab diversity and abundance, by combining results
from different islands.
In all cases, bait uptake was measured daily for 6–10
consecutive days in a systematic way, starting on the same
day of bait broadcast. For all pre-eradication studies bait
was broadcast by hand, whereas for the eradications either
aerial or hand broadcast was used (Table 1). Bait uptake
was measured in fixed circular plots as described by Pott,
et al. (2015). A subset of the resulting dataset was included
in the meta-analysis by Pott, et al. (2015), which showed
the utility of bait availability studies. However, there are
three major differences with the present study. Firstly, Pott
and colleagues only used a subset of the Mexican dataset
due to the limited data available for the other islands
(e.g. data from only two of the 6–10 days available were
analysed). Secondly, the results presented here derived
from standardised monitoring methodologies. Lastly, our
focus is to investigate the role of land crabs in the overall
bait uptake in more detail, distinguishing for crab type
(hermit and burrowing).
Samaniego-Herrera, et al.: Critical role of land crabs
Table 1 General description of the six Mexican islands where land crabs were monitored before and after the successful
rodent eradications.
Archipelago
Island
Area
(ha)
Ecosystem
type
Dominant
vegetation1
Species
eradicated
(year)2
Eradication
method and
bait rate (total
kg/ha)
Main bait competitors
(seasonal fluctuation)
3
Savanna
Shrubs and
grasses
Mus musculus
(2011)
Hand broadcast
17 kg/ha
Hermit crabs
(low fluctuation)
Pérez
13
Savanna
Shrubs
Rattus rattus
(2011)
Hand broadcast
17 kg/ha
Hermit crabs
(low fluctuation)
Muertos
15
Savanna
Shrubs
Mus musculus
(2011)
Hand broadcast
17 kg/ha
Small burrowing crabs
(low fluctuation)
Isabel
82
Tropical
seasonal
forest
Deciduous
forest
Rattus rattus
(2009)
Aerial broadcast
20 kg/ha
Large burrowing crabs3
(high fluctuation)
Arrecife Alacranes
Pájaros
Banco Chinchorro
Cayo Norte
Mayor
30
Tropical
rainforest
Mangroves
& evergreen
forest
Rattus rattus
(2012)
Aerial broadcast
42 kg/ha
Large burrowing crabs
(moderate fluctuation)
Cayo Centro
539
Tropical
rainforest
Mangroves
& evergreen
forest
Rattus rattus
(2015)
Aerial broadcast
60 kg/ha
Large burrowing crabs
(moderate fluctuation)
See Fig. 1.
Always end of dry season.
3
There was virtually no crab interference during the rat eradication; bait lasted for weeks.
Sources: Samaniego-Herrera, et al., 2013; Samaniego-Herrera, et al., 2014; Samaniego-Herrera, et al., 2018.
1
2
Land crab recovery
On all islands, land crab activity was monitored twice
a year, at the end of each dry and wet season, both before
(2–3 years) and after (1–5 years) each rodent eradication.
Every season, several (6–18) fixed plots (25 m × 2 m)
were used to estimate crab density; the exception was on
Isabel Island, where two 300 m × 6 m plots were used.
Plots were walked for 3–5 consecutive nights. One person
with a headlamp walked in the middle of the plot, starting
one hour after sunset. In order to walk all plots within 90
minutes (i.e. the peak activity period), several observers
participated each night on some islands. The number of
land crabs, by species, was recorded. Minimum training
is required to carry out this task given the morphological
differences of the species present.
Data analysis
Bait uptake trends (always measured in the dry season)
were investigated using a linear mixed model (R software
package nlme) for bait availability, where the density
of bait (kg/ha), using the difference of target density
minus measured density as the response variable (i.e.
comparing rates of decline rather than actual densities),
was dependent on time (days) and interactions of time with
fixed effect covariates (i.e. covariates which would affect
bait availability). These fixed effects included whether
the study was conducted prior to or during the eradication
(distinguishing between first and second bait application),
whether rats or mice were the target species and how
abundant they were (according to local mark-recapture
studies by Samaniego-Herrera (2014)), and the abundance
of both types of land crabs: hermit and burrowing (low
– high, based on the monitoring in this study, therefore
standardised). Although each island used different bait
application rates, we were specifically interested in the
rates of decline in bait availability. Inter-island differences
were accounted for by including island as a random effect
in our model. Diagnostic plots were visually checked for
violations of model assumptions.
For land crab activity, an index of density was estimated
as the number of nocturnal surface-active crabs per hectare.
First, we used 2-way ANOVA to test the difference in density
between seasons (dry and wet) and islands only for the preeradication periods (i.e. avoiding potential confounding
effects caused by the eradications), as obvious fluctuations
were occurring at least on some islands. In order to
compare trends during favourable periods (hence closer to
real density, as inactive crabs typically bury themselves),
and given that the lower land crab activity during the dry
season was confirmed for some islands, further analysis
comparing pre- and post-eradication density used data
from wet seasons only. Differences in density among island
types (savanna, seasonal, rainforest) and periods (pre- and
post-eradication) were tested with linear models. Data
were log-transformed to achieve normality. All analyses
were performed in R 3.4.
RESULTS
Land crabs and rodent eradications
The resulting database contains 108 eradication
attempts spread over 101 tropical islands (Appendix 1;
detailed spreadsheet: www.pacificinvasivesinitiative.org/).
On some islands, there were two eradication attempts
targeting a single rodent species or there were two rodent
species being targeted by a single eradication attempt.
Island sizes range from 0.1 ha to 4,310 ha (median = 10 ha).
Most attempts (86.1%) targeted only rats (Rattus exulans,
R. norvegicus, R. rattus or R. tanezumi), 2.8% only mice
(Mus musculus) and 11.1% targeted both rats and mice.
211
Island invasives: scaling up to meet the challenge. Ch 1C Rodents: Lessons
Eighty-nine (82.4%) of the eradication attempts
were successful and 19 (17.6%) failed (Table 2). Land
crab interference was reported as important in 56.2% of
the successful attempts and in 100% of the failed ones.
However, for the latter, in addition to land crab interference
other potential factors that may have contributed to the
failure were also reported. Examples of such factors are
gaps in bait coverage, which in turn can increase in area as
land crabs take bait at the edges.
Over 82% of the eradication attempts used higher bait
rates (x̄ = 25.7 kg/ha, range: 3–163 kg/ha) compared to those
typically used on temperate islands (12 kg/ha; Broome, et
al., 2014). In all cases, the justification for using higher
bait rates was high rodent abundances (either estimated or
assumed) and land crab presence, although abundance of
either was rarely quantified. On some islands, additional
factors such as high abundance of small invertebrates
(e.g. ants and cockroaches, also bait consumers) were also
mentioned.
Land crab abundances were reported as having
been estimated either through measurements (21.5%)
or observations (55.9%) during the planning phase; the
rest (22.6%) of the cases did not try to estimate land
crab abundance. On most islands, land crabs have been
identified to the genus level. Through our research on
current taxonomy, we identified 165 species of land crabs
in 52 genera and 15 families, of which seven genera have
been reported as important bait consumers (Table 3). For
most islands (90.7%), only three or fewer land crab species
were reported to be present.
Considering all islands, the 2-way ANOVA test
revealed significant differences in total bait rates used
depending on which type of land crab was the main source
of interference (F = 11.33, p<0.001) and on the interaction
between crab type and island type (F = 3.65, p<0.001),
whereas island type was marginally significant (F = 3.02, p
= 0.05). Higher bait rates (17–163 kg/ha) were used when
burrowing crabs were the main bait competitors (n= 8),
followed by hermit crabs (3–83.3 kg/ha; n= 55) and cases
with ‘no interference’ (8–33.2 kg/ha; n= 39), i.e. low crab
abundance (Fig. 2). When considering only successful
attempts, the patterns remained the same.
Rodents impacts on land crabs
Accounts of insular land crab populations being
negatively impacted by invasive rodents included 15
populations of ten species across nine countries and
overseas territories (Table 4). These impacts are mainly
in the form of population suppression and ecological
Fig. 2 Total bait used (median, IQR, range and outliers) in
successful rodent eradications on tropical islands with
land crabs, by type of dominant land crab and type of
island. None = land crabs were present at low densities
therefore interference was minimum.
212
Table 2 Success rate of hand and aerial rodent eradication
projects, per target species, on tropical islands with land
crabs.
Target species
Mus musculus
Rattus exulans
Rattus norvegicus
Rattus rattus
Rattus tanezumi
TOTAL
Failed
%
n
28.6
2
17.4
12
0
0
18.5
5
0
0
17.6
19
Successful
%
n
71.4
5
82.6
57
100
3
81.5
22
100
2
82.4
89
extirpation. Pascal, et al. (2004) first suggested possible
rat predation on Gecarcinus ruricola on Hardy Islet, after
they found numerous crab carapaces in rat middens and the
index of crab abundance increased after the rat eradication.
Samaniego-Herrera & Bedolla-Guzmán (2012) and
Samaniego-Herrera (2014) confirmed invasive rats can
indeed cause ecological extirpations of land crabs; they
documented how G. quadratus, a small burrowing crab,
shifted from being extremely rare before the R. rattus
eradication to being the most abundant species four years
post-eradication. Evidence of rodent predation on land
crabs as well as the dramatic recovery responses following
rat eradication is growing (Harper & Bunbury, 2015;
Samaniego-Herrera, et al., 2017). For example, Nigro, et
al. (2017) showed that the recovery of two carnivorous
species (Geograpsus spp.), the smallest of the Palmyra
atoll crabs, led to a dramatic widening of the crab trophic
niche following the rat eradication on Palmyra atoll, which
is altering the ecology of the atoll presumably towards a
more natural state. Likewise, Russell, et al. (2015) showed
land crab trophic position differed depending on what
invasive rat species is present.
Case study: Mexican tropical islands
Bait uptake
Linear mixed models were constructed, including
period (trials or during eradication), rodent species (R.
rattus or M. musculus), rodent abundance (high or low) and
land crab abundance (high or low) per crab type (hermit
or burrowing) as covariates. The model with the greatest
support (49.8%) revealed a complex relationship between
the difference in bait density from target density (response
variable) and all variables tested and some interactions
(Table 5). Essentially, the rate of decline of bait density
depends on (a) days since broadcast, declining over time,
(b) the type of crab, declining faster with burrowing crabs,
(c) the density of burrowing crabs, declining with high
density, (d) the density of hermit crabs, declining with
high density, (e) study type, declining faster during trials,
(f) the broadcast, declining faster after first broadcast, (g)
the density of rats, declining with high density, and (h)
the density of mice, declining with high density. Fig. 3
illustrates the variability in bait density among plots and
islands after the rodent eradications, although in all cases
(trials and eradications) bait was still readily available after
the recommended four nights (Keitt, et al., 2015). The
faster decline in available bait when burrowing crabs are
abundant is a novel result.
Land crab recovery
For the pre-eradication period, the 2-way ANOVA
revealed significant differences in land crab density
depending on island type (highest on rainforest islands)
Samaniego-Herrera, et al.: Critical role of land crabs
Table 3 Current taxonomy of the 52 genera and 165 species of land crabs with island records.
Infraorder
Family
Genus1
Anomura
Coenobitidae
Birgus
Coenobita
Diogenidae
Brachyura
No.
species1
Documented bait
consumer? 2
1
Yes
17
Yes
Clibanarius
4
Calcinus
1
Porcellanidae
Petrolisthes
1
Eriphiidae
Eriphia
2
Gecarcinunidae
Barytelphusa
1
Parathelphusidae
Adeleana
1
Austrothelphusa
1
Geelvinkia
1
Holthuisana
2
Rouxana
3
Terrathelphusa
1
Thelphusula
2
Cardisoma
4
Yes
Discoplax
7
Yes
Epigrapsus
3
Gecarcinus
4
Yes
Gecarcoidea
3
Yes
Gecarcinidae
Documented rodent
vulnerability?3
Yes
Yes
Johngarthia
5
Yes
Yes
Grapsidae
Geograpsus
5
Yes
Yes
Sesarmidae
Aratus
1
Armases
5
Yes
Yes
Chiromantes
4
Episesarma
1
Geosesarma
24
Karstama
3
Labuanium
2
Metasesarma
1
Metopaulinas
1
Neosarmatium
1
Parasesarma
1
Sesarma
8
Sesarmoides
1
Sesarmops
1
Mictyridae
Mictyris
1
Ocypodidae
Afruca
1
Austruca
1
Cranuca
1
Gelasimus
2
Leptuca
12
Minuca
6
Ocypode
6
Tabuca
1
Uca
1
Ucides
2
Geryonidae
Carcinus
1
Potamidae
Cerberusa
1
Potamon
2
Madagapotamon
1
Malagasya
1
Guinotia
2
Potamonautidae
Pseudothelphusidae
Yes
According to Ng, et al. (2008), McLaughlin, et al. (2010) and Shih, et al. (2016).
2
According to this review.
3
See Table 4.
1
213
Island invasives: scaling up to meet the challenge. Ch 1C Rodents: Lessons
Table 4 Documented land crab species negatively impacted by invasive rodents on tropical islands.
Species
Birgus latro
Gecarcinus
lateralis
Mean size Invasive rodent
(mm)
640
37.7
Rodent
impact
Island/archipelago
Reference
Rattus rattus
Population
suppression
Tetiaroa Atoll,
Society Islands,
French Polynesia
Genet & Gaspar, pers.
comm. 2017
Rattus rattus
Ecological
extirpation
Pérez Island, Arrecife
Alacranes, Mexico
Samaniego-Herrera, et
al., 2017
Rattus rattus
Ecological
extirpation
Banco Chinchorro,
Mexico
This study
Rattus rattus
Ecological
extirpation
Half Moon Caye,
Belize
Samaniego-Herrera, et
al., 2015
Gecarcinus
quadratus
38.5
Rattus rattus
Ecological
extirpation
Isabel Island, Mexico
Samaniego-Herrera &
Bedolla-Guzmán, 2012
Gecarcinus
ruricola
69
Rattus rattus
Population
suppression
Hardy Island,
Martinique
Lorvelec & Pascal, 2005
Geograpsus
crinipes
46.4
Rattus rattus
Ecological
extirpation
Palmyra Atoll,
Tropical Pacific
Nigro, et al., 2017
Rattus sp.
Ecological
extirpation
Mañagaha Island,
Northern Mariana Is
Paulay & Starmer, 2011
Rattus norvegicus,
R. exulans
Ecological
extirpation
Raoul Island, New
Zealand
Bellingham, et al., 2010
Rattus rattus
Ecological
extirpation
Palmyra Atoll,
Tropical Pacific
Nigro, et al., 2017
Rattus sp.
Ecological
extirpation
Mañagaha Island,
Northern Mariana Is
Paulay & Starmer, 2011
Geograpsus
grayi
55.3
Johngarthia
planata1
92.5
Rattus rattus
Population
suppression
Clipperton Island,
eastern Pacific
Pitman, et al., 2005
Ocypode kuhlii
50
Rattus norvegicus,
R. exulans
Ecological
extirpation
Raoul Island, New
Zealand
Bellingham, et al., 2010
Ocypode
quadrata
45
Mus musculus,
Rattus rattus
Population
suppression
Alacranes Islands,
Mexico
This study
Rattus exulans, R.
tanezumi
Population
suppression
Wake Atoll,
Tropical Pacific
Carlton & Hodder, 2003
Uca spp.
25–40
Referred to as Gecarcinus planatus before 2008.
1
(F = 28.01, p <0.001) and season (higher in wet season) (F =
15.05, p <0.001). However, Tukey comparisons confirmed
that the increase during wet seasons was significant only
on two islands (Isabel and Muertos).
Given the differences between seasons on some islands,
trends in land crab abundance between pre and post
eradication periods was evaluated only for wet seasons. The
linear model confirmed that land crab densities (pooling
species) are significantly higher (R2 = 0.19, F = 30.27, p
< 0.001) post-eradication (1–5 years later) on all islands,
except on Cayo Norte and Pérez where the increase was
not significant. On Isabel Island, the smallest burrowing
crab (G. quadratus) showed a substantial trend of increase
over a period of five years after the rat eradication (Fig. 4).
DISCUSSION
Land crabs and rodent eradications
Interference by land crabs remains poorly understood,
documented and managed, despite its significant impact on
rodent eradication operations on tropical islands. It appears
that inconsistent application of recommended practices
214
(Wegmann, 2008; PII, 2011; Keitt, et al., 2015) continues
across projects (Broome, 2011), e.g. a lack of estimating
land crab densities and undertaking consumption trials to
inform baiting rates. Without determining how significant
the land crab problem is on an island and how it fluctuates
with seasons, managers tend to automatically apply a
mitigation strategy of high bait application rates.
However, the use of high bait rates increases the cost of
operations, adds logistical complexity and, if unnecessarily
high, potentially increases the risk to non-target vertebrates
(Pitt, et al., 2015) (as invertebrates are not susceptible to
anticoagulants (Broome, et al., 2012)). The highest total
bait rate used in a rodent eradication to date was 186 kg/ha
on Palmyra Atoll (Wegmann, et al., 2012). This high bait
rate (spread over two applications) was determined and
approved on the basis of sound field studies (Wegmann,
et al., 2008; Wegmann, et al., 2011) as it had been
demonstrated that the diverse and abundant land crab
community represented a significant bait ‘sink’, warranting
drastic mitigation measures. Importantly, Palmyra has an
exceptionally wet climate and this eradication case is an
outlier, as the second highest bait rate used to date is 94.2
kg/ha for the eradication of R. exulans on two Gambier
Samaniego-Herrera, et al.: Critical role of land crabs
higher when compared to the three or fewer species usually
recorded by eradication managers.
At present, bait uptake trials (e.g. Pott, et al., 2015)
are the best way to predict bait rates required for rodent
eradication. However, if climatic conditions are not very
similar during implementation, the ‘land crab scenario’
could be very different and significantly affect the bait
consumption rate. In order to better predict the potential
variability of bait uptake in the presence of certain land
crab communities, future monitoring of climatic conditions
and land crab communities on a suite of islands is required.
Note that carnivorous and intertidal crabs may also consume
bait (confirmed for ghost crabs, A. Barnaud pers. comm.),
but due to their generally low abundance (probably due to
vulnerability to rats) and/or limited distribution (coastal),
they tend to be neglected. The implications for failed
eradication attempts are important. Estimations of bait
uptake rates may no longer be true if a second eradication
attempt occurs within a few years of the first one, i.e. when
crab populations are more abundant because rats haven’t
had the time to fully recover and therefore haven’t again
supressed the land crabs.
Fig. 3 Bait availability days after hand or aerial broadcast
during six rodent eradications on Mexican islands.
islets (Kamaka and Makaroa)(Appendix 1). In addition,
the largest rat eradication on a rainforest island (539 ha
Cayo Centro, Banco Chinchorro), where large land crabs
are abundant, was successful using 60 kg/ha over two
applications (Samaniego-Herrera, et al., 2018). Moreover,
there have been several instances where managers reported
that the bait rates used were higher than required based on
their observations of bait availability 5–15 days post bait
application, which also coincided with ‘fewer land crabs
than expected’ (Steve Cranwell, Araceli Samaniego, Elenoa
Seniloli, pers. comm.). Studies (Samaniego-Herrera, et al.,
2014; Samaniego-Herrera, et al., 2018) have shown that
temporal land crab interference can vary substantially
even on rainforest islands, the lowest peaks of activity
being over the driest months. Thus, timing the eradication
operation to coincide with the driest conditions of the
year is recommended, particularly on islands with high
abundance of burrowing crabs. Unquestionably, land crab
activity is only one of the many factors that must be taken
into account while planning an eradication (PII, 2011), so
this has to be done in tandem with the other components,
for example, minimising operational risks and the potential
lethality to non-target species. Land crab interference with
bait stations was outside the scope of this paper, but it is
certainly a problem (Wegmann, 2008).
In addition to land crab abundance, land crab species
composition also affects bait uptake. Burrowing land
crabs, in particular large species (e.g. genera Cardisoma,
Johngarthia and Discoplax), are generally the most
problematic, although coconut crabs (B. latro) can be as
troublesome even though they are usually in low to medium
abundances. This has been reflected in the tendency to
use higher bait rates on islands with abundant burrowing
crabs, and confirmed by our case study where bait
availability was quantified. Hermit crabs appear to be more
widespread globally, so they may cause less interference
on individual islands but affect more islands overall. Note
that burrowing crabs, although larger, are more elusive
than hermits due to their propensity to burrow or seek
shelter under rocks or leaf litter during the day (Bliss,
et al., 1978). Species richness of land crabs per island is
likely to be generally underestimated. On the few islands
for which comprehensive inventories exist (e.g. Christmas
Island, Seychelles or French Polynesia), a list of 5–12
species of land crabs is common (Orchard, 2012), which is
Behaviours as well as consumption capabilities vary
widely among groups and species, hence the importance of
identifying species or at least type of crab (hermit, coconut,
burrowing; Fig. 5). Hermit crabs are small and slow eaters
and walkers compared to the average burrowing crab. They
can take only one piece of bait at a time and they do not
cache food. In contrast, most burrowing crabs are able to
take up to three pieces of bait at a time (depending on crab
size/species) and walk quickly to their burrow where they
cache the bait (G. Harper & A. Samaniego pers. obs.). How
much bait they can accumulate, how fast, and how long it
takes for it to be eaten has not been determined.
Land crab activity is regulated by a combination of
air and soil surface temperature, relative humidity, the
intensity of insolation (solar radiation) and the availability
of protective cover, be it leaf litter, suitable cavities, or
soil for burrowing, which is further influenced by the soil
compaction (Bliss, et al., 1978; Green, 1997; Brook, et al.,
2009). The optimum temperature for land crab activity
appears to be about 30oC, with virtually no activity below
18oC (Bliss, et al., 1978). Hence, in order to mitigate the
desiccating effects of the high temperatures that crabs
require to be active, their activity is largely restricted
to periods and locations with high humidity. To reduce
interference by land crabs, on ‘wet’ subtropical islands
cooler months should be targeted for eradications.
The thermoregulatory abilities of hermit crabs
(Coenobita spp.) are low, which is less of a problem where
there is little temperature variation, as often occurs on wet
tropical islands, but on arid islands they are essentially
nocturnal (Achituv & Ziskind, 1985). Wind strength will
also affect activity through increasing desiccation (Barnes,
1997). For burrowing crabs (e.g. Gecarcinus spp.) on
seasonally arid islands, activity is dictated by relative
humidity, with little or no activity below 77% RH, through
to high activity above 95% RH (Bliss, et al., 1978; Green,
1997; Capistrán-Barradas, et al., 2003). Burrowing crabs
probably occupy burrows to reduce their water loss. Often,
unseasonal rain or even showers will initiate short periods
of activity, but if conditions are very dry land crabs can
remain underground for several months (Bliss, et al., 1978).
The effects of humidity and insolation on land crab
activity strongly suggest nocturnal land crab monitoring
will detect the highest crab activity and species diversity,
particularly if all habitats are sampled and the season is
taken into account. Land crabs are long-lived species, so
monitoring tends to indicate how many land crabs are
active at that time, not how many are actually present.
215
Island invasives: scaling up to meet the challenge. Ch 1C Rodents: Lessons
Table 5 Significant parameters in relation to bait availability after rodent eradications, according to linear mixed
models for six rodent eradications on Mexican tropical islands.
Value
Std Error
DF
t-value
p-value
-1.456
0.275
2768
-5.287
> 0.001
4.204
1.014
2768
4.142
> 0.001
Bait application
-4.658
1.056
2768
-4.407
> 0.001
Day x Burrowing
-2.343
0.262
2768
-8.914
> 0.001
Day Hermit
-1.135
0.338
2768
-3.357
> 0.001
Day Period
-0.896
0.254
2768
-3.521
> 0.001
Day Bait application
1.115
0.275
2768
4.050
> 0.001
Hermit x Bait application
6.912
1.610
2768
4.291
> 0.001
-0.019
0.005
2768
-3.725
> 0.001
1.672
0.431
2768
3.875
> 0.001
Day
Period
x
x
x
Day Rodent
x
Day Burrowing Bait application
x
x
Rodent impacts on land crabs
Except for New Zealand, little has been documented
regarding impacts of invasive rodents on island
invertebrates in general (St Clair, 2011). Furthermore,
few land crab accounts exist, due to the limited research
conducted in the tropics (Brook, et al., 2009) and the low
proportion of rodent eradications carried out in this region
(Howald, et al., 2007; Holmes, et al., 2015). Our list of
land crab species impacted by rodents is most likely to
be severely under-reported as Rodentia comprise over
a quarter of terrestrial mammal species known to forage
on intertidal food sources. Burrowing crabs make up a
substantial proportion of this and the number of species is
highly likely to be an underestimation (Carlton & Hodder,
2003).
Adding to the impacts of invasive rodent predation on a
wide range of plants and animals (Carlton & Hodder, 2003;
St Clair, 2011; Sunde, 2012; Harper & Bunbury, 2015), the
suppression of native ecosystem engineers, such as land
crabs, by rats and other invasive species such as ants, cats
and dogs, could have significant and enduring consequences
on relatively simple island ecosystems (Carlton & Hodder,
2003; O’Dowd, et al., 2003). Land crabs are often the largest
invertebrates on tropical islands, and particularly atolls,
and will occupy the niches of vertebrates on small oceanic
islands (Burggren & McMahon, 1988). They are highly
integrated in the ecosystem energetics of tropical islands,
as they control recruitment and species composition of
Fig. 4 Population increase of Gecarcinus quadratus on
Isabel Island, Mexico after the ship rat eradication in
2009. The single record in 2008 was a new record for
the island.
216
seedlings on the forest floor (Green, et al., 1999; Lindquist
& Carroll, 2004). They also regulate nutrient dynamics
through substantial leaf litter consumption (Kellman &
Delfosse, 1993; Capistrán-Barradas, et al., 2003; Gutiérrez
& Jones, 2006; Gutiérrez, et al., 2006). Hence, they may
govern the growth and productivity of tropical forests
(Lindquist, et al., 2009) and sustain diversity at large
scales (Young, et al., 2013; Nigro, et al., 2017). Moreover,
plant composition will be influenced by soil structure,
which is affected by land crab activity through inland
transfer of marine debris and shells, removal of algae
from rock surfaces subsequently deposited as faeces, and
by increasing leaf litter breakdown through deposition
underground in burrows. Given the critical role land crabs
play in island ecosystems, it is recommended that these
are included as outcome indicators and monitored posteradication.
CONCLUSION
Land crabs are a diverse group. For management
purposes, it is useful to distinguish three groups: hermit
crabs, coconut crabs and burrowing crabs, noting the latter
vary widely in size. The ecology and climatic tolerances of
each group is different, as is their capacity as bait consumers.
Assessing species richness and abundance of land crabs
should be a priority in the planning phase of rodent
eradication projects in the tropics, so that their interference
can be efficiently managed during eradications. Similarly,
changes to the land crab community should be measured
Fig. 5 Types of land crabs in relation to potential interference
with rodent eradications on tropical islands. Sizes refer
to the range of mean size across species.
Samaniego-Herrera, et al.: Critical role of land crabs
post-eradication as indicators of ecosystem recovery. Once
enough data are gathered regarding bait consumption
in a standardised manner, we will improve our ability to
predict appropriate bait rates for the eradication of invasive
rodents on different types of tropical islands, as is currently
done for temperate islands. Where compatible with other
factors (e.g. non-target species and human activities),
rodent eradications on tropical islands should be timed for
the driest conditions or alternatively on ‘wet’ subtropical
islands, the coolest months.
ACKNOWLEDGEMENTS
We would like to thank the following colleagues
for providing unpublished data, project documents and
valuable comments: Ray Pierce, Derek Brown, Graham
Wragg, Thomas Ghestemme, Mike Bell, Elenoa Seniloli,
Richard Griffiths, Sia Rasalato and Gerard Rocamora.
Data from the Mexican islands were collated by A.
Samaniego thanks to the support of GECI. Thanks to Rigel
Sansores for kindly drawing the illustrations and to Anny
Peralta for helping with data collection. Thanks also to
Andy Cox, Dick Veitch and two anonymous reviewers for
the valuable feedback on the manuscript.
REFERENCES
Achituv, Y. and Ziskind, M. (1985). ‘Adaptation of Coenobita scaevola
(Forskål, 1775) (Crustacea, Anomura) to terrestrial life in desertbordered shore line’. Marine Ecology Progress Series 25: 189–197.
Athens, J.S., Tuggle, H.D., Welch, D.J. and Ward, J.V. (2002). ‘Avifaunal
extinctions, vegetation change, and Polynesian impacts in prehistoric
Hawai’i’. Archaeology in Oceania 37(2): 57.
Atkinson, I.A.E. (1985). ‘The spread of commensal species of Rattus
to oceanic islands and their effect on island avifaunas’. In: P.J. Moors
(ed.) Conservation of Island Birds. pp. 35–81. Cambridge: International
Council for Bird Preservation.
Barnes, D.K. (1997). ‘Ecology of tropical hermit crabs at Quirimba
Island, Mozambique: Distribution, abundance and activity’. Marine
Ecology Progress Series 154: 133–142.
Bellard, C., Cassey, P. and Blackburn, T.M. (2016). ‘Alien species as a
driver of recent extinctions’. Biology Letters 12(2): 20150623.
Bliss, D.E., Van Montfrans, J., Van Montfrans, M. and Boyer, J.R.
(1978). ‘Behavior and growth of the land crab Gecarcinus lateralis
(Freminville) in southern Florida’. Bulletin of the American Museum of
Natural History 160(11): 111–152.
Brook, S., Grant, A. and Bell, D. (2009). ‘Can land crabs be used as a rapid
ecosystem evaluation tool? A test using distribution and abundance of
several genera from the Seychelles’. Acta Oecologica 35(5): 711–719.
Broome, K. (2011). ‘Transferring rodent eradication success from
temperate regions into the tropical Pacific – What do we need to learn?’.
PII News. Auckland: Pacific Invasives Initiative.
Broome, K.G., Fairweather, A.A.C. and Fisher, P. (2012). Brodifacoum
Pesticide Information Review. Version 2012/1. Hamilton, New Zealand:
New Zealand Department of Conservation.
Broome, K.G., Cox, A., Golding, C., Cromarty, P., Bell, P. and McClelland,
P. (2014). Rat Eradication Using Aerial Baiting: Current Agreed Best
Practice Used in New Zealand (Version 3.0). Wellington, New Zealand:
New Zealand Department of Conservation.
Burggren, W.W. and McMahon, B.R. (1988). Biology of the Land Crabs,
Cambridge, Cambridge University Press.
Capistrán-Barradas, A., Defeo, O. and Moreno-Casasola, P. (2003).
‘Density and population structure of the red land crab Gecarcinus
lateralis in a tropical semi-deciduous forest in Veracruz, Mexico’.
Interciencia 28(6): 323–327.
Carlton, J.T. and Hodder, J. (2003). ‘Maritime mammals: Terrestrial
mammals as consumers in marine intertidal communities’. Marine
Ecology Progress Series 256: 271–286.
Croll, D.A., Newton, K.M., McKown, M., Holmes, N., Williams, J.C.,
Young, H.S., Buckelew, S., Wolf, C.A., Howald, G., Bock, M.F.,
Curl, J.A. and Tershy, B.R. (2016). ‘Passive recovery of an island bird
community after rodent eradication’. Biological Invasions 18(3): 703–
715.
DIISE (2016). The Database of Island Invasive Species Eradications,
Developed by Island Conservation, Coastal Conservation Action
Laboratory UCSC, IUCN SSC Invasive Species Specialist Group,
University of Auckland and Landcare Research New Zealand. <http://
diise.islandconservation.org>.
Green, P.T. (1997). ‘Red crabs in rain forest on Christmas Island, Indian
Ocean: Activity patterns, density and biomass’. Journal of Tropical
Ecology 13(1): 17–38.
Green, P.T., Lake, P.S. and O’Dowd, D.J. (1999). ‘Monopolization of
litter processing by a dominant land crab on a tropical oceanic island’.
Oecologia 119(3): 435–444.
Green, P.T., O’Dowd, D.J. and Lake, P.S. (2008). ‘Recruitment dynamics
in a rainforest seedling community: Context-independent impact of a
keystone consumer’. Oecologia 156(2): 373–385.
Griffiths, R., Miller, A. and Climo, G. (2011). ‘Addressing the impact
of land crabs on rodent eradications on islands’. Pacific Conservation
Biology 17(4): 347–353.
Gutiérrez, J.L. and Jones, C.G. (2006). ‘Physical ecosystem engineers as
agents of biogeochemical heterogeneity’. Bioscience 56(3): 227–236.
Gutiérrez, J.L., Jones, C.G., Groffman, P.M., Findlay, S.E.G., Iribarne,
O.O., Ribeiro, P.D. and Bruschetti, C.M. (2006). ‘The contribution of
crab burrow excavation to carbon availability in surficial salt-marsh
sediments’. Ecosystems 9(4): 647–658.
Harper, G.A. and Bunbury, N. (2015). ‘Invasive rats on tropical islands:
Their population biology and impacts on native species’. Global
Ecology and Conservation 3: 607–627.
Hilton, G.M. and Cuthbert, R.J. (2010). ‘The catastrophic impact of
invasive mammalian predators on birds of the UK Overseas Territories:
A review and synthesis’. Ibis 152(3): 443–458.
Holmes, N.D., Griffiths, R., Pott, M., Alifano, A., Will, D., Wegmann, A.S.
and Russell, J.C. (2015). ‘Factors associated with rodent eradication
failure’. Biological Conservation 185: 8–16.
Howald, G., Donlan, C.J., Galván, J.P., Russell, J.C., Parkes, J.,
Samaniego, A., Wang, Y., Veitch, D., Genovesi, P., Pascal, M., Saunders,
A. and Tershy, B. (2007). ‘Invasive rodent eradication on islands’.
Conservation Biology 21(5): 1258–1268.
Jones, H.P., Holmes, N.D., Butchart, S.H.M., Tershy, B.R., Kappes,
P.J., Corkery, I., Aguirre-Muñoz, A., Armstrong, D.P., Bonnaud, E.,
Burbidge, A.A., Campbell, K., Courchamp, F., Cowan, P.E., Cuthbert,
R.J., Ebbert, S., Genovesi, P., Howald, G.R., Keitt, B.S., Kress, S.W.,
Miskelly, C.M., Oppel, S., Poncet, S., Rauzon, M.J., Rocamora, G.,
Russell, J.C., Samaniego-Herrera, A., Seddon, P.J., Spatz, D.R., Towns,
D.R. and Croll, D.A. (2016). ‘Invasive mammal eradication on islands
results in substantial conservation gains’. Proceedings of the National
Academy of Sciences 113(15): 4033–4038.
Keitt, B., Griffiths, R., Boudjelas, S., Broome, K., Cranwell, S., Millett,
J., Pitt, W. and Samaniego-Herrera, A. (2015). ‘Best practice guidelines
for rat eradication on tropical islands’. Biological Conservation 185:
17–26.
Kellman, M. and Delfosse, B. (1993). ‘Effect of the red land crab
(Gecarcinus lateralis) on leaf litter in a tropical dry forest in Veracruz,
Mexico’. Journal of Tropical Ecology 9(1): 55–65.
Le Corre, M., Danckwerts, D.K., Ringler, D., Bastien, M., Orlowski, S.,
Morey Rubio, C., Pinaud, D. and Micol, T. (2015). ‘Seabird recovery
and vegetation dynamics after Norway rat eradication at Tromelin
Island, western Indian Ocean’. Biological Conservation 185: 85–94.
Lindquist, E.S. and Carroll, C.R. (2004). ‘Differential seed and seedling
predation by crabs: Impacts on tropical coastal forest composition’.
Oecologia 141(4): 661–671.
Lindquist, E.S., Krauss, K.W., Green, P.T., O’Dowd, D.J., Sherman, P.M.
and Smith, T.J. (2009). ‘Land crabs as key drivers in tropical coastal
forest recruitment’. Biological Reviews 84(2): 203–223.
McLaughlin, P.A., Komai, T., Lemaitre, R. and Rahayu, D.L. (2010).
‘Annotated checklist of anomuran decapod crustaceans of the world
(exclusive of the Kiwaoidea and families Chirostylidae and Galatheidae
of the Galatheoidea) Part I – Lithodoidea, Lomisoidea and Paguroidea’.
The Raffles Bulletin of Zoology 23(1): 131–137.
Myers, N., Mittermeier, R.A., Mittermeier, C.G., Da Fonseca, G.A.B.
and Kent, J. (2000). ‘Biodiversity hotspots for conservation priorities’.
Nature 403(6772): 853–858.
Ng, P.K.L., Guinot, D. and Davie, P.J.F. (2008). ‘Systema Brachyurorum:
Part 1. An annotated checklist of extant brachyuran crabs of the world’.
The Raffles Bulletin of Zoology 17: 1–286.
Nigro, K.M., Hathaway, S.A., Wegmann, A.S., Miller-ter Kuile, A.,
Fisher, R.N. and Young, H.S. (2017). ‘Stable isotope analysis as an
early monitoring tool for community-scale effects of rat eradication’.
Restoration Ecology 10.1111/rec.12511.
217
Island invasives: scaling up to meet the challenge. Ch 1C Rodents: Lessons
O’Dowd, D.J., Green, P.T. and Lake, P.S. (2003). ‘Invasional ‘meltdown’
on an oceanic island’. Ecology Letters 6(9): 812–817.
Orchard, M. (2012). Crabs of Christmas Island, Christmas Island Natural
History Association.
Paine, R.T. (1966). ‘Food web complexity and species diversity’.
American Naturalist 100(910): 65–75.
Pascal, M., Brithmer, R., Lorvelec, O. and Venumiere, N. (2004).
‘Conséquences sur l’avifaune nicheuse de la réserve naturelle des
Îlets De Sainte-Anne (Martinique) de la récente invasion du rat noir
(Rattus rattus), etablies à l’issue d’une tentative d’éradication’. Revue
d’écologie 59(1–2): 309–318.
Paulay, G. and Starmer, J. (2011). ‘Evolution, insular restriction, and
extinction of oceanic land crabs, exemplified by the loss of an endemic
Geograpsus in the Hawaiian Islands’. PLOS ONE 6(5): e19916.
PII. (2011). Resource Kit for Rodent and Cat Eradication. Pacific
Invasives Initiative. <http://pacificinvasivesinitiative.org/rce/index.
html>. [Accessed January 2017].
Shih, H.-T., Ng, P.K., Davie, P.J., Schubart, C.D., Türkay, M., Naderloo,
R., Jones, D. and Liu, M.-Y. (2016). ‘Systematics of the family
Ocypodidae Rafinesque, 1815 (Crustacea: Brachyura), based on
phylogenetic relationships, with a reorganization of subfamily rankings
and a review of the taxonomic status of Uca Leach, 1814, sensu lato and
its subgenera’. The Raffles Bulletin of Zoology 64: 139–175.
St Clair, J.J.H. (2011). ‘The impacts of invasive rodents on island
invertebrates’. Biological Conservation 144(1): 68–81.
Sunde, F.G. (2012). ‘The Effects of Rats on Rocky Intertidal Ecosystems
in Northeastern New Zealand’. MSc thesis. Auckland: University of
Auckland.
Towns, D.R., Atkinson, I.A.E. and Daugherty, C.H. (2006). ‘Have
the harmful effects of introduced rats on islands been exaggerated?’
Biological Invasions 8(4): 863–891.
Towns, D.R., West, C.J. and Broome, K.G. (2013). ‘Purposes, outcomes
and challenges of eradicating invasive mammals from New Zealand
islands: An historical perspective’. Wildlife Research 40(2): 94–107.
Pitman, R.L., Ballance, L.T. and Bost, C. (2005). ‘Clipperton Island: Pig
sty, rat hole and booby prize’. Marine Ornithology 33: 193–194.
Wegmann, A.S. (2008). Land Crab Interference with Eradication
Projects: Phase I – Compendium of Available Information. Auckland:
Pacific Invasives Initiative.
Pitt, W.C., Berentsen, A.R., Shiels, A.B., Volker, S.F., Eisemann,
J.D., Wegmann, A.S. and Howald, G.R. (2015). ‘Non-target species
mortality and the measurement of brodifacoum rodenticide residues
after a rat (Rattus rattus) eradication on Palmyra Atoll, tropical Pacific’.
Biological Conservation 185: 36–46.
Wegmann, A.S., Helm, J., Jacobs, B., Samaniego, A., Smith, W., Drake,
D., Fisher, R., Hathaway, S., Henry, A., Smith, J. and McKown, M.
(2008). Palmyra Atoll Rat Eradication: Biomarker Validation of an
Effective Bait Application Rate. Technical Field Report. Santa Cruz,
California: Island Conservation.
Pott, M., Wegmann, A.S., Griffiths, R., Samaniego-Herrera, A., Cuthbert,
R.J., Brooke, M.d.L., Pitt, W.C., Berentsen, A.R., Holmes, N.D.,
Howald, G.R., Ramos-Rendón, K. and Russell, J.C. (2015). ‘Improving
the odds: Assessing bait availability before rodent eradications to aid in
selecting bait application rates’. Biological Conservation 185: 27–35.
Wegmann, A.S., Buckelew, S., Howald, G., Helm, J. and Swinnerton,
K. (2011). ‘Rat eradication campaigns on tropical islands: Novel
challenges and possible solutions’. In: C.R. Veitch, M.N. Clout and
D.R. Towns (eds.) Island invasives: eradication and management, pp.
239-243. Occassional Paper SSC no. 42. Gland, Switzerland: IUCN and
Auckland, New Zealand: CBB.
Russell, J.C., Caut, S., Anderson, S.H. and Lee, M. (2015). ‘Invasive rat
interactions and over-invasion on a coral atoll’. Biological Conservation
185: 59–65.
Russell, J.C. and Holmes, N.D. (2015). ‘Tropical island conservation: Rat
eradication for species recovery’. Biological Conservation 185: 1–7.
Wegmann, A.S., Flint, E., White, S., Fox, M., Howald, G., McClelland, P.,
Alifano, A. and Griffiths, R. (2012). ‘Pushing the envelope in paradise:
A novel approach to rat eradication at Palmyra Atoll’. In: R.M. Timm
(ed.) 25th Vertebrate Pest Conference, pp. 48–53. Davis: University of
California.
Russell, J.C., Meyer, J.-Y., Holmes, N.D. and Pagad, S. (2017).
‘Invasive alien species on islands: Impacts, distribution, interactions
and
management’.
Environmental
Conservation
10.1017/
S0376892917000297.
Weigelt, P., Jetz, W. and Kreft, H. (2013). ‘Bioclimatic and physical
characterization of the world’s islands’. Proceedings of the National
Academy of Sciences 110(38): 15307–15312.
Samaniego-Herrera, A. (2014). ‘Ecology and Impacts of Invasive Rodents
on Tropical Islands, in Relation to Eradication Operations’. PhD thesis.
Auckland: The University of Auckland.
Young, H.S., McCauley, D.J., Guevara, R. and Dirzo, R. (2013).
‘Consumer preference for seeds and seedlings of rare species impacts
tree diversity at multiple scales’. Oecologia 172(3): 857–867.
Samaniego-Herrera, A., Aguirre-Muñoz, A., Rodríguez-Malagón, M.,
González-Gómez, R., Torres-García, F., Méndez-Sánchez, F., FélixLizárraga, M. and Latofski-Robles, M. (2011). ‘Rodent eradications on
Mexican islands: Advances and challenges’. In: C.R. Veitch, M.N. Clout
and D.R. Towns (eds.) Island invasives: eradication and management,
pp. 350–355. Occassional Paper SSC no. 42. Gland, Switzerland: IUCN
and Auckland, New Zealand: CBB.
Samaniego-Herrera, A. and Bedolla-Guzmán, Y. (2012). ‘Land crabs
(Decapoda, Brachyura, Gecarcinidae) on Isabel Island, Mexico,
including a new record, and its relation to the removal of invasive rats’.
Crustaceana 85(8): 1007–1011.
Samaniego-Herrera, A., Anderson, D.P., Parkes, J.P. and Aguirre-Muñoz,
A. (2013). ‘Rapid assessment of rat eradication after aerial baiting’.
Journal of Applied Ecology 50(6): 1415–1421.
Samaniego-Herrera, A., Russell, J.C., Choquenot, D., Aguirre-Muñoz, A.
and Clout, M. (2014). ‘Invasive rodents on tropical islands: eradication
recommendations from Mexico’. In: R.M. Timm and J.M. O’Brien
(eds.) Proceedings of the 26th Vertebrate Pest Conference, pp. 43–50.
Davis: University of California.
Samaniego-Herrera, A., Clout, M.N., Aguirre-Muñoz, A. and Russell, J.C.
(2017). ‘Rodent eradications as ecosystem experiments: A case study
from the Mexican tropics’. Biological Invasions 19(6): 1761–1779.
Samaniego-Herrera, A., Aguirre-Muñoz, A., Bedolla-Guzmán, Y.,
Cárdenas-Tapia, A., Félix-Lizárraga, M., Méndez-Sánchez, F., ReinaPonce, O., Rojas-Mayoral, E. and Torres-García, F. (2018). ‘Eradicating
invasive rodents from wet and dry tropical islands in Mexico’. Oryx 52:
559–570.
Sherman, P.M. (2002). ‘Effects of land crabs on seedling densities and
distributions in a mainland neotropical rain forest’. Journal of Tropical
Ecology 18(1): 67–89.
Sherman, P.M. (2006). ‘Influence of land crabs Gecarcinus quadratus
(Gecarcinidae) on distributions of organic carbon and roots in a Costa
Rican rain orest’. Revista de Biologia Tropical 54(1): 149–161.
218
Island
Allen Cay
Anchorage
Motu Kena
Motu Kena-iti
Motu Tou
Mabualau
Nukubasaga
Nukupureti
Nukusemanu
Qelelevu
Tauraria
Tuinibeka
Vatu-i-Ra
Vetauua
Hiuveru
Hiveu
Kamaka
Makapu
Makaroa
Mekiro
Motu-o-ari
Teauaone
Temoe
Tenarunga
Tenarunga
Tepapuri
Tiarao
Country/
territory
Bahamas
Cook Islands
Cook Islands
Cook Islands
Cook Islands
Fiji
Fiji
Fiji
Fiji
Fiji
Fiji
Fiji
Fiji
Fiji
French Polynesia
French Polynesia
French Polynesia
French Polynesia
French Polynesia
French Polynesia
French Polynesia
French Polynesia
French Polynesia
French Polynesia
French Polynesia
French Polynesia
French Polynesia
6.9
12.8
1.2
0.7
14.7
3.2
18.0
3.0
1.6
147.0
49.3
2.9
2.0
35.0
3.2
4.7
47.6
11.2
16.4
11.5
4.5
8.8
430.8
424.0
424.0
26.0
4.2
Area
(ha)
Mm
Re
Re
Re
Re
Re
Re
Re
Re
Re
Re
Re
Re
Re
Rr
Rr
Re
Re
Re
Re
Re
Re
Re
Re
Rr
Re
Rr
Target
species
2012
2012
2012
2012
2012
2008
2008
2008
2008
2008
2008
2008
2006
2008
2008
2008
2015
2003
2015
2003
2003
2003
2015
2015
2015
2003
2008
Erad.
year
Dry
Wet
Wet
Wet
Wet
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Wet
Wet
Dry
Dry
Dry
Dry
Dry
Dry
Dry
NA
NA
Dry
Wet
Erad.
season
S
S
S
S
F
S
S
S
S
S
S
S
S
S
F
F
F
S
S
F
F
F
S
S
S
F
F
Erad.
result
Hand
Hand
Hand
Hand
Hand
Hand
Aerial
Aerial
Aerial
Aerial
Aerial
Aerial
Hand
Aerial
Hand
Hand
Aerial
Hand
Aerial
Hand
Hand
Hand
Aerial
Aerial
Aerial
Hand
Hand
Erad.
method
40
20
20
20
20
26
25
25
25
25
25
25
26
25
15
15
94.2
4
94.2
4
3
3
87.6
72.6
72.6
3
15
25D
20R
20R
20R
20R
20R
20R
20R
20R
20R
20R
20R
20R
20R
20R
20R
25W
20R
25W
20R
20R
20R
25W
25W
25W
20R
20R
Total bait Bait
rate (kg/ type
ha)
Savanna
Seasonal
Seasonal
Seasonal
Seasonal
Seasonal
Seasonal
Seasonal
Seasonal
Seasonal
Seasonal
Seasonal
Seasonal
Seasonal
Seasonal
Seasonal
Rainforest
Rainforest
Rainforest
Rainforest
Rainforest
Rainforest
Seasonal
Seasonal
Seasonal
Rainforest
Seasonal
Island type
Hermit
Hermit
Hermit
Hermit
Hermit
Hermit
Hermit
Hermit
Hermit
Hermit
Hermit
Hermit
Hermit
Hermit
Hermit
Hermit
Hermit
Hermit
Hermit
Hermit
Hermit
Hermit
Hermit
Hermit
Hermit
Hermit
Hermit
Main
interference
Low
Medium
Medium
Low
Medium
High
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
None
None
None
Low
Low
Low
Medium
Medium
Medium
Medium
High
Medium
Medium
Medium
Medium
Medium
Medium
Medium
Medium
Low
Low
Medium
Medium
Medium
Medium
High
High
High
Medium
Medium
High
Low
None
Medium
Low
Medium
High
None
Low
None
None
High
Low
Low
None
High
None
None
None
None
None
None
None
None
None
None
None
None
None
Burrowing Hermit crab Coconut
crab density1 density1
crab
density1
Ca, Co, Ge
Bi, Ca, Co, Oc
Bi, Ca, Co, Oc
Bi, Ca, Co, Oc
Bi, Ca, Co, Oc
Ca, Co
Bi, Co
Co
Co
Bi, Co
Bi, Co
Bi, Co
Co
Bi, Co
Co
Co
Co, Gg
Co
Co, Gg
Co
Co
Co
Co
Co
Co
Co
Co, Ge
Land crab
genera2
Appendix 1. Details of rodent eradications on tropical islands with land crabs, up to 2015, conducted either by aerial broadcast (Aerial) or hand broadcast (Hand; Hand*= bait piles on
the ground on a grid) of bait. Target species: Mm= Mus musculus, Re= Rattus exulans, Rn= Rattus norvegicus, Rr= Rattus rattus, Rt= Rattus tanezumi; Eradication result: F= failure,
S= success; Bait type: 20R= Pestoff 20R, 25D= Bell-25D, 25W= Bell-25W; Crab genera: Bi= Birgus, Ca= Cardisoma, Co= Coenobita, Ge= Gecarcinus, Gd= Gecarcoidea, Gg=
Geograpsus, Jo= Johngartia, Oc= Ocypode. 1During the eradication; 2Observed by project managers throughout the project.
Samaniego-Herrera, et al.: Critical role of land crabs
219
220
Island
Toreauta
Vahanga
Big Ambo
Big Fred/Tonga
Big Nimroona
Big Tibo
Birnie
Drum
E isle
East Drum
Enderbury
Isles Lagoon 13
Isles Lagoon 16
Isles Lagoon 2
Isles Lagoon 21
Isles Lagoon 22
Isles Lagoon 23
Isles Lagoon 3
Isles Lagoon 4
Isles Lagoon 5
McKean
North Drum
NW Fred/Tonga
NW Nimroona
NW Tibo
SE Fred/Tonga
SW Islet Koil
SW motu Koil
Country/
territory
French Polynesia
French Polynesia
Kiribati
Kiribati
Kiribati
Kiribati
Kiribati
Kiribati
Kiribati
Kiribati
Kiribati
Kiribati
Kiribati
Kiribati
Kiribati
Kiribati
Kiribati
Kiribati
Kiribati
Kiribati
Kiribati
Kiribati
Kiribati
Kiribati
Kiribati
Kiribati
Kiribati
Kiribati
5.3
380.0
1.4
3.5
6.5
3.8
49.4
6.1
0.8
1.0
608.0
1.2
4.1
1.4
1.2
1.5
0.1
0.5
1.4
0.3
27.0
2.5
1.3
0.6
0.8
0.8
0.1
3.0
Area
(ha)
Rr
Re
Re
Re
Re
Re
Re
Re
Re
Re
Re
Re
Re
Re
Re
Re
Re
Re
Re
Re
Rt
Re
Re
Re
Re
Re
Re
Re
Target
species
2011
2015
2009
2009
2009
2012
2011
2009
2009
2009
2011
2009
2009
2009
2009
2009
2009
2009
2009
2009
2008
2009
2009
2009
2009
2009
2009
2009
Erad.
year
Dry
NA
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Erad.
season
S
S
S
S
S
S
S
S
S
S
F
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
Erad.
result
Hand
Aerial
Hand
Hand
Hand
Hand
Aerial
Hand
Hand
Hand
Aerial
Hand
Hand
Hand
Hand
Hand
Hand
Hand
Hand
Hand
Hand
Hand
Hand
Hand
Hand
Hand
Hand
Hand
Erad.
method
Total bait
rate (kg/
ha)
27.3
72.4
14.3
22.8
19.2
12.5
51
22.15
13.3
15
38.4
16.6
13.9
17.1
16.6
14.6
10
20
10
20
62.8
24
26.9
33.2
12.5
22.5
8
10
25W
25W
20R
20R
20R
20R
20R
20R
20R
20R
20R
20R
20R
20R
20R
20R
20R
20R
20R
20R
20R
20R
20R
20R
20R
20R
20R
20R
Bait
type
Seasonal
Seasonal
Savanna
Savanna
Savanna
Savanna
Savanna
Savanna
Savanna
Savanna
Savanna
Savanna
Savanna
Savanna
Savanna
Savanna
Savanna
Savanna
Savanna
Savanna
Savanna
Savanna
Savanna
Savanna
Savanna
Savanna
Savanna
Savanna
Island type
Hermit
Hermit
None
None
None
None
Hermit
None
None
None
Hermit
None
None
None
None
None
None
None
None
None
Hermit
None
None
None
None
None
None
None
Main
interference
Burrowing Hermit crab Coconut
crab density1 density1
crab
density1
Low
Low
Low
Low
Medium Low
Low
None
None
Low
None
None
Low
None
None
Low
None
None
Low
High
None
Low
None
None
Low
None
None
Low
None
None
Low
High
None
Low
None
None
Low
None
None
Low
None
None
Low
None
None
Low
None
None
Low
None
None
Low
None
None
Low
None
None
Low
None
None
Low
High
None
Low
None
None
Low
None
None
Low
None
None
Low
None
None
Low
None
None
Low
None
None
Low
None
None
Bi, Co, Gg
Co, Ca?
Ca?, Oc?
Ca?, Oc?
Ca?, Oc?
Ca?, Oc?
Ca, Co
Ca?, Oc?
Ca?, Oc?
Ca?, Oc?
Bi, Ca, Co, Gg
Ca?, Oc?
Ca?, Oc?
Ca?, Oc?
Ca?, Oc?
Ca?, Oc?
Ca?, Oc?
Ca?, Oc?
Ca?, Oc?
Ca?, Oc?
Ca, Co
Ca?, Oc?
Ca?, Oc?
Ca?, Oc?
Ca?, Oc?
Ca?, Oc?
Ca?, Oc?
Ca?, Oc?
Land crab
genera2
Appendix 1 (continued) Details of rodent eradications on tropical islands with land crabs, up to 2015, conducted either by aerial broadcast (Aerial) or hand broadcast (Hand; Hand*=
bait piles on the ground on a grid) of bait. Target species: Mm= Mus musculus, Re= Rattus exulans, Rn= Rattus norvegicus, Rr= Rattus rattus, Rt= Rattus tanezumi; Eradication
result: F= failure, S= success; Bait type: 20R= Pestoff 20R, 25D= Bell-25D, 25W= Bell-25W; Crab genera: Bi= Birgus, Ca= Cardisoma, Co= Coenobita, Ge= Gecarcinus, Gd=
Gecarcoidea, Gg= Geograpsus, Jo= Johngartia, Oc= Ocypode. 1During the eradication; 2Observed by project managers throughout the project.
Island invasives: scaling up to meet the challenge. Ch 1C Rodents: Lessons
Island
SW Nimroona
Flat
Flat
Gabriel
Gunner's Quoin
Cayo Centro
Cayo Norte Mayor
Cayo Norte Menor
Isabel
Muertos
Pájaros
Pérez
Dekehtik
Pein Mal
Double
G'I
Laregnere
Mato
Ndo
Nge
Redika
Signal
Table
Tiam'bouene
Uatermbi
Uatio
Uie
Uo
Country/
territory
Kiribati
Mauritius
Mauritius
Mauritius
Mauritius
Mexico
Mexico
Mexico
Mexico
Mexico
Mexico
Mexico
Micrionesia
Micrionesia
New Caledonia
New Caledonia
New Caledonia
New Caledonia
New Caledonia
New Caledonia
New Caledonia
New Caledonia
New Caledonia
New Caledonia
New Caledonia
New Caledonia
New Caledonia
New Caledonia
3.9
249.6
249.6
40.5
65.0
539.0
29.0
15.0
82.0
15.0
3.0
14.0
2.6
2.2
6.0
5.0
0.5
5.0
17.2
7.0
7.0
6.0
11.5
17.0
1.0
5.0
2.0
3.0
Area
(ha)
Re
Rr
Mm
Rr
Rn
Rr
Rr
Rr
Rr
Mm
Mm
Rr
Re
Rr
Re
Re
Re
Rr
Re
Re
Re
Re
Rr
Re
Re
Re
Re
Re
Target
species
2009
1998
1998
1995
1995
2015
2012
2012
2009
2011
2011
2011
2007
2007
2008
1998
1998
1998
1998
1998
1998
1998
2008
2008
1998
1998
1998
1998
Erad.
year
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Erad.
season
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
Erad.
result
Hand
Hand
Hand
Hand
Hand
Aerial
Aerial
Aerial
Aerial
Hand
Hand
Hand
Hand
Hand
Hand
Hand*
Hand*
Hand*
Hand*
Hand*
Hand*
Hand*
Hand
Hand
Hand*
Hand*
Hand*
Hand*
Erad.
method
Total bait
rate (kg/
ha)
22.6
15
15
20
15
60
42
42
20
17
17
17
50
50
20
14
14
14
14
14
14
14
20
20
14
14
14
14
20R
20R
20R
20R
20R
25W
25W
25W
25D
25D
25D
25D
25W
25W
20R
20R
20R
20R
20R
20R
20R
20R
20R
20R
20R
20R
20R
20R
Bait
type
Savanna
Seasonal
Seasonal
Seasonal
Seasonal
Rainforest
Rainforest
Rainforest
Seasonal
Savanna
Savanna
Savanna
Rainforest
Rainforest
Seasonal
Seasonal
Seasonal
Seasonal
Seasonal
Seasonal
Seasonal
Seasonal
Seasonal
Seasonal
Seasonal
Seasonal
Seasonal
Seasonal
Island type
None
Hermit
Hermit
Hermit
Hermit
Burrowing
Burrowing
Burrowing
Hermit
Burrowing
Hermit
Hermit
Burrowing
Burrowing
None
None
None
None
None
None
None
None
None
None
None
None
None
None
Main
interference
Burrowing Hermit crab Coconut
crab density1 density1
crab
density1
Low
None
None
Low
High
None
Low
High
None
High
High
None
Low
Low
None
Medium
Low
None
Medium
Low
None
Medium
Low
None
Low
Low
None
Medium
Low
None
Low
High
None
Low
High
None
High
High
Low
High
High
Low
None
Low
None
Low
Low
None
Low
Low
None
Low
Low
None
Low
Low
None
Low
Low
None
Low
Low
None
Low
Low
None
None
Low
None
None
Low
None
Low
Low
None
Low
Low
None
Low
Low
None
Low
Low
None
Ca?, Oc?
Ca, Co
Ca, Co
Ca, Co
Ca, Co
Ca, Co, Ge
Ca, Co, Ge
Ca, Co, Ge
Co, Ge, Jo
Co, Ge
Co, Ge
Co, Ge
Bi, Ca, Co
Bi, Ca, Co
Co
Co, Ge?
Co, Ge?
Co, Ge?
Co, Ge?
Co, Ge?
Co, Ge?
Co, Ge?
Co
Co
Co, Ge?
Co, Ge?
Co, Ge?
Co, Ge?
Land crab
genera2
Appendix 1 (continued) Details of rodent eradications on tropical islands with land crabs, up to 2015, conducted either by aerial broadcast (Aerial) or hand broadcast (Hand; Hand*=
bait piles on the ground on a grid) of bait. Target species: Mm= Mus musculus, Re= Rattus exulans, Rn= Rattus norvegicus, Rr= Rattus rattus, Rt= Rattus tanezumi; Eradication
result: F= failure, S= success; Bait type: 20R= Pestoff 20R, 25D= Bell-25D, 25W= Bell-25W; Crab genera: Bi= Birgus, Ca= Cardisoma, Co= Coenobita, Ge= Gecarcinus, Gd=
Gecarcoidea, Gg= Geograpsus, Jo= Johngartia, Oc= Ocypode. 1During the eradication; 2Observed by project managers throughout the project.
Samaniego-Herrera, et al.: Critical role of land crabs
221
222
Island
Vua
Fanna
Kayangel
Ducie
Henderson
Oeno
Pitcairn
Desecheo
Conception
Curieuse
Curieuse
Denis
Denis
Fregate
Fregate
Grande Ile
Grande Polyte
Grande Soeur
Ile aux Rats
North
North
Petit Polyte
Petite Soeur
Palmyra
Wake
Wake
Country/
territory
New Caledonia
Palau
Palau
Pitcairn
Pitcairn
Pitcairn
Pitcairn
Puerto Rico
Seychelles
Seychelles
Seychelles
Seychelles
Seychelles
Seychelles
Seychelles
Seychelles
Seychelles
Seychelles
Seychelles
Seychelles
Seychelles
Seychelles
Seychelles
US Islands
US Islands
US Islands
5.0
35.0
112.0
75.0
4,310.0
66.0
476.1
116.0
69.0
289.0
289.0
133.0
133.0
219.0
219.0
143.0
21.0
105.2
1.0
201.0
201.0
1.0
48.0
234.9
696.0
696.0
Area
(ha)
Re
Re
Re
Re
Re
Re
Re
Rr
Rn
Rr
Mm
Rr
Mm
Mm
Rn
Rr
Rr
Rr
Rr
Rr
Rr
Rr
Rr
Rr
Re
Rt
Target
species
1998
2009
2011
1997
2011
1997
1998
2012
2007
2000
2000
2000
2000
2000
2000
2007
2007
2010
2005
2003
2005
2007
2010
2011
2012
2012
Erad.
year
Dry
Dry
Wet
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Wet
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Erad.
season
S
F
F
S
F
S
F
F
S
S
F
F
F
S
S
S
S
S
S
S
S
S
S
S
F
S
Erad.
result
Hand*
Hand
Hand
Hand
Aerial
Hand
Hand
Aerial
Aerial
Aerial
Aerial
Aerial
Aerial
Aerial
Aerial
Aerial
Aerial
Aerial
Hand
Aerial
Aerial
Aerial
Aerial
Aerial
Aerial
Aerial
Erad.
method
Total bait
rate (kg/
ha)
14
50
25
8
16
8
8
26
26.7
23
23
23.6
23.6
35
35
29.7
29.7
35.9
15
31
39.9
29.7
32.1
163
36
36
20R
20R
20R
20R
20R
20R
20R
25D
20R
20R
20R
20R
20R
20R
20R
20R
20R
20R
20R
20R
20R
20R
20R
25D
25D
25D
Bait
type
Main
interference
Burrowing Hermit crab Coconut
crab density1 density1
crab
density1
Seasonal
None
Low
Low
None
Rainforest Burrowing High
Low
Low
Rainforest Hermit
Low
Low
None
Seasonal
Hermit
Low
High
None
Seasonal
Hermit
Low
Medium Low
Seasonal
Hermit
Low
High
Low
Seasonal
Hermit
Low
Low
Low
Savanna
Hermit
Low
High
None
Rainforest Hermit
Low
Low
None
Rainforest Hermit
Low
Medium None
Rainforest Hermit
Low
Medium None
Rainforest Hermit
Low
High
None
Rainforest Hermit
Low
High
None
Rainforest Hermit
Low
Medium None
Rainforest Hermit
Low
Medium None
Rainforest Hermit
Low
Medium None
Rainforest Hermit
Low
Medium None
Rainforest Hermit
Low
Low
None
Rainforest Hermit
Low
Low
None
Rainforest Hermit
Low
Medium None
Rainforest Hermit
Low
Low
None
Rainforest Hermit
Low
Medium None
Rainforest Hermit
Low
Low
None
Rainforest Burrowing High
High
Medium
Seasonal
Hermit
Low
Low
None
Seasonal
Hermit
Low
Low
None
Island type
Co, Ge?
Bi, Co, Gd
Bi, Ca, Co
Co
Bi, Co
Co
Co
Co, Ge
Co
Co, Jo?
Co, Jo?
Co, Jo?
Co, Jo?
Co, Jo?
Co, Jo?
Co, Jo?
Co, Jo?
Co, Jo?
Co, Jo?
Co, Jo?
Co, Jo?
Co, Jo?
Co, Jo?
Bi, Ca, Co
Ca, Co
Ca, Co
Land crab
genera2
Appendix 1 (continued) Details of rodent eradications on tropical islands with land crabs, up to 2015, conducted either by aerial broadcast (Aerial) or hand broadcast (Hand; Hand*=
bait piles on the ground on a grid) of bait. Target species: Mm= Mus musculus, Re= Rattus exulans, Rn= Rattus norvegicus, Rr= Rattus rattus, Rt= Rattus tanezumi; Eradication
result: F= failure, S= success; Bait type: 20R= Pestoff 20R, 25D= Bell-25D, 25W= Bell-25W; Crab genera: Bi= Birgus, Ca= Cardisoma, Co= Coenobita, Ge= Gecarcinus, Gd=
Gecarcoidea, Gg= Geograpsus, Jo= Johngartia, Oc= Ocypode. 1During the eradication; 2Observed by project managers throughout the project.
Island invasives: scaling up to meet the challenge. Ch 1C Rodents: Lessons
A.B. Shiels, D. Will, C. Figuerola-Hernández, K.J. Swinnerton, S. Silander, C. Samra and G.W. Witmer
Shiels, A.B.; D. Will, C. Figuerola-Hernández, K.J. Swinnerton, S. Silander, C. Samra and G.W. Witmer. Trail cameras are a key monitoring tool
for determining target and non-target bait-take during rodent removal operations: evidence from Desecheo Island rat eradication
Trail cameras are a key monitoring tool for determining target and
non-target bait-take during rodent removal operations: evidence
from Desecheo Island rat eradication
A.B. Shiels1, D. Will2, C. Figuerola-Hernández2, K.J. Swinnerton2,3, S. Silander4, C. Samra1 and G.W. Witmer1
USDA, APHIS, WS, National Wildlife Research Center, Fort Collins, CO 80521, USA.
<aaron.b.shiels@aphis.usda.gov>. 2Island Conservation, 2100 Delaware Ave, Suite 1, Santa Cruz, CA 95060, USA.
3
The Island Endemics Foundation, P.O. Box 1908, Boquerón, PR 00622, USA. 4U.S. Fish and Wildlife Service,
Caribbean Islands National Wildlife Refuge Complex, Cabo Rojo, PR 00622, USA.
1
Abstract Efforts to remove invasive rodents (e.g. Rattus spp. and Mus musculus) from islands often use toxicant-laced
baits containing the anticoagulants brodifacoum or diphacinone. Rodenticide baits are generally delivered through aerialor hand-broadcast, or in bait stations. These baits are not rodent-specific and are subject to non-target consumption or
secondary exposure (e.g. an individual preying upon another individual that has consumed bait). During rodenticide
applications, it is generally unknown which animals are visiting and consuming bait; and to quantify this, we recommend
using trail cameras (e.g. Reconyx™ motion-activated infra-red) positioned to monitor individual bait pellets. To
demonstrate the importance and effectiveness of using trail cameras during such operations, we report results of target
(Rattus rattus, black rat) and non-target (native land crab, lizard, insect) bait-interactions after an aerial-broadcast of
Brodifacoum-25D Conservation to eradicate rats from Desecheo Island, Puerto Rico. During the first five days following
bait application, trail cameras (n = 15) revealed that there were 40 incidences of animals contacting bait pellets: 50% rat,
32% hermit crab, 13% Ameiva lizard, and 5% insect. Trail cameras provide temporal and spatial information regarding the
effectiveness of rodent removal, and the last rat pictured by trail cameras on Desecheo was six days after bait application
began. Trail cameras revealed 30 incidences of animals contacting bait pellets 6–20 days after bait application began: 47%
hermit crab, 37% Ameiva lizard, 13% insect, and 3% black crab. Despite viewing ~69,000 images from trail cameras,
lizards were never pictured consuming bait on Desecheo; therefore, any brodifacoum exposure to Desecheo lizards likely
occurred via secondary pathways (e.g. consumption of contaminated insects). Scaling up, we estimate that > 75% of the
total bait distributed on Desecheo was not consumed by rats. Trail cameras help inform the hazards of rodenticide use and
can be easily incorporated into rodent removal operations.
Keywords: aerial rodenticide broadcast, best practice methods, brodifacoum anticoagulant, land crabs, motion-sensing
cameras, Rattus rattus, risk assessment, tropical dry forest
INTRODUCTION
Invasive species, particularly rodents, are among the
greatest threats to native biodiversity on islands. The
breadth of flora and fauna that have been extirpated, or
are currently threatened, by invasive rats (Rattus spp.) and
house mice (Mus musculus) is extensive (see Towns, et
al., 2006; St Clair, 2011; Shiels, et al., 2014). The most
common method to suppress invasive rodent populations,
or eradicate them from islands, is by using toxicantlaced baits such as those containing the anticoagulants
brodifacoum, bromadiolone, or diphacinone (Howald, et
al., 2007; Duron, et al., 2017). These rodenticide baits are
not rodent-specific and are subject to non-target exposure
through their direct consumption of the bait (i.e. primary
exposure) or by an individual preying upon another
individual that has consumed bait directly (i.e. secondary
exposure). Until there is a rodent-specific toxicant
developed that can be effectively delivered to target rodent
species, non-target species that co-habit treatment areas
where rodenticides are used may be at risk to exposure and
possibly death. Therefore, there is a level of risk involved
when using anticoagulant rodenticides that is relevant to
livestock managers and pet owners in domestic settings,
and to conservationists attempting to protect native species
from the negative effects of rodents in natural areas (Hoare
& Hare, 2006).
Existing methods for rodenticide risk assessments
suggest implementation of non-toxic bait-uptake trials
with biomarker-laced bait, and rodenticide residue analysis
of native fauna, both of which can be expensive and may
require harvesting individuals including those that are
threatened or rare (Pott, et al, 2015). Bait uptake trials with
biomarkers are important to determine the level of nontarget exposure to bait, and subsequently help determine
the bait application rates needed at the site. However,
such trials are not always used for island-wide rodent
eradication attempts (Pott, et al, 2015) and rarely used for
rodent suppression projects (Duron, et al., 2017), perhaps
in part because such trials are not a requirement for use of
the rodenticide product, and they necessitate considerable
effort associated with the capture and sampling of the target
and non-target animal community. Although expensive
and requiring the harvest of native animals, rodenticide
residue studies revealed that residues of the used toxicant
establish throughout most of the biological food web
and often result in some non-target animal mortalities
(e.g. Pitt, et al., 2015). The general acceptance of risk
associated with rodenticide use is based on the premise that
benefits to native wildlife outweigh the costs (i.e. native
wildlife populations increase despite losing a few native
individuals from toxicant exposure). A recent example in
Alaska reviewed by Croll, et al. (2016) demonstrates that
the short-term loss of some individuals of native birds
following a rat eradication using brodifacoum has been
overwhelmed by large increases in types and abundances
of native seabirds over the long term.
The use of trail cameras (i.e. motion-triggered infrared cameras) is an underutilised method to assess risk
to non-target animals associated with rodenticide use.
Trail cameras are a means of continuously monitoring
rodenticide bait for animal interactions without having
to be physically present for such observations. Human
observations of animals visiting the bait during rodenticide
applications are rare, due to the inability to watch more
than a few bait pellets at once and the great likelihood of
missing certain animals because of their unique behaviours
during foraging (e.g. being secretive, nocturnal, or confined
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 223–230. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
223
Island invasives: scaling up to meet the challenge. Ch 1C Rodents: Lessons
to particular habitats). Trail cameras can be placed across a
variety of habitats, installed to monitor bait for long periods
(days to months), and reliably record diurnal and nocturnal
visitation while not substantially altering behaviours (some
animals can hear or see cameras/functions; Meek, et al.,
2014) or harming resident animals (Swan, et al., 2004).
When monitoring bait exposure to wildlife, trail cameras
may be less expensive than other methods that require
capturing or harvesting animals, and do not require animal
use permits or animal sampling. Furthermore, the nearly
real-time evidence of bait consumption by target and
non-target species documented by trail cameras provides
the operational staff confidence that the target rodents are
consuming the bait, and allows for adjustments to any
subsequent rodenticide bait applications or non-target
mitigation strategies, if needed.
Desecheo had one of the largest nesting colonies of brown
boobies (Sula leucogaster) in Puerto Rico.
We propose that trail cameras provide critical
information regarding target bait acceptance, effectiveness,
and primary non-target bait exposure during rodent removal
campaigns, and therefore future rodent removal campaigns
should consider employing this tool. To demonstrate how
trail cameras can be used effectively to meet such goals,
we report the results of a field study associated with a
rat eradication project on Desecheo Island, Puerto Rico,
where bait take by target (R. rattus) and non-target animals
(native crab, lizard, insect) were assessed after the aerialbroadcast of Brodifacoum-25D Conservation bait (3 g
pellets, 0.0025% brodifacoum). We used trail cameras
to assess the proportion of bait that rats and non-target
species interacted with, including how much they removed
or consumed, during each of the bait applications. We were
also interested in documenting the spatial and temporal
changes in bait interactions, including when rats were
no longer observed visiting baits. We expected rats to be
early primary consumers of the bait, and their observation
would quickly decline one to two weeks after the first bait
application. Because of the high densities of hermit crabs
(Coenobita clypeatus) on many parts of the island, we
expected that their role in bait consumption and removal
would be formidable and consistent between applications;
yet, we expected much less bait removal and consumption
from other non-targets, such as the three endemic lizard
species that have mostly insectivorous life-histories, and
the few forest birds and seabirds on the island.
Experimental design
Bait application
In March/April 2016 (the dry season), USFWS and
Island Conservation (IC) conducted the bait application
operation on Desecheo using Brodifacoum-25D
Conservation (25 ppm brodifacoum in ~3 g pellets), under
a supplemental label specific to the 2016 eradication effort
(Will, et al., 2019). Bait was applied aerially at 30–45 kg/
ha (depending upon habitat; see Fig. 1) for each of two
applications (18 March and 9 April) in 2016. The 2016
rat eradication attempt used application rates two to three
times greater of Brodifacoum 25-D Conservation than
those used in the unsuccessful 2012 eradication attempt.
There were 11 sites on Desecheo established for
monitoring (Table 1; Fig. 1). These sites were chosen to
occupy the different habitats and bait application regions
(e.g. deflector, coastal overlap, valleys, cliff; Fig. 1) in areas
accessible (often near established trails) on the western
half of the island; the steeply sloped terrain and cliffs were
avoided for safety and logistical concerns. In total, we
established four sites in the ‘interior’ on ridges or slopes,
two sites in ‘valley floor/bottoms’, one ‘cliff’ site, two sites
in the ‘deflector’ zone, which was immediately inland of
the water’s edge and high tide line, and two sites in the
‘coastal overlap’, which was the most inland portion of the
deflector zone and the adjacent inland zone (i.e. interior or
valley floor/bottom). To consistently describe the habitat at
each site, slope and vegetation were described by a single
person (A. Shiels) measuring three variables at each of the
11 sites (Table 1).
At each of the 11 sites, we established a single 150 m
transect that had flags marking each 10 m along the transect.
Transects were established with meter tapes in a straight
line that roughly paralleled walking trails. Once within at
least 150 m of a targeted habitat (i.e. interior, valley floor/
METHODS
Study site and animals
Desecheo (18°23′14″N, 67°28′19″W) is a small (1.2
km2 or 117 ha) island approximately 21 km from the
western shore of the main island of Puerto Rico. The
terrain is rugged with karst limestone as parent material,
and the peak elevation is 218 m. Vegetation is Bursera
simaruba-dominated forest, shrubland, and grassland.
Annual rainfall averages 1020 mm (Seiders, et al., 1972).
The island is a U.S. Fish and Wildlife Service (USFWS)
National Wildlife Refuge. Rattus rattus is abundant on
Desecheo, and was first reported in 1912 (Wetmore, 1918).
The negative impacts of R. rattus to natural areas and
native species on tropical islands are well known (Towns,
et al., 2006; St Clair, 2011; Shiels & Drake, 2011; Pender,
et al., 2013; Shiels, et al., 2013; Shiels, et al., 2014); rats on
Desecheo have been observed eating juvenile lizards and
suspected of consuming other native species (Draft EA,
2015). Desecheo has three endemic lizards (anole: Anolis
desechensis, gecko: Sphaerodactylus levinsi, ameiva
ground lizard: Ameiva desechensis) that may be vulnerable
to rats. Although non-native goats (Capra hircus) and
non-native rhesus monkeys (Macaca mulatta) were
once common to the island, they have been functionally
eradicated (Hanson, et al., 2019). Prior to military actions
and rhesus monkeys being introduced to the island,
224
Fig. 1 Map of Desecheo Island, Puerto Rico, outlining the
different treatment zones for bait application. The entire
island received two applications of Brodifacoum 25D:
Conservation rodenticide bait in 2016 (18 March and
9 April). Bait application rates were 30 kg/ha for both
applications for all parts of the island except the coastal
overlap (#3, #4), cliff faces (#11), and valley floors (#5,
#6), which each received a total of 45 kg/ha during both
applications. For orientation, there are three main valleys
on the island, where (left to right, or west to east) West
Valley (containing #6) is the smallest and most western
(also where camp was set up at the base), Long Valley is
the middle valley (containing #5), and East Valley is the
eastern valley. See Table 1 for details of each site.
Shiels, et al.: Trail cameras are a key
Table 1 Bait application rates (mean +/- SE bait pellets per m2) and ground cover vegetation (0–1 m height) measured
on the ground in 1 m2 plots (n = 15 for each site) along 150 m transects on Desecheo Island, Puerto Rico. Target
application rates were either 30 kg/ha (equivalent to 1 bait pellet per m2), or 45 kg/ha (equivalent to 1.5 bait pellets per
m2, and listed in bold), as each pellet weighed 3.06 ± 0.09 g (n = 49).
Site
No (see
Fig. 1)
1
2
3
4
5
6
7
8
9
10
11
Site
Habitat
description
Deflector #1 (coastline
of Long Valley [L.V.])
Deflector #2 (coastline
of West Valley [W.V.])
Coastal; rocky
with herb/grass
Coastal; sand
with little to no
vegetation
Coastal Overlap #1
Mixed shrubland
(50–80 m inland of high with herbs, grass,
tide line, L.V.)
small trees
Coastal Overlap #2
Thick grassland
(50–80 m inland of high and scattered
tide line, W.V.)
shrubs
Valley Bottom #1 (L.V.) Forest
Valley Bottom #2
(W.V.)
Ridge/Slope #1 (West
Ridge of W.V.)
Ridge/Slope #2 (Headslope of L.V.)
Ridge/Slope #3 (Ridge
and slope of island
peak)
Ridge/Slope #4 (Slope
of L.V. northwest wall)
Cliff (northeast cliff and
windward slope)
Average &
(Maximum) Slope (%)
Canopy
Height (m)
0.2 ± 0.1
2.4 ± 0.8
(2.5 ± 0.4)
0.1 ± 0.0
4.4 ± 2.2
(0.7 ± 0.3)
Forest
Forest edge and
open shrubland
Forest
Forest edge and
open shrubland
Forest
Windswept
shrubland with
herbs and grass
bottoms, cliff, deflector, coastal overlap), the start of a
transect was randomly established by blindly throwing an
object over one’s shoulder while standing on the walking
trail and then beginning the transect from where the object
landed. The 10 m interval flagging marked the location of
the 1 m2 plots for which we monitored bait pellets (15 1 m2
plots per transect; 165 total plots for each application at all
11 sites).
A total of 15 trail cameras (12 Reconyx HyperFire
models HC500 and HC600, and three Browning Model
No: BTC-6HD) were placed to monitor bait pellets to
help identify animals visiting and consuming the pellets.
Each of the 11 sites always had at least one plot with a
trail camera monitoring baits, and some sites had up to
three cameras positioned at randomly assigned plots. Only
one camera was placed per plot, and each camera was
secured to the lower 30–70 cm of a tree or rock. Within
15–120 minutes of the helicopter applying bait to the site,
two bait pellets were gathered from the surrounding 2 m2
of a respective plot and the trail camera was aimed at the
two bait pellets that were placed side-by-side, 40–90 cm
away from the camera. A pin-flag was placed next to the
two bait pellets in each plot so their presence could be
monitored with subsequent visits. All other bait pellets in a
1 m diameter around the pin flag that marked the two target
pellets were removed from the area so as not to confuse
the observer monitoring pellets. The cameras were set to
be triggered by motion, but also were programmed to take
a picture each hour (on the hour), and sometimes more
frequently (15 or 30 min) at set intervals to help account
Application 1
(Pellets/m2)
(March 18,
2016)
1.6 ± 0.4
Application 2
(Pellets/m2)
(April 9, 2016)
1.6 ± 0.3
0.6 ± 0.2
1.6 ± 0.4
1.3 ± 0.2
(4.0 ± 0.1)
7.3 ± 1.5
0.9 ± 0.2
1.2 ± 0.3
0.7 ± 0.1
(3.0 ± 0.2)
4.4 ± 1.5
1.8 ± 0.2
0.7 ± 0.3
3.3 ± 0.1
(7.0 ± 1.1)
3.5 ± 0.2
(9.3 ± 0.7)
2.6 ± 0.3
(6.9 ± 0.7)
3.1 ± 0.3
(7.8 ± 0.7)
2.3 ± 0.2
(5.4 ± 0.5)
15.4 ± 1.9
0.8 ± 0.2
2.1± 0.4
18.4 ± 2.4
1.4 ± 0.3
2.1 ± 0.4
10.4 ± 2.4
1.2 ± 0.3
1.7 ± 0.3
8.0 ± 3.2
1.2 ± 0.3
1.3 ± 0.3
28.1 ± 3.3
1.1 ± 0.2
0.9 ± 0.2
19.6 ± 6.0
0.5 ± 0.2
0.9 ± 0.3
14.3 ± 4.8
0.7 ± 0.2
1.7 ± 0.2
4.2 ± 0.2
(10.4 ± 0.9)
0.8 ± 0.1
(2.9 ± 0.2)
for periods where bait disappeared or was visited without
an animal triggering the camera (e.g. insects rarely trigger
these cameras). Once a Reconyx camera was triggered
by motion, it would take 10 consecutive pictures over 20
seconds; Browning cameras would take one picture each
time triggered.
Cameras were serviced (batteries and SD cards
changed, checked for functioning) as needed, and if both
bait pellets were removed from a plot with a motioncamera, the camera would be moved to another plot
within the site, where bait pellets were still present. Upon
activating the cameras on the day of each bait application,
the baits and cameras were checked daily for at least seven
consecutive days, which was the duration that field staff
was on the island; the bait pellets and cameras were also
checked at day 20 after the first application because that
day preceded the second (and final) application and field
staff had returned to the island.
For our analysis, we scored the number of incidences
where an animal was observed contacting the bait (i.e.
touching, eating, removing). An incidence ended when
the animal left the camera’s field of view, or when a series
of pictures produced by one triggering event ended. The
trail cameras monitored for 27 continuous days, which
began the first day of application one and ended seven
days after application two. Results were presented in three
time-periods: 1) application one until the date rats were
last observed contacting bait (i.e. day five), 2) days 6–20
post-application one, and 3) the first seven days following
application two.
225
Island invasives: scaling up to meet the challenge. Ch 1C Rodents: Lessons
RESULTS
From the 15 cameras deployed, ~38,000 pictures were
taken between application one and two (i.e. 20 days of
continuous monitoring). We reviewed each picture from
all 11 sites, and found 2,686 pictures where an animal was
present. Most of the pictures that captured animals showed
that they were not in contact with the bait, but instead
they were passing by the bait (e.g. ameiva in Fig. 2), or
perhaps searching or foraging nearby the bait. Seventy
pictures from application one showed an animal in contact
with a bait pellet. The first five days following application
one was the only time period that rats were observed in
contact with the bait (18–22 March), and of the 40 pictures
involving animals during this period, 20 were of individual
rats (Fig. 3). Although rats dominated bait contact (Fig.
4) during the first five days following bait application
(especially so during the first two days), hermit crabs (Fig.
5) comprised 32% of bait contact events (Fig. 3). Most
rats and hermit crabs contacting bait either removed it or
consumed it in place. Ameivas, which contacted the bait
in 13% of the pictures during the first five days, usually
had a part of their body (e.g. leg, tail) contacting the bait,
or they occasionally touched it with their snout, or on one
occasion licked the bait and moved out of the frame. Thus,
other than a single lick of the bait, ameivas were never seen
consuming (biting, chewing, swallowing) or removing the
bait. Finally, there were two insects (one appeared to be a
grasshopper) seen in contact with a bait pellet during the
first five days following bait application one (Fig. 3).
The last day when a rat was captured by motioncameras on Desecheo was 23 March, which was the sixth
day following application one. On this day, there was one
rat pictured at Coastal Overlap #2 (grass/shrubland), and
one at Ridge #2 (forest). Neither rat came into contact with
the bait, but instead passed within 12 cm and 1 m of the
bait pellets. These were the last two rats pictured by trail
cameras on Desecheo despite the cameras being active and
bait present in their field of view through to 15 April 2016.
There were 30 pictures from days 6–20 (23 March–7
April) following application one that showed an animal
Fig. 2 An adult ameiva (Ameiva desechensis) triggers a trail
camera positioned to monitor brodifacoum bait pellets
on Desecheo Island, March 2016. Notice the two green
bait pellets at the base of a pin-flag at the lower central
position of the photo. Ameivas rarely were pictured
in contact with the bait and were never documented
consuming or removing the bait pellets.
Fig. 4 A black rat (Rattus rattus) triggers a trail camera
positioned to monitor brodifacoum bait pellets on
Desecheo Island, March 2016. Notice the bait pellet the
rat is nearly touching with its nose. Black rats, being the
target species, were pictured consuming and removing
the bait pellets for the first five days following the first bait
application (18 March 2016).
Fig. 3 Percentage of all trail camera results when an animal
was in contact with a bait pellet (e.g. touching, eating,
removing) during the first five days (18–22 March) after
bait application one, on Desecheo Island, Puerto Rico.
There was a total of 40 animals in contact with bait
during this period (20 rats, 13 hermit crabs, five ameivas,
and two insects), and these pictures were taken on the
following five sites (Cliff, Overlap #2, Deflector #1, Ridge
#4, and Long Valley #1; see Table 1 for site descriptions).
226
Fig. 5 A hermit crab (Coenobita clypeatus) triggers a trail
camera while approaching a bait pellet on Desecheo
Island. Hermit crabs were the primary visitors and
consumers of bait pellets after the first week following
application one.
Shiels, et al.: Trail cameras are a key
in contact with a bait pellet. Because rats were no longer
present or otherwise not pictured by the trail cameras,
the proportion of animals documented contacting the bait
shifted (compare Fig. 3 and Fig. 6), such that hermit crabs
comprised nearly half (i.e. 14 of 30) of the pictures, and
ameivas were pictured contacting the bait in 37% of the
pictures during this period (Fig. 6). Insects, primarily
grasshoppers, were contacting the bait in four pictures,
and there was one picture of a black land crab (Gecarcinus
ruricola) consuming a bait pellet during this period (Figs
6 & 7).
Sites tended to differ in the types of animals, and
their relative abundances, captured on camera contacting
bait pellets. In total, there were only five sites following
application one that had pictures of animals contacting
bait, even though all 11 sites had one to three cameras
monitoring bait pellets and all 11 sites had pictures of
some animals in the view. For example, the Cliff site only
had pictures of hermit crabs contacting bait, whereas the
Deflector #1 site only had pictures of insects (primarily
grasshoppers) contacting bait (Fig. 8). Coastal overlap #2
and Deflector #1 were the only sites that had rats pictured
contacting bait, and Long Valley #1 (valley bottom) and
Coastal Overlap #2 were the only sites that had ameivas
pictured contacting the bait pellets following application
one (Fig. 8). It should be noted here that trail cameras were
only monitoring, although continuously, a small subset of
the total bait applied to Desecheo (i.e. only about 30 baits;
15 cameras monitoring two baits each).
Bait pellets were monitored during the first seven
days following bait application two (Day 21–27), which
occurred on 9 April 2016. There were approximately
31,000 pictures taken and reviewed during this period,
and 176 pictures contained an animal. Similar to our
findings after the first application, most of the pictures
that captured animals showed that they were not in contact
with the bait. There were 16 incidences where animals
were in contact with bait pellets during the week following
application two. There tended to be few proportional
changes in animal-bait interactions that occurred from the
6–20 days of monitoring after bait application one and
the first seven days of bait application two (Day 21–27).
Hermit crabs continued to dominate bait interactions, and
insect consumption of the bait had risen to the highest
proportional levels of all previous measurements (Fig. 9).
Ameiva interactions tended to decrease after application
two relative to the 6–20 days following application one
(Fig. 9). There were five incidences of animals contacting
bait pellets during days six and seven: two hermit crabs,
two insects, and one ameiva; thus, the first five days of bait
interaction would have been similar to the first seven days
of bait interaction. Furthermore, the pictures that captured
animals interacting with bait occurred at five of the 11 sites
(Cliff, Overlap #1, Ridge #1, Ridge #4, and Long Valley
#1) during the week following bait application two. As
Fig. 6 Percentage of all trail camera results when an animal
was in contact with a bait pellet (e.g. touching, eating,
removing) during days 6–20 (23 March-7 April) following
bait application one, on Desecheo Island, Puerto Rico.
There was a total of 30 animals in contact with bait during
this period (14 hermit crabs, 11 ameivas, four insects,
and one black land crab), and these pictures were taken
on the following five sites (Cliff, Overlap #2, Deflector
#1, Ridge #4, and Long Valley #1; see Table 1 for site
descriptions). Note that there were no rats pictured
interacting with bait after five days, and rats were not
pictured at all after six days following bait application
one.
Fig. 7 A black land crab (Gecarcinus ruricola) triggers a
trail camera while consuming a bait pellet on Desecheo
Island. Black land crabs were rarely observed, and only
active at night, on Desecheo Island.
Fig. 8 Percentage of all trail camera results, separated by
site, depicting when an animal was in contact with a bait
pellet (e.g. touching, eating, removing) during the initial
five days (18–22 March), and days 6-20 (23 March-7
April), following bait application one, on Desecheo
Island, Puerto Rico. There was a total of 70 animals in
contact with bait during this period (i.e. 40 during the
initial five days, 30 from 6–20 days), and these pictures
were taken at the following five sites (Cliff, Overlap
#2, Deflector #1, Ridge #4, and Long Valley #1; see
Table 1 for site descriptions). Note that there were no
rats pictured interacting with bait after five days, and
rats were not pictured at all after six days following bait
application one.
227
Island invasives: scaling up to meet the challenge. Ch 1C Rodents: Lessons
with all previous pictures, hermit crabs and insects were
observed consuming bait pellets, yet ameivas were not
seen consuming bait.
DISCUSSION
Trail camera usage during the 2016 rat eradication
on Desecheo Island allowed us to quantify, in near “realtime” fashion, the proportional visitation, removal, and
consumption of bait pellets, and the timing of such visitation,
by target rats and non-target species. Such quantification
of bait interactions allows for upscaling to whole habitats
and an island-wide understanding of the risks to nontarget native species and the potential effectiveness of
the eradication campaign at various timescales following
initial bait application. Initially, most bait interactions
involved rats, and the last rat documented by cameras was
on the sixth day after initial bait application. Non-targets
that consumed, removed, or otherwise contacted the bait
pellets were numerous during the continuous 27 days of
cameras monitoring bait pellets on Desecheo, and hermit
crabs, ameiva lizards, and insects were the main non-target
visitors to the bait pellets. Trail camera usage can therefore
better inform rodent removal campaigns of potential animal
exposure pathways and confirm target bait acceptance as
they are occurring, and therefore should be considered for
future rodent control and eradication operations.
Trail cameras revealed that bait was readily consumed
by invasive rats on Desecheo during the 2016 rat
eradication campaign. Results during the first five days
following bait application, when averaged across all
monitored habitats, revealed half of the bait that animals
on Desecheo interacted (i.e. made contact) with was by
rats, and these were most-likely bait consumption events.
Without implementing trail cameras to monitor bait pellets,
our sole indication that rats were consuming the bait would
have not occurred until four days post-application when
the first rats turned up dead (Shiels, et al., 2017a). Live rats
were rarely observed during the day prior to and following
bait application, and bait was never observed being visited
Fig. 9 Percentage of all trail camera results when an animal
was in contact with a bait pellet (e.g. touching, eating,
removing) during days 0–7 (9–16 April) following bait
application two, on Desecheo Island, Puerto Rico. There
was a total of 16 animals in contact with bait during this
period (seven hermit crabs, three ameiva, six insects),
and these pictures were taken at the following five sites
(Cliff, Overlap #1, Ridge #1, Ridge #4, and Long Valley
#1; see Table 1 for site descriptions).
228
by rats without the aid of trail cameras (Shiels, et al.,
2017a). Furthermore, carcasses of rats may not always be
found because of the expense to keep monitoring crews on
the island for extended periods following bait application,
rodents suffering from toxicosis often die belowground,
and dead rats are quickly scavenged on many islands
with a substantial land crab population (Pitt, et al., 2015).
Although non-toxic bait uptake trials using biomarkers
were performed prior to the 2012 rat eradication attempt
on Desecheo (USFWS, 2011), trail cameras provided
evidence during the 2016 rat eradication that rats were
indeed consuming the bait.
If we use the trail camera findings to scale-up to the whole
island, and assume that all pictures with rats contacting
the bait resulted in the bait pellet being consumed by the
rat, over half of the 5,325 kg of bait that was distributed
across Desecheo in application one, and most (or all) of the
5,325 kg of bait in application two, was not consumed by
rats. Furthermore, > 75% of the bait applied to Desecheo
was consumed by non-target species or did not result
in animal consumption (i.e. the bait disintegrated into
the soil or was consumed by the microbial community).
Clearly, accounting for non-target bait consumption is
a critical part of the best practices associated with initial
determination of bait application rates for island-wide rat
eradications (Pott, et al., 2015). For example, six- to eighttimes as much bait as the Brodifacoum 25W: Conservation
parent label includes was applied to Palmyra Atoll, in the
tropical Pacific, to account for the high density of land crab
populations (Pitt, et al., 2015). Land crabs are a well-known
non-target species that, like all other invertebrates, are not
affected by the brodifacoum toxicant when they consume
the bait, but they render the bait unavailable to target
rodents (Cuthbert, et al., 2012). Our evidence from trail
cameras during the Desecheo rat eradication demonstrates
how common non-target bait interactions can be when
rodenticides, such as brodifacoum bait pellets, are used
for rodent removal. Furthermore, trail cameras revealed
the importance of applying additional bait to Desecheo to
account for non-targets, primarily hermit crabs, rendering
the bait pellets unavailable to rats.
Substantial spatial variation of rat and non-target bait
pellet interactions was present during the period following
bait application on Desecheo, as bait interactions involving
particular animal species tended to differ by habitat (Fig.
8). We must remind the reader that only a very small subset
of the bait pellets applied to Desecheo were monitored
with trail cameras, and there were far more appearances
of animals in the camera view than there were animals
that contacted the bait pellets. Additionally, several of
the sites that had trail cameras continuously monitoring
bait pellets did not have any rats that contacted the bait
pellets. The spatial heterogeneity of rat and non-target
events in various habitats also highlights the need for trail
camera replication, and we feel that our sample size of 15
cameras is modest, and that substantially fewer cameras
would be insufficient for an island of size and habitat
heterogeneity like Desecheo. Additionally, we benefited
from programmed interval-triggering for the cameras
that supplemented motion-triggering because this helped
capture insects and other small or slow-moving animals
that would not trigger the cameras (Newey, et al., 2015).
However, the trade-off of programmed interval-triggering,
and 10 pictures per triggering, is the added human labour
needed to view and analyse the large number of pictures.
Temporal variation of target rodent visitors to bait
pellets can inform operational use of the rodenticide, and
the trail cameras revealing an absence of rats after six days
on Desecheo may suggest modifications to the operation
plan to shorten the length of bait availability on the island.
However, adjustments to operational plans are generally
Shiels, et al.: Trail cameras are a key
made to be more conservative (i.e. more bait for longer
periods) rather than less conservative.
Our concerns of primary brodifacoum bait exposure to
the Desecheo endemic lizard community were abated by
the trail camera results, as there was an absence of pictures
where lizards were observed consuming bait despite their
interactions with the pellets. Additionally, there were
no population level impacts to the lizard community as
observed by the mark-and-recapture work completed in
2012 (Herrera Giraldo, et al., 2019). Ameivas were the only
lizard species that were pictured in contact with the bait
pellets during our monitoring, and there was no evidence of
bait consumption aside from a single lick of the bait pellet
by one individual. Most events where ameivas contacted
the bait were by brushing the tail or legs on the pellet
when passing by. Ameivas, and the other lizard species on
Desecheo, are primarily insectivorous, and are commonly
seen foraging in the leaf litter for insects (Shiels, et al.,
2017a). Based on brodifacoum residue analysis following
bait application, all three endemic lizard species had
detectable levels of brodifacoum in their livers or bodies
(Shiels, et al., 2017a), and the trail cameras and general
diets of these lizards support consumption of contaminated
insects as the most-likely pathway for such brodifacoum
exposure. Although we could not definitively conclude that
insects pictured on the bait pellets were consuming them,
at minimum they would have gained exposure to the bait
through direct contact, which probably facilitated exposure
to higher trophic level predators. We were surprised that
birds, particularly pearly-eyed thrashers (Margarops
fuscatus), were not pictured consuming bait pellets as the
few birds collected for residue analysis had evidence of
brodifacoum exposure (Shiels, et al., 2017a); however, their
omnivorous diet that includes invertebrates and vertebrates
(Wetmore, 1916) favours brodifacoum exposure through
this secondary pathway.
Trail cameras are a cheaper method than residue
analysis to document primary exposure of target and nontarget species during rodenticide campaigns. The USDA
NWRC Chemistry Unit commonly charges between
US$150–US$250 per sample for brodifacoum residue
analysis, and this is a comparable fee to other laboratories.
Additionally, brodifacoum residue analysis generally takes
several weeks to complete. There is a wide price range in
trail cameras, but some of the least expensive trail cameras
can be purchased for <US$100 per camera (e.g. see https://
www.amazon.com/). Inexpensive trail cameras are often
adequate for most rodent removal campaigns because
these cameras produce an image that is identifiable as a
rat or a non-target (e.g. Bushnell brand from 2005 used
in Shiels & Drake (2011)); the reliability, quality of the
image, and flexibility of the cameras in customising
image quality, triggering frequency, and sensitivity are all
factors that are generally better in the Reconyx Hyperfire
cameras (US$450–US$550 for those used in our study;
http://www.reconyx.com/product/Outdoor_Series)
than
the less expensive alternatives (see Newey, et al. (2015)
for a review). An important component that trail cameras
cannot easily produce is evidence of secondary exposure
of non-target species. One could, however, position
rodent carcasses (or non-target carcasses of interest) on
the ground such that trail cameras could document the
scavengers of those carcasses. The potential brodifacoum
exposure of local raptors is worrisome (e.g. Rueda, et
al., 2006), and on Desecheo there are only a few resident
kestrels, and several non-resident raptor visitors (several
species of hawks), that would not be easily observable in
their consumption of carcasses or any mortalities that may
occur from rodenticide exposure on Desecheo.
Prior to rodenticide use, trail cameras can also help
in surveying the potential target and non-target species
at a site. Either singly or in combination with non-toxic
bait uptake trials (Pott, et al., 2015), trail cameras can
inexpensively help identify the potential animals without
catching or harming them. Because rodenticide bait pellets
are a mostly cereal-grain matrix, setting out ‘home-made’
mixtures or placing local fruits and seeds on the ground
with monitoring cameras (see Shiels & Drake, 2011)
may be a first step in determining some of the potential
animal species that may visit rodenticide baits. This may
be applicable for planning purposes, especially on isolated
islands where visits to the island may be short or infrequent.
Additionally, advanced trail camera technology now allows
pictures to be checked remotely, via cellular transmission
of the pictures to a cell phone or email account (Eason, et
al., 2017).
Additional benefits of using trail cameras include
assistance in the confirmation that the target rodent species
is indeed the only rodent species on the island. Trail
cameras producing high quality pictures, and multiple shots
that can reveal multiple angles of the animal, allow for
distinguishing features (e.g. tail length, ear size, body size)
to be revealed and assessed. Furthermore, there are some
occasions where rat-eradications have resulted in surprises
such as house mouse populations ‘suddenly present’, or an
explosion in their abundance, due to the mice being masked
by the dominance of rats prior to rat eradication (Witmer,
et al., 2007); trail cameras would be a viable method to
document and act upon such surprises. Trail cameras may
also be implemented to assess the particular prey (e.g. fruit
and seed) that are most attractive or vulnerable to rodent
predation (e.g. Shiels & Drake, 2011), and to document
biological change after rodent removal by quantifying
before and after native prey survival (e.g. Pender, et al.,
2013). On Desecheo, there was a major caterpillar outbreak
coinciding with rat removal (Shiels, et al., 2017b), and
trail cameras could have been used to better document the
development of the outbreak.
The use of trail cameras is an underutilised method of
risk assessment for rodenticide use, particularly assessing
primary rodenticide exposure that could be a substitute for,
or an improvement upon, more expensive methods that
require animal handling or sacrifice. Trail cameras can be
placed across a variety of habitats, installed to monitor bait
for extensive periods (days to months), and reliably record
diurnal and nocturnal visitation of target and non-target
animals while not substantially altering behaviours or
harming resident animals. Trail cameras provide temporal
and spatial information regarding the effectiveness of
rodent removal, help inform the hazards of rodenticide
use, and can be easily incorporated into rodent removal
operations.
ACKNOWLEDGEMENTS
We are grateful for the multi-agency partnership
(USFWS, IC, USDA, DNER, NZ pilot, AK mechanic)
involved in the Desecheo rat eradication project of 2016.
Field assistance was provided by Erick Bermúdez, Juan
Carlos Alejandro, Ricardo Colón, Juan García-Cancel,
Jorge Gutiérrez, Armando Feliciano, Claudia Lombard,
Ricardo Albarracín, and José Luis Herrera-Giraldo.
REFERENCES
Croll, D.A., Newton, K.M., McKown, M., Holmes, N., Williams, J.C.,
Young, H.S., Buckelew, S., Wolf, C.A., Howald, G., Bock, M.F.,
Curl, J.A. and Tershy, B.A. (2016). ‘Passive recovery of an island bird
community after rodent eradication’. Biological Invasions 18: 703–715.
229
Island invasives: scaling up to meet the challenge. Ch 1C Rodents: Lessons
Cuthbert, R.J., de La Brooke, M. and Torr, N. (2012). ‘Overcoming
hermit-crab interference during rodent-baiting operations: A case study
from Henderson Island, South Pacific’. Wildlife Research 39: 70–77.
Draft E.A. (2015). Restoration of Habitat on the Desecheo National
Wildlife Refuge Through the Eradication of Non-native Rats, Desecheo,
Puerto Rico (Environmental Assessment). Boquerón, Puerto Rico,
USA: USFWS.
Duron, Q., Shiels, A.B. and Vidal, E. (2017). ‘Control of invasive rats and
priorities for future action’. Conservation Biology 31: 761–771.
Eason, C.T., Shapiro, L., Ogilvie, S., King, C. and Clout, M. (2017).
‘Trends in the development of mammalian pest control technology in
New Zealand’. New Zealand Journal of Zoology 44: 267–304.
Hanson, C., Hall, T.J., DeNicola, A.J., Silander, S., Keitt, B. and
Campbell, K. (2019). ‘Rhesus macaque eradication to restore the
ecological integrity of Desecheo National Wildlife Refuge, Puerto
Rico’. In: In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and
C.J. West (eds.) Island invasives: scaling up to meet the challenge, pp.
249–255. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Herrera-Giraldo, J.L., Figuerola-Hernandez, C., Swinnerton, K.J.,
Holmes, N., Bermúdez-Carambot, E.N., Gonzalez-Maya, J.F. and
Gomez-Hoyos, D.A. (2019). ‘Survival analysis of two endemic lizard
species during a rat eradication on Desecheo Island, Puerto Rico’. In:
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West
(eds.) Island invasives: scaling up to meet the challenge, pp. 191–195.
Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Hoare, J.M. and Hare, K.M. (2006) ‘The impact of brodifacoum on nontarget wildlife: Gaps in knowledge’. New Zealand Journal of Ecology
30: 157–167.
Howald, G., Donlan, J.C., Galván, J.P., Russell, J.C., Parkes, J.,
Samaniego, A., Wang, Y., Veitch, D., Genovesi, P., Pascal, M., Saunders,
A. and Tershey, B. (2007). ‘Invasive rodent eradication on islands’.
Conservation Biology 21: 1258–1268.
Meek, P.D., Ballard, G-A, Fleming, P.J.S., Schaefer, M., Williams, W. and
Falzon, G. (2014). ‘Camera traps can be heard and seen by animals’.
PLOS ONE 9(10): e110832.
Newey, S., Davidson, P., Nazir, S., Fairhurst, G., Verdicchio, F., Irvine, R.J.
and van der Wall, R. (2015) ‘Limitations of recreational camera traps
for wildlife management and conservation research: A practitioner’s
perspective’. Ambio 44: 624–635.
Pender, R.J., Shiels, A.B., Bialic-Murphy, L. and Mosher, S.M. (2013).
‘Large-scale rodent control reduces pre- and post-dispersal seed
predation of the endangered Hawaiian lobeliad, Cyanea superba subsp.
superba (Campanulaceae)’. Biological Invasions 15: 213–223.
Pitt, W.C., Berentsen, A.R., Shiels, A.B., Volker, S.F., Eisemann, J.D.,
Wegmann, A. and Howald, G. (2015). ‘Non-target species mortality and
the measurement of brodifacoum rodenticide residues after a rat (Rattus
rattus) eradication on Palmyra Atoll, tropical Pacific’. Biological
Conservation 185: 36–46.
Pott, M., Wegmann, A.S., Griffiths, R., Samaniego-Herrera, A., Cuthbert,
R.J., de L. Brooke, M., Pitt, W.C., Berentsen, A.R., Holmes, N.D.,
Howald, G.R., Ramos-Rendn, K. and Russell, J.C. (2015). ‘Improving
the odds: Assessing bait availability before rodent eradications to aid in
selecting bait application rates’. Biological Conservation 185: 27–35.
Rueda, D., Campbell, K.J., Fisher, P., Cunninghame, F. and Ponder, J.B.
(2006). ‘Biologically significant residual persistence of brodifacoum
in reptiles following invasive rodent eradication, Galapagos Islands,
Ecuador’. Conservation Evidence 13: 38.
Seiders, V.M., Briggs, R.P. and Glover, L. (1972). ‘Geology of Isla
Desecheo, Puerto Rico, with notes on the great southern Puerto Rico
fault zone and Quaternary stillstands of the sea’. US Geological Survey
Professional Paper 739.
Shiels, A.B. and Drake D.R. (2011). ‘Are introduced rats (Rattus rattus)
both seed predators and dispersers in Hawaii?’. Biological Invasions
13: 883–894.
Shiels, A.B., Flores, C.A., Khamsing, A., Krushelnycky, P.D., Mosher,
S.M. and Drake, D.R. (2013). ‘Dietary niche differentiation among three
species of invasive rodents (Rattus rattus, R. exulans, Mus musculus)’.
Biological Invasions 15: 1037–1048.
Shiels, A.B., Pitt W.C., Sugihara, R.T. and Witmer, G.W. (2014). ‘Biology
and impacts of Pacific island invasive species. 11. Rattus rattus, the
black rat (Rodentia: Muridae)’. Pacific Science 68: 145–184.
Shiels, A.B., Witmer, G.W., Samra, C., Moulton, R.S., Ruell E.W.,
O’Hare J.R., Eisemann J.D., Volker, S.F. and Goldade, D.A. (2017a).
Assessment of Bait Density, Bait Availability, and Non-target Impacts
During an Aerial Application of Rodenticide to Eliminate Invasive Rats
on Desecheo Island, Puerto Rico. Final Report QA 2588. Ft Collins,
Colorado, USA: USDA, APHIS, WS, NWRC.
230
Shiels, A.B., Haines, W.P., Swinnerton, K.J., Silander, S., FiguerolaHernández, C., Will, D., García-Cancel, J.G, and Torres-Santana,
C.W. (2017b). ‘Sudden appearance and outbreak of Eunica monima
(Lepidoptera: Nymphalidae) on Desecheo Island, Puerto Rico’. Florida
Entomologist 100: 176–179.
St Clair, J.J.H. (2011). ‘The impacts of invasive rodents on island
invertebrates’. Biological Conservation 144: 68–81.
Swan, C.E., Hass, C.C., Dalton, D.C. and Wolf, S.A. (2004). ‘Infraredtriggered cameras for detecting wildlife: an evaluation and review’.
Wildlife Society Bulletin 32: 357–365.
Towns, D.R., Atkinson, I.A.E. and Daugherty, C.H. (2006). ‘Have
the harmful effects of introduced rats on islands been exaggerated?’.
Biological Invasions 8: 863–891.
USFWS. (2011). Final Environmental Assessment: Desecheo National
Wildlife Rat Eradication to Promote Ecosystem Restoration. Puerto
Rico: Caribbean Islands National Wildlife Refuge Complex. Boquerón,
Puerto Rico, USA: USFWS.
Wetmore, A. (1916). ‘Birds of Puerto Rico’. U.S. Department of
Agriculture Bulletin 326: 1–140.
Wetmore, A. (1918). ‘The birds of Desecheo Island, Porto Rico’. The Auk
35: 333–340.
Will, D., Swinnerton, K., Silander, S., Keitt, B., Griffiths, R., Howald,
G.R., Figuerola-Hernandez, C.E. and Herrera-Giraldo, J.L. (2019).
‘Applying Lessons Learnt from Tropical Rodent Eradications: a Second
Attempt to Remove Invasive Rats from Desecheo National Wildlife
Refuge, Puerto Rico’. In: In: C.R. Veitch, M.N. Clout, A.R. Martin,
J.C. Russell and C.J. West (eds.) Island invasives: scaling up to meet
the challenge, pp. 154–161. Occasional Paper SSC no. 62. Gland,
Switzerland: IUCN.
Witmer, G.W., Boyd, F. and Hillis-Starr, Z. (2007). ‘The successful
eradication of introduced roof rats (Rattus rattus) from Buck Island
using diphacinone, followed by an irruption of house mice (Mus
musculus)’. Wildlife Research 34: 108–115.
Sposimo, P.; D. Capizzi, T. Cencetti, F. De Pietro, F. Giannini, C. Gotti, F. Puppo, G. Quilghini, E. Raganella Pelliccioni, G. Sammuri, V. Trocchi, S. Vagniluca, F. Zanichelli and N. Baccetti. Rat and
lagomorph eradication on two large islands of central Mediterranean: differences in island morphology and consequences on methods, problems and targets
Rat and lagomorph eradication on two large islands of central
Mediterranean: differences in island morphology and
consequences on methods, problems and targets
P. Sposimo1, D. Capizzi2, T. Cencetti1, F. De Pietro3, F. Giannini3, C. Gotti4, F. Puppo3, G. Quilghini5,
E. Raganella Pelliccioni4, G. Sammuri3, V. Trocchi4, S. Vagniluca6, F. Zanichelli3 and N. Baccetti4
NEMO srl, Piazza D’Azeglio 11, Firenze I-50121, Italy. <sposimo@nemoambiente.com>. 2Regione Lazio, Direzione
Ambiente e Sistemi Naturali, Via del Pescaccio 96, Roma I-00166, Italy. 3Parco Nazionale Arcipelago Toscano, Enfola,
Portoferraio LI I-57037, Italy. 4ISPRA, via Cà Fornacetta 9, Ozzano Emilia BO I-40064, Italy. 5UnitàTerritoriale
Carabinieri per la Biodiversità, via Bicocchi 2, Follonica GR I-58022, Italy. 6Gruppo Carabinieri Forestale, via degli
Asili, Livorno I-57126, Italy.
1
Abstract Montecristo and Pianosa islands, although approximately equal in surface area (c. 1,000 ha), differ greatly
in substrate, human presence, vegetation and altitude (650 m vs. 30 m asl, respectively). The former island hosts one of
the largest yelkouan shearwater (Puffinus yelkouan) populations in Italy, the latter a depleted remnant of once numerous
Scopoli’s shearwaters (Calonectris diomedea). Two consecutive EU-funded LIFE projects have been designed to protect
these seabird populations. On Montecristo, rough and inaccessible, aerial delivery of toxic baits in January-February 2012
eradicated black rats (Rattus rattus) and feral rabbits (Oryctolagus cuniculus) (originally a non-target species), with no
permanent consequences on a local, ancient population of wild goats (Capra hircus). Eradication on Pianosa, currently
underway (started January 2017), is being performed by ground baiting, delivered by 4,750 dispensers placed on a 50 m
× 50 m grid throughout the island. The latter operation is included in a multi-species eradication aimed at several other
target species, among which was the brown hare (Lepus europaeus), apparently introduced around 1840. Genetic analyses
on the first trapped hares showed that this was the last uncontaminated and viable population of L. europaeus subsp.
meridiei in existence. Whether of natural origin or introduced, the commencement of eradication of this population has
instead created the awareness of a taxon otherwise unavailable for conservation elsewhere. While both projects address
the same conservation issues (protection of shearwater colonies and restoration of natural communities), they differ
greatly regarding economic cost, public perception, effort needed to maintain results in the long term and effects on nontarget species. In the present paper, specific attention has been paid to the comparison between bait delivering techniques,
results obtained, the array of problems originating from the complex regulatory framework and reactions by the general
public.
Keywords Capra hircus, Lepus europaeus meridiei, Montecristo, Oryctolagus cuniculus, Pianosa, Rattus rattus, Tuscan
Archipelago
P. Sposimo, D. Capizzi, T. Cencetti, F. De Pietro, F. Giannini, C. Gotti, F. Puppo, G. Quilghini, E. Raganella Pelliccioni, G. Sammuri, V. Trocchi, S. Vagniluca, F. Zanichelli and N. Baccetti
INTRODUCTION
We present here two eradications of invasive mammals
carried out on Mediterranean islands, directly concerning
black rat (Rattus rattus), but also affecting two species
of lagomorphs (feral rabbit (Oryctolagus cuniculus) and
brown hare (Lepus europaeus)) and the wild goat (Capra
hircus) (Table 1). These actions are part of a comprehensive
recovery programme of nesting areas of seabirds in the
Italian islands (Capizzi, et al., 2016). The two islands
concerned, Montecristo and Pianosa, belong to the Tuscan
Archipelago National Park and are almost equal in size but
very different in morphology, vegetation, fauna and human
presence. These differences influenced the eradication
approaches.
Montecristo is a 1,080 ha island located in the central
northern Mediterranean Sea, in an intermediate position
between Corsica and the Italian Peninsula (Fig. 1), with
a rugged topography and a maximum altitude of 650 m.
It is uninhabited, and access is strictly limited. There are
very few trails and no roads. As a consequence, the only
realistic method to eradicate rats was the aerial distribution
of bait. The main conservation target here was the
yelkouan shearwater (Puffinus yelkouan), with 400–750
breeding pairs whose reproductive success was heavily
affected by predation on eggs and chicks by the black rat
(Baccetti, et al., 2009). A population of feral rabbits was
also present. Given the bait distribution technique chosen,
the eradication of this species was considered possible,
although unlikely, and was not declared as a project target.
The species considered at risk of unwanted mortality
(Table 1) by direct consumption of baits (<http://www.
montecristo2010.it>) were mainly the yellow-legged
gull (Larus michahellis) and Montecristo wild goat.
The latter was introduced on the island in pre-Roman
times from founders that were still at a very early stage
of domestication (or just tamed). The species’ historical
origin, together with its current uniqueness in the western
Mediterranean, motivated its role as a flagship species and
led to the founding of Montecristo State Nature Reserve
in the early 1970s. The cultural/historical value of this
peculiar population makes it deserving of appropriate
conservation efforts (Gotti, et al., 2014).
Pianosa is a 1,040 ha island, that is entirely flat (< 30
m altitude), 30 km NW of Montecristo (Fig. 1) with a
Fig. 1 Location of Montecristo and Pianosa islands.
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 231–235. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
231
Island invasives: scaling up to meet the challenge. Ch 1C Rodents: Lessons
small village and some scattered (and usually abandoned)
settlements. It was occupied as a prison until 1998,
making it inaccessible even for researchers during that
time. Currently it is permanently inhabited by 20–30
detainees with two–three guards and it is open to guided
tours with a daily limitation of 330 visitors. The extensive
road network is maintained in reasonably good condition;
vegetation is relatively accessible, especially on formerly
cultivated areas (roughly half of the island). The main
conservation target is Scopoli’s shearwater (Calonectris
diomedea), threatened by black rat predation on eggs and
chicks (Table 1), consisting of 30–50 nesting pairs on
Pianosa and 150–250 on La Scola, a satellite islet located
240 m to the east of Pianosa (Capizzi, et al., 2016). Rats
were removed from La Scola, only 1.6 ha in size, in 2001;
however, the short distance from the main island allows
periodical rat incursions (three in the period 2001–2011, by
single individuals), which till now have been successfully
eliminated (Capizzi, et al., 2016) by a set of bait stations
permanently installed and refilled when necessary.
Black rat eradication on Pianosa is part of a multispecies eradication programme aimed at the restoration of
the native animal community <(http://www.restoconlife.eu/
en>), which originally included the removal of the brown
hare (Lepus europaeus). The house mouse (Mus musculus)
is widespread on the island and is not an explicit eradication
target of the project (Table 1), due to the spacing of the
bait stations chosen for an island of the size of Pianosa and
for the primary target species. The existence of permanent
settlements, the presence of tourists during the summer and
the occurrence of several non-target species, together with
legal constraints on distribution methods, forced the choice
of a ground-based eradication. Diurnal raptors, owls,
yellow-legged gulls and hooded crows (Corvus cornix)
and finally some domestic cats (Felis catus) are among the
non-target species potentially threatened by the operation
(Table 1). For non-domesticated feral cats living in the wild
another specific eradication action has been conducted.
METHODS AND RESULTS
Rodent eradications
On Montecristo the first aerial baiting was conducted
on 8 and 9 January 2012, with a pellet density of 10 kg/ha
on the ground. The baits consisted of 2 g cereal (Brocum®,
0.005% brodifacoum as active ingredient, produced by
Colkim Ltd).
A 30 ha area was excluded from the aerial drop (Fig.
2). This included unoccupied human dwellings, and an
enclosure of about 25 ha, where 44 wild goats (at least 24%
of the population size, assessed through direct counts, the
rest having remained free) were kept to ensure survival of
the population. This area was treated either by bait stations
in the goat enclosure hand-broadcast of pellets outside
the goat enclosure. Wax blocks (Solo Blocs®, 0.005%
brodifacoum as the active ingredient, produced by Bell
Ltd) were installed in the bait-stations.
The aerial distribution was originally planned along
parallel transects, 50 m apart, to obtain a roughly complete
overlap between parallel transects. However, the pilot
Fig. 2 Location of the goat enclosure, houses, and bait
stations involved in the Montecristo biosecurity plan.
had difficulties flying along such predefined routes
without veering significantly from the flight line, as the
geographic positioning system (GPS) based guidance
system malfunctioned and caused frequent interruptions
of the baiting. We changed our plans and opted in favour
of baiting along two different sets of 100 m wide parallel
flight lines, at right angles to each other. This would allow
a greater tolerance for the helicopter’s distance from the
scheduled flight-lines and smaller areas without overlap.
We then covered the coast and endeavoured to cover
obvious gaps (Fig. 3). The second distribution of baits
was initially expected to occur two weeks later, but the
exceptionally dry weather that allowed pellets to persist
on the ground in good shape and adequate amounts,
made it unnecessary for a much longer time. The second
baiting was done 45 days after the first delivery, covering
110 ha only, corresponding to the most critical areas: the
coastline, a buffer zone around the excluded areas and an
area where the first distribution appeared to have been less
than optimal. The bait density was lower than in the first
Table 1 Species involved in the Montecristo and Pianosa rat eradication projects. Among non-target species, strong
negative impacts on local populations have been recorded on Montecristo for the breeding pair of Corvus corax (with
permanent reoccupation of the site in 2015–2016) and on Pianosa for Tyto alba.
Conservation target
species
Montecristo Puffinus yelkouan
Invasive species
Pianosa
Rattus rattus
Island
232
Calonectris diomedea
Rattus rattus
Invasive non target
Non target species
species
Oryctolagus cuniculus Larus michahellis, Capra hircus,
Corvus corax
Mus musculus
Larus michahellis, Corvus cornix,
owls, diurnal raptors
Sposimo, et al.: Rat & lagomorph central Mediterranean
Fig. 4 Bait consumption on Pianosa during the first four
sessions.
Lagomorphs
Fig. 3 Helicopter flight lines during the first bait drop.
distribution, i.e. 4 kg/ha on the ground. The eradication
was successful, as the last sign of rats was detected 15 days
after the first distribution (Sposimo, 2014). The cost of the
whole operation, excluding preliminary analyses, planning
and devices for protection of goats, was €226,800 (US
$280,000).
On Pianosa Island, the eradication took place via
ground-baiting with second generation anticoagulants
inside bait-stations, placed at the nodes of a 50 m × 50 m
grid covering the entire island. Bait density was doubled
along the coastline and in urban areas, in consideration
of locally higher rat densities. Approximately 4,750 baitstations were deployed in January 2017, then checked
and refilled monthly until May 2017. The percentage of
consumed baits and/or any sign of rats were recorded. Bait
stations were retrieved in October 2017. The bait consisted
of wax-blocs with brodifacoum (Solo Blocs®, 200 or 20
g), except in the area occupied by human settlements where
it was replaced by wax-blocks with 0.005% bromadiolone
(Notrac Blox®, 225 or 28 g, produced by Bell Ltd) during
the first and second baiting events, to reduce the risk of
secondary poisoning of domestic animals (Buckle &
Smith, 2015).
Rates of bait consumption are detailed in Fig. 4. After
the initial and very high rate of bait consumption, the rate
decreased by one order of magnitude during each of the
two successive sessions; low and steady final values were
assumed to be mainly due to house mice and invertebrates.
In May, only one credible sign of black rat was found
across the entire island, but the success of the eradication
has to be confirmed by the implementation of monitoring
activities that are still ongoing; the presence of house
mouse was detected in nine bait-stations, suggesting that
eradication of this species, as expected, is unlikely. This
result would be consistent with other evidence of house
mice being more difficult to eradicate than Rattus species
(MacKay, et al., 2007), because of smaller home-range
size that allows the survival of some individuals in baitfree areas between 50 m spacing of stations. The cost of the
eradication, excluding preliminary analyses and planning,
was €477,600 (US $590,351).
Feral rabbit: the aerial delivery of rat bait took place
on Montecristo during mid-winter, when feral rabbits
appear to be in the most critical annual phase and when
local decreases in the population had been observed.
Nevertheless, the presence of a 25 ha fenced area where the
bait distribution had been implemented with bait-stations
inaccessible to rabbits, and the lack of any specific effort
to cull surviving rabbits (Murphy, et al., 2010), made the
eradication of this species quite unlikely. Nonetheless,
rabbits do seem to have been eradicated, probably due to
the rabbit-permeable fencing of the goat enclosure and
unusually long duration of pellets outside it, because of
the dry weather. After the bait delivery of February 2012
there was just a single observation of rabbits (an adult
with young) in April 2014. Traps were immediately set
in the observation area without any result. The lack of
any further finding of individuals or sign more than three
years later makes it unlikely that any survivors might still
be present on the island. Despite the lack of any specific
monitoring, rabbits would not have escaped detection in
the surroundings of the fenced gardens, where full time
employed wardens are present, or along the transects
regularly covered for goat counts.
Brown hare: the Pianosa multi-species eradication
programme originally included netting and shooting
brown hares, which had likely been introduced on the
island for hunting purposes as early as 1840. Tissue and
blood samples obtained from 35 individuals netted at the
beginning of the operation, or shot with hunter-dog teams,
were collected and genetically analysed, leading to the
unpredicted result that the Pianosa brown hare population
represented the last uncontaminated and viable population
of Lepus europaeus meridiei, a subspecies once distributed
in central and northern Italy, believed to be extinct due
to genotype contamination with exotic hares massively
imported in Italy for hunting (Mengoni, et al., 2018).
Pianosa’s insular status, habitats, and past restricted access
due to the prison have resulted in maintaining a viable
and representative stock of high conservation value. The
number of individuals captured and culled in early 2016
had no consequence on the population size after one to two
breeding seasons.
Biosecurity
Reinvasion risk varies greatly between the two islands,
because of the different human presence and accessibility
(Russell, et al., 2008b). On Montecristo an incoming rat
would likely follow a single pathway covering the small
area of the wharf and buoys. After the eradication, a
biosecurity interception system was set up consisting of 15
bait stations placed all around the entry point at the landing
233
Island invasives: scaling up to meet the challenge. Ch 1C Rodents: Lessons
bay. We experimentally tested the effectiveness of this
system in December 2016, releasing 14 black rats, all adult
males, equipped with VHF transmitters. Animals were
released on the pier individually, over 19 hours, simulating
a reinvasion event. Each bait station was armed with a
lethal snap-trap. Individuals were tracked for 65 hours.
Twelve individuals were intercepted by a bait station, the
majority (10) within 20 m from the release point, while one
disappeared shortly after release (and is suspected of dying
from hypothermia after having been observed swimming)
and the last one escaped the interception system to hide
in a stone wall 100 m inland from the wharf. The average
time between release and capture was 3.4 h, which is very
short if compared with the results of similar experiments
on brown rats (Russell, et al., 2008a). The biosecurity
system was modified after these results and organised in
three sub-systems (Fig. 2): one permanently active (eight
bait-stations), concentrated in the vicinity of the wharf, and
two more to be activated in case of potentially rat-infested
boats docking at the wharf or buoys. A plan for contingency
response has been set-up as well.
Biosecurity measures to be implemented in Pianosa,
beginning in October 2017, are directed both towards
ferries, to reduce the presence of rodents on board, and
towards the implementation of an island-based interception
system, roughly following Montecristo’s scheme. In the
likely case of an unsuccessful mouse eradication, the land
system will require a more frequent bait replenishment,
together with a permanent mouse-control strategy in the
harbour area.
months after bait delivery, where gull casualties were
recorded for at least a four month period. On Pianosa,
the operators who checked bait-stations every month
collected all corpses they found. Standardised counts were
performed on both islands to assess any negative effects
on non-target populations. Deaths recorded on Montecristo
only occurred for two species, the wild goat (n=35) and
yellow-legged gull (n=891), while the local pair of
common raven (Corvus corax) was no longer observed,
indicating presumed extirpation. Ravens permanently
reoccupied the site only in 2015–16. Annual monitoring
of the Montecristo goat population by distance sampling
methods showed a temporal decrease of approximately
30–40% in the summer following the aerial treatment,
while counts performed in all subsequent years attested
to a fast recovery of the population to the pre-eradication
level. On the contrary, yellow-legged gulls dropped from
1,036–1,833 breeding pairs in the two years before baits
were delivered to 591 in 2012, 292 in 2013, with a steady,
slight increase in all following years, up to 499 in 2017.
On Pianosa the impact of rodenticide indirectly
consumed by native predators was more diverse, more
concentrated in time, but less thoroughly recorded:
findings of fresh corpses ceased around mid-March 2017
and included nocturnal raptors of two species (eight
individuals), diurnal raptors of three species (seven
individuals) and at least three hooded crows; no gulls were
affected. Effects on native populations at Pianosa seem
to be limited to the expected extirpation of breeding barn
owls (Tyto alba).
Effects on conservation targets
DISCUSSION
The effects of the Montecristo eradication on its
conservation target species are shown in Table 2, where
a dramatic increase in breeding success of yelkouan
shearwater has taken place since the year of the bait
delivery (2012), whereas breeding performance of
Scopoli’s shearwater on Pianosa has constantly remained
poor. Evidence of new breeding sites, including nest boxes,
being occupied by yelkouan shearwaters on Montecristo
since 2012 is available, but an increase of nest density or
population size remains to be quantified as of yet. A number
of benefits were recorded on non-target avian species, such
as minor increases of breeding scops owl (Otus scops)
and European nightjar (Caprimulgus europaeus), and an
obvious increase of sedentary and alien chukar partridges
(Alectoris chukar).
Black rats have been successfully eradicated on
Montecristo and Pianosa seems to be on a similar trajectory,
thanks to two projects performed with radically different
techniques.
In order to maintain these achievements – and
investments – in the long term, different efforts are needed.
Biosecurity measures are relatively simple for Montecristo,
as long as the current management of access is allowed. A
field test has shown that currently adopted measures are
adequate, suggesting minor adjustments enabling a slight
reduction of effort needed for their maintenance.
Reinvasion risk is significantly higher for Pianosa,
this island being affected by a permanent, yet currently
moderate, flow of supplies and visitors, that could
strongly increase in the near future due to already planned
restoration of many buildings. This, together with the
probable survival of the house mouse, results in the need
for more complex and costly (due to bait consumption
by mice) biosecurity measures. Risks for native species
and the insular ecosystem deriving from a house mouse
increase following black rat eradication was considered to
be low, due to the presence of several species of specialised
or generalist predators of rodents (three breeding species of
owls and one of snake).
The unexpected disappearance of rabbits from
Montecristo can likely be related to timing of the
operations, that coincided with seasonal lows of the
population, and to random factors such as a very unusual
and prolonged drought for the season (January–April
rainfall of 34.6 mm in 2012, vs. an average of 112.1 mm in
the same period for the previous five years), which allowed
longer bait persistence and possibly impacted rabbits
more strongly. Similarly, unexpected results for different
reasons were obtained in the case of the Pianosa brown
hare, representing a taxon believed to be extinct and, thus,
deserving appropriate management in future.
Effects on conservation target species were, and are
expected to be, very positive. Although this is easily
understandable – and already evident for Montecristo
Effects on non-target species
Thorough searches for gull and goat corpses were
repeatedly carried out on Montecristo from one to four
Table 2 Breeding success of the two conservation target
species on Montecristo (black rats eradicated 2012)
and Pianosa (treated 2017, first rat free season still in
progress).
Year
2010
2011
2012
2013
2014
2015
2016
234
Montecristo target:
Puffinus yelkouan
No.
Reprod.
nests
success
18
0.06
19
0.96
28
0.93
27
0.78
26
0.80
35
0.80
Pianosa target:
Calonectris diomedea
No.
Reprod.
nests
success
6
0.17
16
0.19
19
0.16
17
0.12
Sposimo, et al.: Rat & lagomorph central Mediterranean
yelkouan shearwaters, based on the resulting local
population size and productivity – in the case of Pianosa,
for Scopoli’s shearwaters a full evaluation should include: i)
the huge potential for breeding sites, most of them currently
unused by the depleted breeding stock; and ii) the effortless
maintenance of permanently rat-free conditions on adjacent
La Scola islet, where a large Scopoli’s shearwater colony is
already present. Moreover, since Pianosa is almost devoid
of burrowing seabirds at present, but has suitable breeding
sites, its value in the future attraction of species that are
currently absent (e.g. yelkouan shearwaters from nearby
Montecristo and Mediterranean storm petrel [Hydrobates
pelagicus melitensis]) might even exceed its importance
for Scopoli’s shearwater.
Consequences on non-target species have varied
greatly, depending on bait deployment methods, geomorphological features and faunal composition of the two
islands. Non-target mortality of Montecristo goats did not
prevent the recovery of the population to its initial level in
a few years and, after the aerial treatments, the widespread
presence of goats did not limit the availability of baits
for rats. The presence of the large enclosure to protect
some goats and prevent population extirpation has to be
considered as a prudent measure to ensure the long-term
persistence of this valuable population. However, it posed
a risk to the success of the rat eradication and demanded
an alternative approach (bait delivery inside bait-stations).
The higher mortality of diurnal and nocturnal raptors
observed on Pianosa (and probably underestimated) can
be attributed to several factors: 1) their greater abundance,
2) the presence of house mice that are preyed upon by
small-sized raptors that do not feed ordinarily on rats,
and 3) possibly also the delivery of poison through bait
stations, that may allow rodents to consume a much higher
amount of poison than during an aerial distribution. The
most striking difference between the casualties of the two
projects was the massive impact of the aerial treatment on
the yellow-legged gulls, compared to the absence of any
effect on this species in the bait station-based operation.
Losses could have been minimised with an aerial delivery
planned earlier in the season, when fewer birds are on
the breeding sites. Nevertheless, even these losses – of
a human-dependent and super-abundant species – are
negligible compared to the benefits achieved. A slightly
earlier seasonal planning, however, has to be recommended
in consideration of possible public reactions to the issue
of gull mortality, which was a population decrease lasting
more than one year following the operation’s conclusion.
Losses of barn owls on Pianosa, and their probable (albeit
possibly only temporary) extirpation, represent possibly
the highest biological cost of the programme.
Both projects triggered negative reactions from
the public, particularly harsh towards the Montecristo
eradication. Evidently, aerial baiting was perceived as a
more threatening method by non-experts. Thus, projects
planning to use this technique should employ greater
communication efforts at different levels. The strategy
for communicating with the public, structured as in many
other LIFE projects (e.g. via a dedicated website) was
clearly ineffective, despite the projects being on islands
lacking human populations which should have restricted
the potential audience. People from nearby towns on the
coast or from nearby Elba Island were often unresponsive
to any outreach efforts and usually exploited debated topics
in favour of other agendas, such as anti-Park personal
positions, or criticising the ‘waste’ of money (Baccetti,
et al., 2016). Ambiguity of national regulations in force
during the Montecristo eradication led to legal actions
being taken against project managers, but were finally
positively concluded for the defendants. Currently, the EU
Biocide Regulation 528/2012 clarifies the situation, but the
possibility to carry out aerial baiting remains undefined,
depending on specific authorisations issued by nationally
competent authorities.
Aerial baiting has allowed the black rat to be eradicated
from an island primarily relevant for Mediterranean seabird
conservation that could not be otherwise treated with
traditional ground-based methods. Tangible drawbacks
are not larger than those observed during a comparable
operation implemented by a bait station distribution, while
the economic cost was certainly lower. Nevertheless, at
present, the opportunity to carry out similar operations is
extremely uncertain in Italy, as well as across the rest of
the EU.
ACKNOWLEDGEMENTS
Both operations described have been realised thanks
to EU LIFE programme co-financing, as for the majority
of Italian eradications. Project monitors Michele Lischi
and Chiara Caccamo were particularly helpful. Many
people offered their collaboration in the course of the
programmes, starting from Jacopo Primicerio (Institute
of Biometeorology of the National Research Council) and
Barbara Lastrucci (NEMO Ltd), responsible for the GIS and
GPS guidance system during the Montecristo eradication,
the permanent staff of the two islands (Montecristo wardens
Luciana Andriolo and Giorgio Marsiaj, as well as several
Pianosa prison guards and detainees) and many colleagues
from the UTB Forestry office of Follonica, Tuscan
Archipelago National Park, Penitentiary offices of Porto
Azzurro and ISPRA. We are most grateful to the reviewers
and editors for their comments and English edits.
REFERENCES
Baccetti, N., Capizzi, D., Corbi, F., Massa, B., Nissardi, S., Spano, G.
and Sposimo, P. (2009). ‘Breeding shearwaters on Italian islands:
population size, island selection and co-existence with their main alien
predator, the black rat’. Rivista Italiana di Ornitologia – Research in
Ornithology 78(2): 83–100.
Baccetti, N., Capizzi, D. and Sposimo, P. (2016). ‘Rat Eradications on
Italian Islands for the Conservation of Breeding Seabirds’. In: P. Yésou,
J. Sultana, J. Walmsley and Azafzaf (eds.) Conservation of Marine and
Coastal Birds in the Mediterranean. Proceedings of the UNEP-MAPRAC/SPA Symposium, Hammamet, pp.106–113. Tunisia: The Regional
Activity Centre for Specially Protected Areas.
Buckle, A.P. and Smith, R.H. (eds.) (2015). Rodent Pests and their
Control. Wallingford: CAB International.
Capizzi, D., Baccetti, N. and Sposimo, P. (2016). ‘Fifteen Years of Rat
Eradication on Italian Islands’. In: F.M. Angelici (ed.) Problematic
Wildlife, pp. 205–227. Switzerland: Springer International Publishing.
Gotti, C., Palladini, A. and Raganella Pelliccioni, C (2014). ‘La Capra di
Montecristo una Popolazione Unica in Mediterraneo/ The Montecristo
Goat a Unique Mediterranean Population’. In: F. Zanichelli, F. Giannini,
F. De Pietro and F. Puppo (eds.) Quaderni del Parco, Documenti Tecnici,
volume 2 PROGETTO LIFE + MONTECRISTO 2010. Eradication of
Invasive Alien Plants and Animals and Protection of Native Species
and Habitats in the Tuscan Archipelago, pp. 34–43. Portoferraio: Parco
Nazionale Arcipelago Toscano.
MacKay, J.W.B., Russell, J.C. and Murphy, E.C. (2007). ‘Eradicating
Mice from Islands: Successes, Failures and the Way Forward’. In: G.W.
Witmer, C.W. Pitt and K.A. Fagerstone (eds.) Managing Vertebrate
Invasive Species: Proceedings of an International Symposium, pp.
294–304. Fort Collins: USDA/APHIS/WS, National Wildlife Research
Center.
Mengoni, C., Trocchi, V., Mucci, N., Gotti, C., Giannini, F., Mallia, E.,
Geminiani, C. and Baccetti, N. (2018). ‘The secret of Pianosa island:
An Italian native population of European brown hare (Lepus europaeus
meridiei Hilzheimer, 1906)’. Conservation Genetics: 10.1007/s10592018-1077-4.
Murphy, E., Crowell. M. and Henderson, W. (2010). Workshop
Proceedings: Improving the Efficiency of Rabbit Eradications on
Islands. 4–5 February 2010, Christchurch. Canberra, Australia:
Invasive Animals Cooperative Research Centre.
Russell, J.C., Beaven, B.M., MacKay, J.W.B., Towns, D.R. and Clout,
M.N. (2008a). ‘Testing island biosecurity systems for invasive rats’.
Wildlife Research 35: 215–221.
Russell, J.C., Towns, D.R. and Clout, M.N. (2008b). ‘Review of Rat
Invasion Biology. Implications for Island Biosecurity’. Science for
Conservation 286. Wellington, NZ: Department of Conservation.
Sposimo, P. (2014). ‘The Eradication of the Black Rat on Montecristo’. pp.
20–25. I Quaderni del Parco, documenti tecnici volume 2. Portoferraio:
Parco Nazionale Arcipelago Toscano, <www.montecristo2010.it/
stealthV3_pubblica/0840566ARG4637055061.pdf>.
235
236
Chapter 2: Other taxa
With Sections: A Mammals
B Birds
C Herpetofauna
D Invertebrates
E Plants
F Aquatic
237
D. Algar, M. Johnston and C. Pink
Algar, D.; M. Johnston and C. Pink. Big island feral cat eradication campaigns: an overview and status update of two significant examples
Big island feral cat eradication campaigns: an overview
and status update of two significant examples
D. Algar1, M. Johnston1 and C. Pink2
Science and Conservation Division, Department of Biodiversity, Conservation and Attractions, Locked Bag 104,
Bentley Delivery Centre, Western Australia, Australia 6983. <dave.algar@dbca.wa.gov.au>. 2Christmas Island
National Park, P.O. Box 867, Christmas Island, Indian Ocean, Australia 6798.
1
Abstract Feral cats have been known to drive numerous extinctions of endemic species on islands. Also, predation by
feral cats currently threatens many species listed as critically endangered. Island faunas that have evolved in the absence
of predators are particularly susceptible to cat predation. Australian islands, such as Dirk Hartog and Christmas, both
formerly known to be high biodiversity islands, are no exception. In this paper we outline the techniques being used in the
two eradication campaigns currently underway and provide an update on the status of the programmes on these islands
that differ significantly in terms of climate, topography and habitation. Poison baiting and trapping are the methods used
on both islands but have been managed differently to suit the local conditions.
Keywords: Christmas Island, Dirk Hartog Island, domestic cat, fence, land crab, planning, poison baiting
INTRODUCTION
There is extensive evidence that domestic cats (Felis
catus) introduced to offshore and oceanic islands around
the world have had deleterious impacts on endemic land
vertebrates and breeding bird populations (e.g. Van Aarde,
1980; Moors & Atkinson, 1984; King, 1985; Veitch, 1985;
Bloomer & Bester, 1992; Bester, et al., 2002; Keitt, et
al., 2002; Pontier, et al., 2002; Blackburn, et al., 2004;
Martinez-Gomez & Jacobsen, 2004; Nogales, et al.,
2004; Ratcliffe, et al., 2009; Bonnaud, et al., 2010). Feral
cats have been known to drive numerous extinctions of
endemic species on islands and have contributed to at least
14% of all 238 vertebrate extinctions recorded globally by
the IUCN (Nogales, et al., 2013). In addition, predation by
feral cats currently threatens 8% of the 464 species listed
as critically endangered (Medina, et al., 2011; Nogales, et
al., 2013). Island faunas that have evolved for long periods
in the absence of predators are particularly susceptible to
cat predation (Dickman, 1992). Dirk Hartog and Christmas
Islands, both documented as high biodiversity islands are
no exception.
Dirk Hartog Island (DHI), an area of 620 km2, is the
largest island off the Western Australian coast (Abbott &
Burbidge, 1995). Since the 1860s, DHI has been managed
as a pastoral lease grazed by sheep (Ovis aries) and goats
(Capra hircus). More recently, tourism has been the main
commercial activity on the island undertaken by the former
pastoralist family, the only permanent inhabitants on the
island. Cats were probably introduced by early pastoralists
and became feral during the late 19th century (Burbidge,
2001). Ten of the 13 species of native terrestrial mammals
once present are now locally extinct (Baynes, 1990;
McKenzie, et al., 2000) probably due to predation by cats
(Burbidge, 2001; Burbidge & Manly, 2002; Algar, et al.,
2011a). The house mouse (Mus musculus) has become
established on the island, but other invasive species such as
European rabbit (Oryctolagus cuniculus), red fox (Vulpes
vulpes) and black rat (Rattus rattus) are not present.
Christmas Island (CI) occupies an area of 135 km2 and
is famous for the annual migration of tens of millions of
red crabs (Gecarcoidea natalis) (Orchard, 2012; Misso &
West, 2014). CI has a resident multi-cultural population
of 2,239 residents (2015 records, <http://www.abs.gov.
au/>), predominately Chinese, Malays and Europeans,
who reside on the north-eastern tip of the island. Phosphate
mining is a major economic driver on the island, with
ecotourism becoming increasingly important. Cats were
taken to CI at the time of first settlement in 1888 and a
feral population established soon thereafter (Tidemann,
et al., 1994). Four of the five mammal and two reptile
species that were present on the island at settlement have
since become extinct, with the introduction of cats playing
a crucial role (Beeton, et al., 2010; Martin, et al., 2012).
Two endemic rats, the bulldog rat (Rattus nativitatis) and
Maclear’s rat (R. macleari) disappeared shortly after black
rats were introduced in 1900 (Green, 2014). In addition,
several extant CI birds are listed as species likely to be
adversely affected by cats (Beeton, et al., 2010).
Across Australia, cats have caused or contributed to
population declines and extinctions on many offshore
islands (Dickman, 1992; Dickman, 1996; Burbidge, et al.,
1997; Burbidge, 1999). Today, the impact of cats is broadly
acknowledged and control of feral cats is recognised as
one of the most important fauna conservation issues in
Australia. As a consequence of this, a national ‘Threat
Abatement Plan (TAP) for Predation by Feral Cats’ has
been developed (EA, 1999; DEWHA, 2008; DE, 2015).
The TAP seeks to protect affected native species and
ecological communities, and to prevent further species
and ecological communities from becoming threatened. In
particular, the first objective of the TAP is to “prevent feral
cats from occupying new areas in Australia and eradicate
feral cats from high-conservation-value islands”.
DHI was established as a National Park in November
2009, and this now provides the opportunity to reconstruct
the native mammal fauna (Algar, et al., 2011a). The island
could potentially support one of the most diverse mammal
assemblages in Australia and contribute significantly to
their long-term conservation. Successful eradication of feral
cats is considered to be a necessity prior to reintroductions.
Similarly, the impact of cats on much of the biodiversity
of CI has been of significant concern to island land
management agencies and local residents. Eradication of
cats on the island is necessary to mitigate the socio/health
impacts and threat to those remaining extant species and
to allow successful re-wilding of species such as the bluetailed skink (Cryptoblepharus egeriae) that are currently
restricted to captive breeding programmes.
The islands differ markedly in environmental and
human factors but are linked in the agencies involved, that
have iteratively resolved site-specific challenges associated
with the removal of cat impacts on wildlife populations.
In this paper we outline the cat eradication programmes
currently underway on both islands, describe the strategies,
techniques and application methodology and provide an
update on the campaigns’ progress.
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
238
up to meet the challenge, pp. 238–243. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Algar, et al.: Two cat eradication examples
MATERIALS AND METHODS
Site descriptions
DHI (25° 50’ S, 113° 0.5’ E) lies within the Shark Bay
World Heritage Property of Western Australia, 1.5 km
from mainland Australia. The island is approximately 79
km long and a maximum of 11 km wide with its long axis
in a south-east to north-west direction. Detailed description
of geology and vegetation is provided elsewhere (Beard,
1976; Payne, et al., 1987; Algar, et al., 2011a). The climate
of the region is ‘semi-desert Mediterranean’ (Beard, 1976;
Payne, et al., 1987). The mean annual rainfall for Denham,
located 37 km to the east of DHI is 224 mm (Bureau of
Meteorology, 2017; long-term records 1893–2016).
CI (10° 25’ S, 105° 40’ E) is located in the Indian Ocean,
360 km south of the Indonesian capital of Jakarta. The
oceanic island is composed primarily of Tertiary limestone
overlying volcanic andesite and basalt (Tidemann, et al.,
1994; EA, 2002). The island consists of a series of fringing
limestone terraces, separated by rugged limestone cliffs and
scree slopes, rising to an internal central plateau at about
200 m and extending to 360 m above sea level. A National
Park was established in 1980 and extended in 1986 and
1989 to include most of the rainforest; it now covers 63%
of the island (EA, 2002). There are four main vegetation
types described in detail by Claussen (2005). CI has a
typical tropical, equatorial climate with a wet and a dry
season. The wet season is from December to April when the
north-west monsoon blows. For the rest of the year southeast trade winds bring slightly lower temperatures and
humidity, and much less rain. The island has a mean annual
rainfall of 2,183 mm, high humidity (80–90%) which varies
little between months and consistent temperatures (mean
daily temperature: 22.9–27.4o C) (Bureau of Meteorology,
2017).
Planning
To date, feral cats have been successfully eradicated
from four Western Australian offshore islands: Serrurier
Island (Moro, 1997); Hermite Island in the Montebellos
(Algar, et al., 2002); Faure (Algar, et al., 2010) and Rottnest
Islands (Algar, et al., 2011b) to enable reconstruction of
the original fauna or protection of extant species. These
successes and knowledge gained provide the confidence
to tackle the more ambitious challenges of DHI and CI.
There is a number of key elements used in the operational
planning of a successful eradication strategy. The plan may
include a pilot study that assesses the efficacy of proposed
techniques as well as documenting the procedures to be
used in the sequenced eradication phases, the monitoring
programmes and the surveillance period prior to verifying
eradication has been achieved. Plans for the DHI/CI
eradication programmes build strongly on previous research
conducted on both islands that examined eradication and
monitoring techniques (Algar & Brazell, 2008; Algar, et
al., 2010; Algar, et al., 2011a).
Central to the planning for DHI was the construction
of a 13 km cat barrier fence. The island’s size, in particular
its length, poses logistical constraints on conducting
an eradication campaign across the entire island
simultaneously. It is not practical to monitor for cat activity
over such a large area and therefore, the eradication
campaign is being conducted in stages either side of the
barrier fence. The fence was constructed with a ‘floppy
top’ and electrical hotwires facing to the north to prevent
reinvasion of the southern area once it had been cleared
(see Fig. 1). Use of a barrier fence has been demonstrated
to reduce the cost and increase the overall likelihood of
successful eradication on the island (Bode, et al., 2013).
Crucial in the planning for CI was the presence of a
domestic cat population. Key land management agencies
initiated the preparation of a cat management plan as a
Fig. 1 Dirk Hartog Island.
critical first step. The plan (Algar & Johnston, 2010) was
developed with these agencies, interest groups and the
broader community. It was supported and endorsed by
the various organisations and has been embraced by the
public. Initially, local cat management laws were revised to
include a prohibition on the importation of cats, promoting
responsible cat ownership, compliance and enforcement
of cat management laws. A staged approach to eradicate
cats entirely from the island has been adopted, which is
complemented by the gradual decrease of owned cats as the
de-sexed domestic population dies out. The amended local
legislation required all domestic cats to be neutered, microchipped and registered with the Shire (Stage 1). Surveys
of domestic cats and veterinary programmes are outlined
by Algar, et al. (2011c) and Algar, et al. (2014). Stage 2
requires the removal of all stray cats within the township.
Without implementation of Stage 2 a significant source
of cats, particularly natal recruits, would be available to
disperse into or reinvade territories vacated across the
rest of the island. Stage 3 involves the implementation
of the island-wide (i.e. the national park, mine leases and
Unallocated Crown Land) feral cat eradication programme.
Eradication effort
Baits and baiting application
Baiting is recognised as the most effective method for
controlling feral cats on mainland Australia (Short, et al.,
1997; EA, 1999; Algar & Burrows, 2004; Algar, et al.,
2007; Algar, et al., 2013), and has been used as the primary
technique for eradicating cats on islands (Algar, et al.,
2002; Algar, et al., 2010). World-wide, cat eradications
have been attempted on a number of islands with 82
successful campaigns that range in size from 5–29,000 ha
(Campbell, et al., 2011). There have also been eradication
attempts on a further 15 islands that have failed (ibid.). All
successful campaigns on islands >2,500 ha used primary
poisoning with toxic baits, with the exception of Santa
239
Island invasives: scaling up to meet the challenge. Ch 2A Other taxa: Mammals
Catalina (3,020 ha). Interestingly, seven failed campaigns
on the five largest islands (all >400 ha) did not use toxicants
(Campbell, et al., 2011). A locally developed bait known as
Eradicat® (Algar & Burrows, 2004) containing 4.5 mg of
directly injected toxin ‘1080’ (sodium monofluoroacetate)
is used on both DHI and CI.
The primary eradication technique to be used in
the DHI programme was aerial baiting. A pilot study
conducted during March–May 2009 assessed the efficacy
of this strategy (Johnston, et al., 2010; Algar, et al., 2011a).
This achieved very positive results with 80+% of the feral
cat population poisoned following bait consumption (ibid).
These results demonstrated that a baiting programme, with
the Eradicat® bait as the primary eradication technique,
could be highly effective on DHI.
Deployment of baits from an aircraft was not considered
feasible on CI at the commencement of this campaign
due to the removal of baits by the abundant land crabs.
However, targeted aerial baiting into discrete difficult to
access areas is now being contemplated for late in the dry
season when land crabs are less active (Johnston, et al.,
2016). Preliminary baiting exercises on the island where
baits were placed on the ground, highlighted the potential
problem of non-target species removing ground-laid baits.
Red crabs, robber crabs (Birgus latro), which dominate
the forest floor, black rats and feral chickens (Gallus
domesticus) readily removed baits laid on the ground. Bait
removal by non-target species reduces bait availability to
feral cats and therefore eradication efficacy. In a later trial,
Algar & Brazell (2008) demonstrated a device to suspend
baits above the ground that effectively reduced bait
removal by non-target species yet provided ready access
to feral cats. A key finding from this trial was that the bait
suspension devices (BSD) would provide an effective
primary cat eradication technique on the island. During the
eradication campaign, BSD are located at 100 m intervals
on both sides of the extensive 160 km road/track, staggered
at 50 m intervals across the road/track. Each BSD suspends
two Eradicat® baits tied at the link, considered a single
bait for analysis purposes, at a height of about 550 mm
using 6–8 lb fishing line. Baits are replaced when taken
and as required to maintain palatability. Suspended baits
were also deployed off-track throughout the forest at 50 m
intervals in 2015 and, due to unprecedented rainfall, to a
lesser extent in 2016.
The total number of toxic baits removed indicates the
maximum number of individuals poisoned. The minimum
number of individuals poisoned is calculated by ascribing
bait removals from consecutive BSDs to the same animal.
The actual number of feral cats poisoned would be between
these two extremes. While one Eradicat® bait contains a
lethal dose, it is likely that some cats would visit multiple
BSDs given the delay between bait consumption and death.
programmes was used to provide a measure of baiting
efficacy.
On CI, the trapping programme commenced in the
township to remove stray cats. Initially, cage traps were
used rather than padded leg-hold traps to minimise the
risk of injury to domestic cats. Cats were captured using
wire cage traps (60 × 20 × 20 cm) with treadle plates
(Sheffield Wire Products, Welshpool, Western Australia).
All traps were covered with a hessian sack to provide
shelter and protection to the captured animals until they
could be collected. The traps were usually baited with
cooked chicken wings. Outside the township, elevated trap
platforms (ETPs) – where trap sets are raised above ground
level – are used to exploit cats’ agility and ability to jump,
while preventing trap interference from ground-dwelling
non-target wildlife such as land crabs. Traps along roads
and tracks are generally set on cleaned half 200 l fuel/oil
drums in the same configuration and lured as ground sets
on DHI.
Monitoring
Monitoring programmes use evidence of actual
presence through camera trap images, spotlight records
and sign, whether it be footprints, scats or hair, to detect the
presence/absence of individuals in an area. In eradication
campaigns, monitoring programmes provide information
on where further effort is required and whether additional
measures and/or resources are needed. A key component
of these eradication campaigns is to employ monitoring
methods that will provide quantitative estimates of the
effectiveness of eradication operations; the techniques
must also be capable of detecting animals at low density
populations. The physical characteristics of DHI and CI
differ significantly and required the adoption of a different
suite of monitoring techniques across the two islands.
Of necessity, the monitoring of feral cat activity must
be conducted across the entire island. CI has an extensive
road/track network (see Fig. 2) whereas, on DHI, much
of the former pastoral road network has regenerated,
with many roads and fence lines being impassable. The
monitoring programme on DHI is being conducted from
All Terrain Vehicles (ATVs) which can traverse the entire
island in a safe and efficient manner. Prior to implementing
the monitoring programme, it was necessary to construct
a network of survey tracks to allow monitoring of cat
activity across the island. The spacing of these tracks
Trapping
Trapping programmes are being used as the secondary
eradication effort to remove those animals that survive the
baiting programmes. On DHI, cats are being captured in
padded leg-hold traps; (Victor ‘Soft Catch’ traps No. 3
(Woodstream Corp., Lititz, PA.; U.S.A.) using a mixture of
cat faeces and urine as the lure. Trapped cats are destroyed
using a 0.22 calibre rifle. All animals captured are sexed
and weighed; a broad estimation of age (as either kitten,
juvenile or adult) is recorded using weight as a proxy for
age. The pregnancy status of females is determined by
examining the uterine tissue for embryos. Stomach contents
are removed for diet analysis and a sample of hair and
tissue taken for DNA microsatellite profiling. Also, prior to
the commencement of the two aerial baiting programmes,
a number of cats were trapped and fitted with a GPS datalogger/radio-telemetry collar (Sirtrack Ltd, New Zealand).
Mortality of radio-collared animals following the baiting
240
Fig. 2 Christmas Island.
Algar, et al.: Two cat eradication examples
needed to permit detection of any cat during the survey
period (i.e. two weeks each month) and therefore provide
confidence in the sensitivity of the survey technique.
Information obtained from the GPS data-logger radiocollars during the pilot study (Algar, et al., 2011a) was used
to determine the likelihood of detection and to optimise the
proposed spacing of the survey tracks for the eradication
programme. Track lines were parallel to the long axis of
the island and the orientation of the dune system. This was
the preferred course for survey tracks for logistic reasons
and also to minimise disturbance and erosion to dunes.
Analysis of daily movement patterns, pooled for all cats,
suggested that placement of monitoring tracks at a width
of approximately 2.0 km across the full length of the island
(see Fig. 1) would be sufficient to enable detection of these
animals within each survey period. Choice of this spacing
for the monitoring tracks and separation of camera traps
(see later) was further strengthened by data collected on
home ranges (100% Minimum Convex Polygon) of the
radio-collared cats in the pilot study which were 12.7 km2
for males and 7.8 km2 for females (Johnston, et al., 2010).
Thus, every cat has a very high probability of its sign being
observed over a 10-day monitoring period (Algar, et al.,
2011a).
Camera trapping
Camera trap studies are useful in providing information
on feral cat presence/absence and provide an ideal technique
for monitoring the impact of eradication measures through
the progression of the eradication campaign as they will
allow remote monitoring of cats following each period. On
DHI, camera traps were established at 2 km intervals along
and overlooking the track network with 105 Reconyx
HC600 (Reconyx, Wisconsin; USA) cameras north of
the barrier fence and 64 cameras to the south (see Fig. 1).
Additional cameras were installed at key locations such as
fence ends and around the tourist resort on freehold land. A
variety of visual, olfactory or audible attractants were used
at camera sites, including no lure. On CI, 84 Scoutguard
SG-560C (HCO Outdoors, Norcross, GA, USA) non-lured
camera traps were located approximately 1.0 km apart in
an island-wide array, with six spatially explicit transects
nested within (see Fig. 2). Occupancy analysis and spatial
mark/resight modelling was conducted to estimate density
over time (The Analytical Edge Pty Ltd., Hobart, Australia).
Sign searches
The sandy surface on DHI enables the search and
detection of cat footprints. The network of management
tracks is searched daily by skilled observers riding ATVs
over a 10-day period on a seasonal basis, that is, four
times per year. Circuits ranging in length from 80–140
km, are ridden at a speed of <20 km/h which is adequate
to identify footprints on the track surface. The observers
alternate the direction of travel and the circuit they inspect
on a daily basis. The track surface on CI is hard and does
not lend itself to identification of footprints. Other sign
monitoring techniques are currently being developed that
will complement the use of camera traps to survey for cat
activity.
Surveillance period and independent verification
The final phase of the campaign on DHI, an intensive
and simultaneous island-wide surveillance period was
initiated in October 2016 on the belief that that eradication
had been achieved. Assuming no more cats are found, this
third phase is expected to be of a two-year duration and will
be used to confirm eradication success in October 2018.
On DHI, surveillance monitoring for cat activity
is being conducted over a 10-day period in each of
the southern and northern sectors every three months.
Surveillance monitoring is employing both camera trap
recording and cat sign searches. The cat sign searches
are being conducted along the pre-existing tracks and the
monitoring grid network. Opportunistic cat sign searches
along beaches and other areas of interest (e.g. caves and
seabird colonies) are also being conducted. The monitoring
is undertaken across the entire zone the same day to avoid
any issues associated with cat movement.
In addition, on DHI specialist detector dogs and their
handlers (Latitude 42 Environmental Consultants Pty
Ltd., Tasmania, Australia) have been contracted to further
independently verify the absence of cats and corroborate
that eradication has been successfully achieved. A team of
six dogs and experienced handlers undertake the intensive
search effort for cat sign during the winter when weather
conditions are the most favourable.
On CI, surveillance monitoring, which is yet to
commence, will primarily utilise the island-wide camera
array with a range of lures as on DHI. Detector dogs are
not being considered for use on CI because of quarantine
regulations for re-importation back onto the mainland,
the difficult terrain and cultural issues associated with the
presence of dogs on the island. A community reporting
system will be maintained as well as implementing an
intensive and comprehensive spotlighting effort around the
island.
Finally, independent verification of eradication success
on both islands is to be undertaken by an impartial
organisation using data summaries provided.
RESULTS
Dirk Hartog Island
Logistical issues associated with transport of fencing
materials prevented construction of the fence on DHI until
following the completion of baiting monitoring in 2014.
As a result, most of the island (90%) was baited in 2014.
However, once completed, the fence alignment has played
a key role in restricting the ranging of cats on the northern
side.
Data on cat home range size and degree of overlap from
the 2009 pilot study were used to derive a best estimate
of cat population size pre-eradication effort. This analysis,
with multiple assumptions, suggested that a total of 439 cats
(range 309–503) was likely present prior to the eradication
campaign. Prior to the first baiting campaign in 2014, 17
cats were trapped and fitted with VHF/GPS collars in the
southern zone during April 2014. Trapped cats were released
at the location of capture. Of these, fifteen were known to
be alive when Eradicat® baits were applied on the 27–28
May. Fourteen of these animals (>90%) died following bait
consumption. The fate of the remaining cat is uncertain
but as it was last detected alive in June 2014 and has not
be relocated by VHF or photographed since this time.
Five cats were trapped, fitted with VHF/GPS collars and
released at the location of capture in the northern sector in
April 2015 prior to the second baiting programme. All were
alive when baits were applied on 25 May 2015. Only one
of these cats died following consumption of an Eradicat®
bait, the remaining four were recovered by trapping.
The combined monitoring programmes have detected 36
individual cats that survived the baiting programmes and
these animals have subsequently been trapped. January
and April seasonal surveillance programmes have failed
to detect the presence of any further cat activity. Detector
dogs did not locate any fresh sign of cat activity south of
the barrier fence in 2016 and examine the area north of the
fence during July 2017.
Christmas Island
One hundred and eighty-four domestic owned cats have
been registered on CI since 2010, with only 74 domestic
241
Island invasives: scaling up to meet the challenge. Ch 2A Other taxa: Mammals
cats remaining at the conclusion of the 2017 domestic
cat survey. Deregistered cats had either died from natural
causes or road fatalities, or were euthanised as the owners
had moved off-island. Although the programme on CI
commenced in late 2010, funding to commence the islandwide eradication effort (Stage 3) was not secured until
2015. Short-term control programmes were conducted
around the township in 2013 and 2014 to protect the
significant investment and gains achieved in controlling
stray cats until a new source of funding could be obtained.
Over the period 2011 to 2015, 336 stray/feral cats were
trapped within the township and a further 216–311 were
poisoned along roadsides/tracks that surrounded the area.
From 2015 to 2017, cage trapping removed 46 stray cats
within the township, outside the township limited ETP leghold trapping resulted in the removal of (12), shooting (11)
and roadside BSDs a further (158–216) cats. An unknown
number of cats was removed from forest baiting in 2015
and 2016 due to uncertainty in determining bait uptake
by cats. Based on the upper and lower estimation method
of baits taken on BSD, between 779–932 stray/feral cats
have been removed since 2010. Preliminary results from
the 2016 island-wide array camera monitoring estimated
that a population of 225 (SE=23) feral cats remains across
the island.
DISCUSSION
Globally, the Dirk Hartog project will become the
largest island feral cat eradication campaign attempted to
date and Christmas Island is a relatively large island with
significant human inhabitation. The restoration of former
species richness on DHI and recovery of the threatened
wildlife populations on CI has required management of
feral and domestic cats. The strategies used to achieve
the reduction in cat populations have been tailored to
suit the specific circumstances applicable to each island.
Perhaps the largest challenge on DHI was to ensure that
the monitoring tools were sufficiently sensitive given
the scale of the island. Removal of cats from Christmas
Island is characterised by improving the management of
owned cats as well as mitigating the impact of land crabs
on poison baiting operations. The guiding principles for
successful eradication (Bomford & O’Brien, 1995) have
been successfully met in both of these island programmes,
although it is worth noting that maintaining the appropriate
socio-political environment has been an ongoing and
time-consuming component of both programmes. The
eradication programmes on both islands have followed a
logical progression of intensiveness that aimed to reduce the
population rapidly from base levels and then use follow-up
trapping to target remaining cats, that is, initial population
knockdown with a low cost/broad-scale method followed
by high cost/labour intensive mopping up. The monitoring
programmes suggest that the cat populations have been
reduced to low (CI) or non-detectable (DHI) levels bearing
out the prescriptions provided in the operational plans.
Residents on CI have been involved in the development
and maintenance of the owned cat population. This has
also involved a compliance programme and importation
ban that was necessary to maintain the closed population.
Maintaining quarantine on DHI has been a more
straightforward process given that one family is involved
who are invested in the ecological restoration of the island
given the anticipated benefits to their tourism enterprise.
Poison baiting has formed a critical part of the
eradication tools on both islands but the variable results
achieved in these programmes should be noted in
preparation for similar programmes in the future. A low
baiting efficacy consequently leads to a requirement for
greater follow-up control with respect to investments in
time and labour. A probable explanation for the observed
242
differences in baiting success in 2014 and 2015 on DHI
relates to the meteorological factors at the time of baiting.
Just prior to 2014 baiting, a pulse of cooler weather was
recorded which would have had the effect of reducing the
availability of alternative prey such as small reptiles and
mammals. In contrast, the 2015 season was characterised
by a rodent irruption that may have been triggered by
rainfall associated with Tropical Cyclone Olwyn. On CI,
unprecedented rainfall in 2016 reduced baiting efficacy
significantly and prompted the development of alternative
trap sets that were effective under wet conditions.
Alternative removal tools, such as different trap sets, must
be ready to implement in situations where baiting is less
successful (Robinson, et al., 2015). Project governance
and budgeting would ideally include sufficient contingency
to adapt or permit operational flexibility to account for
environmental factors that influence on-ground outcomes
(Springer, 2016).
Ultimately, the success of these programmes will be
measured by the response of native wildlife species. On CI,
there has been a dramatic increase in the nesting success
rate in the red-tailed tropicbird (Phaethon rubricauda)
populations following improvement in the management of
urban cats (Algar, et al., 2012) as well as anecdotal reports
of a positive response in forest birds such as the Christmas
Island emerald dove (Chalcophaps indica natalis). It is
premature to make claims about the recovery of extant
species on DHI other than to note detections of species on
cameras which were not detected in 2014. These include
the little long-tailed dunnart (Sminthopsis dolichura),
painted button quail (Turnix varius) and bush stone curlew
(Burhinus grallarius). The wildlife response on DHI will
be intensively monitored in subsequent years during the
ecological restoration of the island.
ACKNOWLEDGEMENTS
The Dirk Hartog Island Ecological Restoration Project
is funded by the Gorgon Barrow Island Net Conservation
Benefits Fund. The DHI cat team comprises Gary
Desmond, Jason Fletcher, Neil Hamilton, Mike Onus and
Cameron Tiller. The Christmas Island Cat Eradication
Project is currently funded by the Director of National
Parks, the Threatened Species Commissioner and an
environmental mining offset acquired from Phosphate
Resources Ltd (Christmas Island). Funding has also been
provided by Department of Infrastructure and Regional
Development, the Shire of Christmas Island, Christmas
Island Phosphates and Border Force. The CI cat team
comprises Neil Hamilton, Jason Fletcher, and Cameron
Tiller together with the island-wide survey baiting team
Dion Maple, Rob Muller, Samantha Flakus, Matt Hudson,
Tanya Detto, Jol Muller, Trent Lane, Joanna Chan, Simone
Richardson, Brendan Tiernan, Renata de Jonge, Jason
Turl, Sean White, Simon Pahor and John Jaycock. The
Department of Biodiversity, Conservation and Attractions
Animal Ethics Committee approved protocols 2009/35,
2012/41 and 2015/39 which describe activities undertaken
in this project.
REFERENCES
Abbott, I. and Burbidge, A.A. (1995). ‘The occurrence of mammal
species on the islands of Australia: a summary of existing knowledge’.
CALMScience 1(3): 259–324.
Algar, D., Burbidge, A.A. and Angus, G.J. (2002). ‘Cat eradication on the
Montebello Islands’. In: C.R. Veitch and M.N. Clout (eds.) Turning the
tide: the eradication of invasive species, pp. 14–18. Occasional Paper
SSC no. 28. IUCN SSC Invasive Species Specialist Group. IUCN,
Gland, Switzerland and Cambridge, UK.
Algar, D. and Burrows, N.D. (2004). ‘Feral cat control research: Western
Shield review, February 2003’. Conservation Science Western Australia
5(2): 131–163.
Algar, D., Angus, G.J., Williams, M.R. and Mellican, A.E. (2007).
‘Influence of bait type, weather and prey abundance on bait uptake
by feral cats (Felis catus) on Peron Peninsula, Western Australia’.
Conservation Science Western Australia 6(1): 109–149.
Algar, et al.: Two cat eradication examples
Algar, D. and Brazell, R.I. (2008). ‘A bait-suspension device for the
control of feral cats’. Wildlife Research 35: 471–476.
Algar, D., Angus, G.J., Brazell, R.I., Gilbert, C. and Withnell, G.B.
(2010). ‘Eradication of feral cats on Faure Island, Western Australia’.
Journal of the Royal Society of Western Australia 93: 133–140.
Algar, D. and Johnston, M. (2010). Proposed Management Plan for Cats
and Black Rats of Christmas Island. Western Australian Department of
Environment and Conservation. 63 pp.
Algar, D., Johnston, M. and Hilmer, S.S. (2011a). ‘A pilot study for
the proposed eradication of feral cats on Dirk Hartog Island, Western
Australia’. In: C.R. Veitch, M.N. Clout and D.R. Towns (eds.) Island
invasives: eradication and management, pp. 10–16. Occassional Paper
SSC no. 42. Gland, Switzerland: IUCN and Auckland, New Zealand:
CBB.
Algar D, Angus G.J. and Onus M.L. (2011b). ‘Eradication of feral cats
on Rottnest Island, Western Australia’. Journal of the Royal Society of
Western Australia 94: 439–443.
Algar, D., Hilmer, S., Nickels, D. and Nickels, A. (2011c). ‘Successful
domestic cat neutering: first step towards eradicating cats on Christmas
Island for wildlife protection’. Ecological Management and Restoration
12(2): 93–101.
Algar, D., Hamilton, N., Holdsworth, M. and Robinson, S. (2012). ‘Cat
removal implicated in improved nestling success in iconic red-tailed
tropicbird rookery’. Landscope 27(4): 43–47.
Algar, D., Onus, M.L. and Hamilton, N.A. (2013). ‘Feral cat control
as part of ‘Rangeland Restoration’ at Lorna Glen (Matuwa), Western
Australia: the first seven years’. Conservation Science Western Australia
8(3): 367–381.
Algar, D., Hamilton, N., and Pink, C. (2014). ‘Progress in eradicating
cats, Felis catus, on Christmas Island to conserve biodiversity’. Raffles
Bulletin of Zoology Supplement No. 30: 44–52.
Baynes, A. (1990). ‘The Mammals of Shark Bay, Western Australia’.
In: P.F. Berry, S.D. Bradshaw and B.R. Wilson (eds.) Research in
Shark Bay—Report of the France-Australe Bicentenary Expedition
Committee, pp. 313–325. Perth, WA: Western Australian Museum.
Beard, J. (1976). Vegetation Survey of Western Australia, Sheet 6. 1:1 000
000 Map Sheet and Explanatory Notes. Perth, Australia: UWA Press.
Beeton, B., Burbidge, A., Grigg, G., Harrison, P., How, R., McKenzie, N.
and Woinarski, J. (2010). Final Report of the Christmas Island Expert
Working Group to Minister for the Environment Protection, Heritage
and the Arts. Canberra: nt of the Environment, Water, Heritage and the
Arts.
Bester, M.N., Bloomer, J.P., Van Aarde, R.J., Erasmus, B.H., van
Rensburg, P.J.J, Skinner, J.D., Howell, P.G. and Naude, T.W. (2002).
‘A review of the successful eradication of feral cats from sub-Antarctic
Marion Island, Southern Indian Ocean’. South African Journal of
Wildlife Research 32(1): 65–73.
Blackburn, T.M., Cassey, P., Duncan, R.P., Evans, K.L. and Gaston, K.J.
(2004). ‘Avian extinction and mammalian introductions on oceanic
islands’. Science 305: 1955–1958.
Bloomer, J.P. and Bester, M.N. (1992). ‘Control of feral cats on subAntarctic Marion Island, Indian Ocean’. Biological Conservation 60:
211–219.
Bode, M., Brennan, K.E.C., Helmstedt, K., Desmond, A., Smia, R. and
Algar, D. (2013). ‘Interior fences reduce cost and uncertainty when
eradicating invasive species from islands’. Methods in Ecology and
Evolution 4(9): 819–827.
Bomford, M. and O’Brien, P. (1995). ‘Eradication or control for Vertebrate
Pests’? Wildlife Society Bulletin. 23(2): 249–255.
Bonnaud, E., Zarzosa-Lacoste, D., Bourgeois, K., Ruffino, L., Legrand, J.
and Vidal, E. (2010). ‘Top-predator control on islands boosts endemic
prey but not mesopredators’. Animal Conservation 13(6): 556–567.
Burbidge, A.A. Williams, M.R. and Abbott, I. (1997). ‘Mammals of
Australian islands: factors influencing species richness’. Journal of
Biogeography 24: 703–715.
Burbidge, A.A. (1999). ‘Conservation values and management of
Australian islands for non-volant mammal conservation’. Australian
Mammalogy 21: 67–74.
Burbidge, A. (2001). ‘Our largest island’. Landscope 17(2): 16–22.
Burbidge, A.A. and Manly, B.F.J. (2002). ‘Mammal extinctions on
Australian islands: causes and conservation implications’. Journal of
Biogeography 29: 465–473.
Bureau of Meteorology (2017). ‘Climate data online’. <http://www.bom.
gov.au/climate/cdo/about/cdo-selecting-data.shtml>. Accessed July
2017.
Campbell, K.J., Harper, G., Algar, D., Hanson, C.C., Keitt, B.S. and
Robinson, S. (2011). ‘Review of feral cat eradications on islands’.
In: C.R. Veitch, M.N. Clout and D.R. Towns (eds.) Island invasives:
eradication and management, pp. 37–46. Occassional Paper SSC no.
42. Gland, Switzerland: IUCN and Auckland, New Zealand: CBB.
Claussen, J. (2005). Native Plants of Christmas Island. Perth: Success
Print.
DE (Department of the Environment) (2015). Threat Abatement Plan for
Predation by Feral Cats. Commonwealth of Australia. 47 pp.
DEWHA (Department of the Environment, Water, Heritage and the Arts)
(2008). Threat Abatement Plan for Predation by Feral Cats. Canberra:
Department of the Environment, Water, Heritage and the Arts.
Dickman, C.R. (1992). ‘Conservation of Mammals in the Australasian
Region: the Importance of Islands’. In: J.N. Coles and J.M. Drew (eds.)
Australia and the Global Environmental Crisis, pp. 175–214. Canberra:
Academy Press.
Dickman, C.R. (1996). Overview of the Impact of Feral Cats on Australian
Native Fauna. Report to Australian Nature Conservation Agency.
EA (Environment Australia) (1999). Threat Abatement Plan for Predation
by Feral Cats. Environment Australia, Biodiversity Group, Canberra.
EA (Environment Australia) (2002). Third Christmas Island National
Park Management Plan. Environment Australia, Canberra.
Green, P. (2014). ‘Mammal extinction by introduced infectious disease
on Christmas Island (Indian Ocean): The historical context’. Australian
Zoologist 37: 1–14.
Johnston, M., Algar, D., Onus, M., Hamilton, N., Hilmer, S., Withnell, B.
and Koch, K. (2010). A Bait Efficacy Trial for the Management of Feral
Cats on Dirk Hartog Island. Arthur Rylah Institute for Environmental
Research Client Report. Heidelberg, Victoria: Department of
Sustainability and Environment.
Johnston, M., McCaldin, G. and Rieker, A. (2016). ‘Assessing the
availability of aerially delivered baits to feral cats through rainforest
canopy using unmanned aircraft’. Journal of Unmanned Vehicle
Systems 4(4): 276–281.
Keitt, B.S.; Wilcox, C.; Tershy, B.R.; Croll, D.A. and Donlan, C.J. (2002).
‘The effect of feral cats on the population viability of black-vented
shearwaters (Puffinus opisthomelas) on Natividad Island, Mexico’.
Animal Conservation 5: 217–223.
King, W.B. (1985). ‘Island Birds: Will the Future Repeat the Past’? In:
P.J. Moors (ed.) Conservation of Island Birds, pp. 3–15. International
Council for Bird Preservation Technical Publication No. 3. Cambridge,
England.
Martin, T.G., Nally, S., Burbidge, A.A., Arnall, S., Garnett, S.T.,
Hayward, M.W, Lumsden, L.F., Menkhorst, P., McDonald-Madden,
E. and Possingham, H.P. (2012). ‘Acting fast helps avoid extinction’.
Conservation Letters 5: 274–280.
Martinez-Gomez, J.E. and Jacobsen, J.K. (2004). ‘The conservation status
of Townsend’s shearwater Puffinus auricularis auricularis’. Biological
Conservation 116: 35–47.
McKenzie, N.L., Hall, N. and Muir, W.P. (2000). ‘Non-volant mammals
of the southern Carnarvon Basin, Western Australia’. Records of the
Western Australian Museum Supplement No. 6: 479–510.
Medina, F.M., Bonnaud, E., Vidal, E., Tershy, B.R., Zavaleta, E.S.,
Donlan, C.J., Keitt, B.S., Le Corre, M., Horwath, S.V. and Nogales,
M. (2011). ‘A global review of the impacts of invasive cats on island
endangered vertebrates’. Global Change Biology 17: 3503–3510.
Misso, M. and West, J. (2014). ‘Conservation management of the terrestrial
biodiversity of Christmas Island: challenges and perspectives’. Raffles
Bulletin of Zoology Supplement No. 30: 17–23.
Moors, P.J. and Atkinson, I.A.E. (1984). ‘Predation on Seabirds by
Introduced Animals, and Factors Affecting its Severity’. In: J.P. Croxall,
P.J.H. Evans and R.W. Schreiber (eds.) Status and Conservation of
the World’s Seabirds, pp. 667–690. International Council for Bird
Preservation Technical Publication No. 2. Cambridge, England.
Moro, D. (1997). ‘Removal of a feral cat from Serrurier Island’. Western
Australian Naturalist 21: 153–156.
Nogales, M., Martin, A., Tershy, B.R., Donlan, C.J., Veitch, D., Puerta,
N., Wood, B. and Alonso, J. (2004). ‘A review of feral domestic cat
eradication on islands’. Conservation Biology 18(2): 310–319.
Nogales, M., Vidal, E., Medina, F.M., Bonnaud, E., Tershy, B.R.,
Campbell, K.J. and Zavaleta, E.S. (2013). ‘Feral cats and biodiversity
conservation: the urgent prioritization of island management’.
BioScience 63(10): 804–810.
Orchard, M. (2012). Crabs of Christmas Island. Christmas Island Natural
History Association. Christmas Island, Indian Ocean, Australia.
Payne, A.; Curry, P. and Spencer, G. (1987). ‘An Inventory and Condition
Survey of Rangelands in the Carnarvon Basin, Western Australia’.
Technical Bulletin No.73. Perth, Australia: Western Australian
Department of Agriculture.
Pontier, D., Say, L., Debias, F., Bried, J., Thioulouse, J., Micol, T. and
Natoli, E. (2002). ‘The diet of cats (Felis catus L.) at five sites on the
Grande Terre, Kerguelen archipelago’. Polar Biology 25: 833–837.
Ratcliffe, N., Bell, M., Pelembe, T., Boyle, D., White, R.B.R., Godley,
B., Stevenson, J. and Sanders, S. (2009). ‘The eradication of feral cats
from Ascension Island and its subsequent recolonization by seabirds’.
Oryx 44(1): 20–29.
Robinson, S., Gadd, L., Johnston, M. and Pauza, M. (2015). ‘Long-term
protection of important seabird breeding colonies on Tasman Island
through eradication of cats’. New Zealand Journal of Ecology 39(2):
316–322.
Springer, K. (2016). ‘Methodology and challenges of a complex multispecies eradication in the sub-Antarctic and immediate effects of
invasive species removal’. New Zealand Journal of Ecology 40(2):
273–278.
Short, J., Turner, B., Risbey, D.A. and Carnamah, R. (1997). ‘Control
of feral cats for nature conservation. II. Population reduction by
poisoning’. Wildlife Research 24: 703–714.
Tidemann, C.R., Yorkston, H.D. and Russack, A.J. (1994). ‘The diet of
cats, Felis catus, on Christmas Island, Indian Ocean’. Wildlife Research
21: 279–286.
Van Aarde, R.J. (1980). ‘The diet and feeding behaviour of feral cats,
Felis catus, on Marion Island’. South African Journal of Wildlife
Research 10: 123–128.
Veitch, C.R. (1985). ‘Methods of eradicating feral cats from offshore
islands in New Zealand’’. In: P.J. Moors (ed.) Conservation of Island
Birds, pp. 125–141. International Council for Bird Preservation
Technical Publication No. 3. Cambridge, England.
243
M. Auld, B. Ayling, L. Bambini, G. Harper, G. Neville, S. Sankey, D.B.A. Thompson and P. Walton
Auld, M.; B. Ayling, L. Bambini, G. Harper, G. Neville, S. Sankey, D.B.A. Thompson and P. Walton.
Safeguarding Orkney’s native wildlife from non-native invasive stoats
Safeguarding Orkney’s native wildlife from non-native invasive stoats
M. Auld1, B. Ayling1, L. Bambini2, G. Harper3, G. Neville4, S. Sankey1, D.B.A. Thompson4 and P. Walton2
RSPB East Scotland Regional Office, 10 Albyn Terrace, Aberdeen AB10 1YP, UK. 2RSPB Scotland Headquarters,
Edinburgh Park, Edinburgh EH12 9DH, UK. 3Biodiversity Restoration Specialists, PO Box 65, Murchison 7053, New
Zealand. 4Scottish Natural Heritage, Great Glen House, Inverness IV3 8NW, UK. <graham.neville@nature.scot>.
1
Abstract The Orkney Islands, off the north-east coast of Scotland, support highly significant cultural and natural
heritage. The combined land area of the 70 islands is 990 km2 (380 sq mi), <1% of the UK, but they host over 20% of
the UK’s breeding hen harriers (Circus cyaneus) (declining over much of its mainland range), 8% of breeding curlews
(Numenius arquata) (one of only two UK populations not in decline) and an internationally important assemblage of
breeding seabirds. The Orkney Islands are naturally free of mammalian predators, and all bird species, including raptors,
are ground-nesting in the largely treeless landscape. Rats (Rattus spp.), hedgehogs (Erinaceus europaeus) and feral cats
(Felis catus) are present across the archipelago. Stoats (Mustela erminea) are native to mainland UK but not Orkney,
yet were detected on Orkney Mainland in 2010. Orkney Mainland has an area of 523 km2 (202 sq mi). Early attempts
at removing them were not successful. By 2013 stoats were present across the Orkney Mainland and connected isles. In
2016, SNH and RSPB formed a partnership to eradicate stoats to protect the native wildlife and designated sites of the
Orkney islands, and to secure the wider socio-economic and cultural benefits of thriving native wildlife. Difficulties faced
in developing the project include predicting the effort required to remove stoats at a rate faster than they can reproduce,
securing community support and access to private land and, in particular, funding large scale biodiversity restoration
projects. A feasibility study determined that stoat eradication would be possible using DOC200 kill traps, and search
dogs in later stages of the eradication. There are no legally available poisons that could be used on stoats in the UK. A
Biosecurity Plan has been produced for the archipelago, with a current focus on preventing the spread of stoats to the uninvaded isles. The partnership is working to secure funds and community support for what will be the world’s largest stoat
eradication attempted to date. We present the findings of the feasibility study and our proposed methodology.
Keywords: biosecurity, feasibility, hen harrier, Orkney vole, predator eradication, Scotland, short-eared owl
INTRODUCTION
The Orkney Islands are situated 10 km from the northeast coast of Caithness, at the northernmost point of
mainland Scotland. The archipelago is made up of around
70 islands, of which 20 are inhabited with a total population
of around 21,000 residents (National Records of Scotland,
2016). The largest island, Orkney Mainland, is some 523
km2 and is home to 75% of the human population. The
islands of Burray and South Ronaldsay are connected
to the Orkney Mainland by causeways carrying road
infrastructure. The remaining islands are not physically
linked, and are reachable via air or inter-island ferry.
Orkney supports a wide range of natural and cultural
heritage for which it is world-famous. There is abundant
native wildlife with seabirds, ground-nesting and wading
birds, corncrake and sea mammals – all important parts
of the ecosystem. Although the islands represent only
0.4% of the UK land area, they are home to a significant
proportion of native UK seabirds and terrestrial species.
About 14% of the UK breeding kittiwake (Rissa tridactyla)
population, 34% of arctic skua (Stercorarius parasiticus),
10% of puffin (Fratercula arctica), 25% of arctic tern
(Sterna paradisaea) and 14% of the global population
of breeding great skuas (Catharacta skua) are found in
Orkney. Orkney is a UK stronghold for hen harrier (Circus
cyaneus) and short-eared owl (Asio flammeus). Over 20%
of the UK population of hen harrier is known to breed in
Orkney. The islands are one of the few remaining core
areas for breeding corncrake (Crex crex) and a stronghold
for breeding waders and especially the curlew (Numenius
arquata) – the UK’s highest conservation priority bird
species. At a time of large-scale decline across the UK,
wader populations in the lowlands of Orkney are thought
to be stable or increasing, bucking the national and global
trends. Densities in these areas are amongst the highest
recorded in Europe.
The quality and importance of Orkney’s natural
heritage is recognised through the number of nationally
and internationally designated sites across the islands.
These cover approximately 30% of the islands’ land area.
There are 13 Special Protection Areas (SPAs), strictly
protected sites for rare and vulnerable birds, and for
regularly occurring migratory species, under the EU Birds
Directive, as well as five Special Areas for Conservation
(SACs) which offer strict protection for threatened habitats
under the EU Habitats Directive. Orkney also has 36 Sites
of Special Scientific Interest (SSSIs), a designation for
sites which best represent Scottish natural heritage and
are designated under the Nature Conservation (Scotland)
Act 2004; one National Scenic Area (NSA), a designation
representing Scotland’s finest landscapes; two nature
conservation Marine Protected Areas (NC MPAs), nature
conservation sites in the marine environment designated
under the Marine (Scotland) Act 2010; and three proposed
marine SPAs. The unique natural and historic heritage
of the islands underpins Orkney’s distinctive culture and
economy and supports a thriving tourism industry.
Although stoats (Mustela erminea) are native,
widespread and common throughout mainland Britain and
Ireland they are not native to the Orkney archipelago. The
first confirmed sighting of a stoat in Orkney was reported
in August 2010, following verbal reports of possible
sightings in June and July of the same year. It is not known
how stoats arrived in Orkney; possible vectors include
accidental release from imported hay or straw, shipping,
or deliberate release (e.g. to control rabbits). Sightings of
stoats reported to Scottish Natural Heritage (SNH) have
increased in frequency and stoats are now considered to be
present across the entire Mainland and linked isles.
This paper examines the risks to Orkney’s native wildlife
from the impact of predation by stoats and describes efforts
to date to deal with the problem. We present the findings
of the feasibility study into eradication, our proposed
methodology for eradication and outline some of the major
challenges to what will be the largest removal of stoats
anywhere in the world.
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
244
up to meet the challenge, pp. 244–248. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Auld, et al.: Invasive stoats on Orkney Islands
MATERIALS AND METHODS
This paper summarises the results and discussions
arising from stoat sightings across the islands, a desk
study conducted to predict the likely impact of stoats on
native wildlife, an independent technical feasibility study
of stoat eradication (Harper, 2017a), and a Biosecurity
Plan (Harper, 2017b) which identifies measures to prevent
increase in range and re-colonisation post-eradication.
RESULTS
Likely impact of stoats
When stoats are introduced into ecosystems that have no
native mammalian predators, such as those of the Orkney
islands, they can have a devastating impact on the native
species present. In New Zealand, the stoat is thought to be
the main driver of declines and some local extirpations of
many native bird populations (Dowding & Murphy, 2001).
A desk study that was conducted predicts that the ecological
consequences of stoat introduction to Orkney are likely to
be devastating (Fraser, et al., 2015). It is highly likely that
the presence of stoats on Orkney will have a catastrophic
effect on ground nesting birds and mammals on Orkney
due to the absence of other mammalian predators including
the red fox (Vulpes vulpes) “… it is highly likely that the
introduction of stoats will profoundly change the ecology
of Orkney and its value for birds of prey and the SPAs
that have been designated for protecting these species.”
(Fraser, et al., 2015). Stoats have never been part of the
ecosystem in Orkney and therefore many native species,
cannot respond rapidly enough to the introduction of this
predator. The potential scale and range of the impact of this
non-native predator is such that little wildlife in Orkney is
currently safe. Impact will be of national and international
significance due to the proportion of populations living on
the islands.
One critical linkage within the Orkney ecosystem is the
predator-prey relationship of the Orkney vole with the hen
harrier and short-eared owl. Fraser, et al. (2015) suggest that
a decline in Orkney voles will have direct consequences on
the hen harrier and short-eared owl populations because
both of these species rely to varying extents on the Orkney
vole as a component of their diet. The short-eared owl
has developed a specialist hunting behaviour to match
Orkney vole activity (Reynolds & Gorman, 1999). A range
of evidence suggests that the abundance of Orkney voles
(which do not display the cyclical population abundance
observed in other Microtus species) is directly linked to the
breeding success of the hen harrier and short-eared owl.
It is therefore suggested that significant depredation of
voles by introduced stoats will have an indirect detrimental
impact on these species. The RSPB Orkney reserves data
on numbers of hen harriers fledging suggests a sustained
decline which started in 2011 (RSPB 2016, unpublished
data). In relation to other species, Fraser, et al. (2015) raise
concerns over a range of ground-nesting birds including
curlew (another bird on the UK Red list of conservation
concern), as stoats are well known to be significant
predators, especially where other terrestrial mammalian
predators such as foxes are absent, as is the case in Orkney
– and so curlew and similar birds are now under severe
risk.
completed) Islands Visitor Survey (Visit Scotland, 2013)
shows that just over 142,800 people visited Orkney,
spending over £31 million in the local economy over the
period of a year. The main influence on visitors deciding
to come to Orkney was an interest in the archaeology and
history of Orkney, followed by the scenery and landscape.
Given the importance of wildlife tourism to the overall
tourism market in Orkney, the predicted declines of many
native species caused by stoat predation is a cause for
concern amongst tourism businesses.
Stoats are also predicted to affect free-range poultry
operations. Free-range poultry farming is common practice
in Orkney, where the absence of mammalian predators
makes this an economically viable management option
for poultry. If stoats continue to persist on the archipelago,
future impacts on this industry are expected to include loss
of stock to stoat predation as well as the financial impact
of implementing predator control and mitigation measures.
Population expansion
No population estimates are available for the stoat
population in Orkney, and the available information on
their range comes from sightings reported by members
of the public. Sightings (both those reported directly and
through the ‘Stoats in Orkney’ Facebook page, maintained
by interested local volunteers) have increased steadily
since 2010, with a marked increase in sightings since
2016. However, caution must be used when correlating
this to any indication of abundance of stoats. Press activity
and awareness-raising campaigns by SNH and RSPB
Scotland (including posters designed to increase recording
of sightings – see Fig. 2) may have increased peoples’
awareness of stoats. This could partially lie behind the
recent increases in the rate of sightings, although no
particular spikes in sightings have been recorded after
media activity in the past (Fraser, et al., 2015). The overall
distribution map of all sightings (Fig. 1) tends to reflect
where people live, work and travel, rather than any accurate
estimate of the distribution of stoats per se.
Stoats were first reported in two areas, one on Orkney
Mainland and one on South Ronaldsay in 2010. Since this
time their numbers and range have increased rapidly and
they are now known to be distributed across Mainland
Whilst there are other threats to Orkney’s native
wildlife, including climate change and changes in land
management practices, the stoat is considered to be the
most pressing and widespread current threat.
Decline in the native wildlife populations is predicted
to have a significant effect on socio-economic benefits
that Orkney’s nature and landscapes provide in terms of
tourism and farming. The 2012–2013 (the most recently
Fig. 1 Distribution of all sightings of stoats in Orkney as
reported to SNH (Scottish Natural Heritage) – 2010 to
2016.
245
Island invasives: scaling up to meet the challenge. Ch 2A Other taxa: Mammals
Fig. 2 Poster to raise awareness of stoats and their potential impact – and encourage
sightings to be reported to SNH (Scottish Natural Heritage).
Orkney and the linked isles of Burray and South Ronaldsay
(Fig. 1). Figure 3 shows the clear increase in sightings from
2010–2016. There are obvious seasonal peaks in stoat
sightings which would correlate with seasonal activity of
the animals (Fig. 4).
Although a number of live and lethal trapping efforts
were implemented, they were unsuccessful in completely
removing the species or controlling range and population
growth. When stoats were first confirmed in 2010 an early
‘rapid response’ was put in place with volunteers using
live traps to remove stoats from the two sites in South
Ronaldsay and the west Mainland. Any stoats caught were
relocated to mainland Scotland. This was most successful
in South Ronaldsay where sightings decreased to two
in 2011 and none at all in 2012 and 2013. However, it
clearly did not remove the problem on West Mainland
where sightings continued to be reported, and eventually
increased in numbers and over a wider area.
In 2011–2013, SNH employed a contractor to remove
stoats using kill traps mainly on the west Mainland,
246
across an area of some 300 km2 (115 sq mi). This was
unfortunately unsuccessful, possibly due to inexperience
of the contractor and/or an inadequate methodology. Fewer
than five stoats were caught – nowhere near the numbers
required to contain or remove the problem. In 2014, SNH
recruited a relatively large number of volunteers from the
local community following an awareness raising campaign
and increasing concern from interested parties. Over 50
volunteers were trained in the use of kill traps and a trapping
project was launched to remove stoats. The purpose of this
project was three-fold; in addition to attempting to remove
or control the spread of stoats, it was to test whether a
large scale volunteer effort was feasible and sustainable in
Orkney, and finally to trial approaches to data handling and
management for an eradication project.
The volunteering project was ultimately scaled down in
2016 as it had been shown that sustaining a large volunteer
force of this size was very resource intensive, few stoats
(under 10) were trapped and keeping volunteers motivated
when stoats were not being caught was very difficult.
Auld, et al.: Invasive stoats on Orkney Islands
landowner permission, and the inclusion of landowning
interests in the partnership, is of paramount importance.
Fig. 3 Number of stoat sightings reported each month on
Orkney between June 2010 and June 2017.
Feasibility of stoat eradication
A proposed methodology for eradicating stoats was
initially developed by SNH following their experience
gained during the Hebridean Mink Project (HMP) and the
Uist Wader Project (UWP) which carries out live trapping
and translocation of hedgehogs. It was revised following
advice from the project’s Technical Advisory Group (TAG).
This methodology and the technical, political and
environmental feasibility of stoat eradication was assessed
independently in an independent Technical Feasibility
Study commissioned by RSPB Scotland and completed by
Grant Harper of Biodiversity Restoration (Harper, 2017a)
following best practice guidelines.
The feasibility study determined that an eradication
project was feasible given the current range, but that a
new methodology should be adopted. Draft costings were
developed for this methodology to determine capacity to
eradicate. This methodology has now been assessed by the
TAG and adopted by the partnership.
The only legal method to remove stoats in the UK is
humane trapping. There are no approved viruses or poisons
available for use at present, nor are any likely to be approved
within the time frame of eradication. The feasibility study
identifies only NZ Department of Conservation (DOC) and
Goodnature A24 self-setting traps as AIHTS (Agreement
on the International Humane Trapping Standard) compliant
in the UK. In Scotland only DOC traps are currently legally
compliant. Dogs can be used to locate but not harm or kill
an animal.
The legal circumstances surrounding land access are
simplified as SNH has the power to issue a Species Control
Order which allows them to compulsorily access land in
order to control invasive species in the event that landowner
permission is withheld. This is of course a last resort, and
Environmental acceptability was assessed to determine
if the impact of the project on native wildlife can be reduced
to an acceptably low level and ensure eradication will not
lead to permanent negative changes impact on non-target
species. It is accepted that there is always a short-term risk
to non-target species, and the project is designed so that this
balances out to give a positive long-term change. Harper
shows that the species most at risk are rats (which are also
non-native to Orkney), but a smaller number of Orkney
vole, mice and (potentially) small feral cats are also at risk
(Harper, 2017a). However, based on New Zealand data,
none of these bycatches are predicted to be of big enough
volume to impact population size or stability of non-target
species or will be anywhere near the estimated impact of
stoats on their long-term populations (Harper, 2017a).
Setting and inspecting traps will inevitably create
disturbance in areas of high breeding density for native
wildlife. This risk will be minimised in the eradication
phase by sensitive placing and minimising disturbance
through remote monitoring of trap triggers. The effect of
this is also asserted to be much less than the effect of stoats
on the native wildlife.
To achieve eradication, a methodology must be
implemented that removes animals at a rate faster than
they can reproduce and target all of the animals within
the population. Stoats in the UK can have home range
sizes from 2–254 ha but average about 40 ha. Their home
range in Orkney is unknown but a home-range analysis
now, when the stoats still have room for expansion, would
delay the project unacceptably. A precautionary approach
is proposed that works on the assumption of the smallest
home range size. This is considered to be required due to
the year round food supply, density of food supply and
novelty of the predator.
A methodology is proposed that uses a uniform trapping
density of 16 traps per km2 in the first instance. Baited
DOC200 kill-traps in standard housing will be used and
it is proposed that Goodnature A24 self-resetting traps are
also deployed, to target trap-shy individuals that may avoid
the DOC traps if these traps are made legal in Scotland.
Dogs are also an option to locate trap shy individuals.
Within the currently affected area, all land is considered
easily accessible and well provisioned with access routes.
This density of trapping gives a total number of traps
of 9270, each will be roughly 250 m apart based on a
square grid. It is expected to take roughly two months to
set these traps, based on 10 trappers setting 20–30 traps
per day. Utilisation of habitats is likely to vary and it is
expected that, in the largely open farmland habitat, stoats
will use field margins that provide both cover and more
food. In these habitats, traps will be set according to linear
features including fences, walls, ditches, roads and tracks.
The proposed grid will be used as a guide for the placement
of the traps but trappers will have the discretion to move
the traps up to 50 m from the proposed position to the best
location on the ground for the interception of stoats.
Trap density and distribution in each habitat type will
be reviewed as the project develops using an adaptive
management approach.
Fig. 4 Seasonality of reported stoat sightings on Orkney.
Draft costings were developed to determine the financial
feasibility and the capacity of the relevant organisations to
deliver the project in a timely manner. These figures have
been further developed by the partnership to ensure social
feasibility through community support and involvement.
The cost of the project is around £4.5 million. This resource
is not available through government so external funding
will need to be sourced.
247
Island invasives: scaling up to meet the challenge. Ch 2A Other taxa: Mammals
The feasibility study did identify some aspects unique
to this project and in particular interaction with the manmade environment. To date stoat eradications have been
carried out in largely uninhabited areas so association
and behaviour of stoats with areas such as gardens, and
farmyards is not known. It is also worth noting that, whilst
effectiveness of traps, baits and lures is well documented in
New Zealand, this is less well known in the UK and there
have been no trials on Orkney.
of stoats on Orkney’s native wildlife through comparison
with impacts in other areas of the globe where they are
not native. This report clearly demonstrates an ecological
imperative to eradicate (Fraser, et al., 2015).
Biosecurity plan
An independent feasibility study (Harper, 2017a) has
clearly shown that it is possible to eradicate stoats given
the current range of presence across the Mainland and
connected isles (as shown in Fig. 1). An assessment of
sustainability has shown that we can sufficiently reduce
the risk of re-invasion but that we must act now to prevent
spread to other islands which could threaten feasibility of
eradication (Harper, 2017b).
Finally, the sustainability of the proposed project
has been assessed through an RSPB commissioned
independent biosecurity plan (Harper, 2017b) to determine
if the likelihood of reinvasion is suitably low, or the risks of
re-invasion can be reduced sufficiently, through affordable
biosecurity measures. All potential invasion and re-invasion
risks have been assessed. Two main pathways have been
identified for the re-introduction of stoats to Orkney, firstly
through intentional introduction and secondly accidentally
through haulage of fodder and bedding. A high level of
community engagement in an eradication and increased
vigilance is considered to reduce the risk of intentional
introduction and on-going use of sniffer dogs and an
isolation and trapping method for cargoes to sufficiently
reduce the risk of accidental re-introduction. It should be
noted that this risk is considered small. For decades Orkney
has had a high volume of bedding and fodder imported
each year, but only the 2010 incident of stoat introduction.
Measures can also be taken to reduce the source risk by
wrapping and timely movement of bales.
The Biosecurity plan does identify a major risk of
expansion to new islands. While no comprehensive survey
of stoat population on Orkney has been carried out, the
distribution and extent of stoat sightings across the whole
of the Mainland and linked isles suggests that the stoat
population has ended the invasion phase and it is suggested
(Harper, 2017b) that they are at or near carrying capacity
within the current range. There are over 60 islands that are
still thought to be stoat-free. This means that the situation
for stoats on Orkney is at a critical stage, as dispersal to
other islands is highly likely. Stoats are thought to be good
swimmers, and written accounts exist of stoats swimming
400 m in Ireland, and much further between islands in
New Zealand (Veale, 2013), although this has not yet
been observed on Orkney. There are many islands within
theoretical swimming distance for stoats which could act
as staging-posts for dispersing animals onto non-linked
isles which are currently biological refuges. Whilst the
plan clearly identifies eradication of stoats from the current
range as the most effective biosecurity measure to prevent
dispersal to new islands it is suggested that trapping on
the Orkney Mainland can reduce the risk of spread until
eradication can begin. There are five areas of Orkney
Mainland coast that have been identified where stoatfree islands are within swimming distance. Immediate
deployment of DOC traps in these areas will reduce risk
of dispersal. Whilst eradication is considered currently
feasible a successful eradication would already be three
times greater in area than the largest successful eradication
to date. Any extension of range, particularly to islands with
less accessible land could make an eradication no longer
technically or financially feasible.
DISCUSSION
Due to the native status of the stoat through most of
the UK, there is little direct evidence of impact of stoat
predation. However, due to the importance of the Orkney
islands in a national and international context for wildlife,
the SNH commissioned a report, assessing likely impacts
248
Our learning from early unsuccessful attempts to
remove and contain stoats through volunteer response and
small contracts has been put to good use in demonstrating
that a full scale professional eradication project is required
to deal with this issue.
Since this work has been completed the SNH and RSPB
Partnership have focused on developing a costed project,
applying for funding for both the eradication and ongoing
biosecurity measures, developing community support and
implementing immediate biosecurity measures to prevent
spread to more islands in Orkney.
We have developed a fully-costed plan, called the
Orkney Native Wildlife Project, costing £4.5 million. We
have applied to the Heritage Lottery Fund for support
and are also in the process of developing and submitting
two further grant applications for the eradication and
supporting ongoing biosecurity measures. RSPB Scotland
is about to commence biosecurity trapping measures on
the Orkney Mainland, in accordance with the biosecurity
plan. These measures will be kept in place until we start
eradication. We also have developed a trial trapping phase
within project development that will investigate success
of traps, lures and baits in different habitats and will fill
gaps in knowledge and be used to fine tune our trapping
methodology.
The Orkney Native Wildlife Project is unique. It will be
the first eradication of stoats in Europe, and also the first
project to consider stoat eradication in areas which include
urban and rural settlements.
REFERENCES
Dowding, J.E. and Murphy, E.C. (2001). ‘The impact of introduced
mammals on endemic shorebirds in New Zealand: a conservation
perspective’. Biological Conservation 99: 47–64.
Fraser, E.J., Lambin, X., McDonald, R.A. and Redpath, S.M. (2015).
‘Stoat (Mustela erminea) on the Orkney Islands – Assessing Risks to
Native Species’. Scottish Natural Heritage Commissioned Report No.
871.
Harper, G. (2017a). ‘Methodology Review and Assessment of Technical
Feasibility of Eradicating Stoats from the Orkney Islands’. RSPB
Scotland commissioned report.
Harper, G. (2017b). ‘Stoat Biosecurity Plan for the Orkney Islands’.
RSPB Scotland Commissioned Report.
National Records of Scotland (2016). ‘Orkney Islands Council Area
– Demographic Factsheet’. <https://www.nrscotland.gov.uk/files/
statistics/council-area-data-sheets/orkney-islands-factsheet.pdf>.
Accessed 15 February, 2018.
Reynolds, P. and Gorman, M.L. (1999). ‘The timing of hunting in shorteared owls (Asio flammeus) in relation to the activity patterns of Orkney
voles (Microtus arvalis orcadensis).’ Journal of Zoology 247: 371–379.
Veale, A.J. (2013). ‘Observations of stoats (Mustela erminea) swimming’.
New Zealand Journal of Zoology 40(2): 166–169.
Visit Scotland (2013). ‘Orkney Final Report (Islands Visitor Survey
2012–2013)’. Island Visitor Survey. <http://www.visitscotland.org/
research_and_statistics/visitor_research/all_markets/islands_visitor_
survey.aspx>. Accessed 2017.
C.C. Hanson, T.J. Hall, A.J. DeNicola, S. Silander, B.S. Keitt and K.J. Campbell
Hanson, C.C.; T.J. Hall, A.J. DeNicola, S. Silander, B.S. Keitt and K.J. Campbell. Rhesus macaque
eradication to restore the ecological integrity of Desecheo National Wildlife Refuge, Puerto Rico
Rhesus macaque eradication to restore the ecological integrity
of Desecheo National Wildlife Refuge, Puerto Rico
C.C. Hanson¹, T.J. Hall¹, A.J. DeNicola², S. Silander³, B.S. Keitt¹ and K.J. Campbell1,4
¹Island Conservation, 2100 Delaware Ave. Suite 1, Santa Cruz, California, 95060, USA. <chad.hanson@
islandconservation.org>. ²White Buffalo Inc., Connecticut, USA. ³U.S. Fish and Wildlife Service, Caribbean Islands›
NWR, P.O. Box 510 Boquerón, 00622, Puerto Rico. 4School of Geography, Planning & Environmental Management,
The University of Queensland, St Lucia 4072, Australia.
Abstract A non-native introduced population of rhesus macaques (Macaca mulatta) was targeted for removal from
Desecheo Island (117 ha), Puerto Rico. Macaques were introduced in 1966 and contributed to several plant and animal
extirpations. Since their release, three eradication campaigns were unsuccessful at removing the population; a fourth
campaign that addressed potential causes for previous failures was declared successful in 2017. Key attributes that led
to the success of this campaign included a robust partnership, adequate funding, and skilled field staff with a strong
eradication ethic that followed a plan based on eradication theory. Furthermore, the incorporation of modern technology
including strategic use of remote camera traps, monitoring of radio-collared Judas animals, night hunting with night vision
and thermal rifle scopes, and the use of high-power semi-automatic firearms made eradication feasible due to an increase
in the probability of detection and likelihood of removal. Precision shooting and trapping were the primary methods used
throughout the campaign. Long-term monitoring using camera traps and observed sign guided a management strategy that
adapted over time in response to population density and structure. Lessons learnt include, 1) macaques quickly adjusted
their behaviour in response to human presence and removal methods, 2) camera traps and thermal scopes provided high
detection likelihood compared to other methods, and 3) the use of Judas animals and night hunting with thermal and night
vision rifle-scopes facilitated removals. The removal of macaques from Desecheo Island appears to be the first introduced
non-hominid primate eradication from an island.
Keywords: conservation, invasive species, island restoration, Judas, Macaca mulatta, primate
INTRODUCTION
Islands occupy ∼5.5% of Earth’s terrestrial surface area
but contain more than 15% of terrestrial species (Kier, et
al., 2009), 61% of all recently extinct species, and 37% of
all critically endangered species on the International Union
for Conservation of Nature (IUCN) Red List (Tershy, et
al., 2015). Non-hominid primates (NHPs) are intelligent
and adaptable animals (Fooden, 2000). World-wide, 78
introduced insular populations are known on 63 islands
(Jones, et al., 2018). Despite their potential for ecological
impacts, including being implicated in 69 insular species
extinctions and extirpations globally (Jones, et al., 2018),
management is problematic as NHP’s demonstrate
behavioural traits making them challenging to remove
and few practitioners are experienced in their control or
eradication (Evans, 1989; Feild, et al., 1997; Breckon,
2000; Kemp & Burnett, 2003; Strier, 2016; Jones, et al.,
2018). Six eradication attempts have been documented
globally and all were unsuccessful (Jones, et al., 2018).
Desecheo Island (Desecheo), has been the site of half of
these attempts targeting a population of invasive rhesus
macaques (Macaca mulatta).
Historically, Desecheo was a major seabird rookery. In
the early 1900s tens of thousands of seabirds representing
seven species were nesting on the island (Bowdish, 1900;
Wetmore, 1918; Struthers, 1927; Meier, et al., 1989;
Noble & Meier, 1989). The most numerous species,
brown boobies (Sula leucogaster), numbered 8,000 15,000 individuals (Danforth, 1931 cited by Noble &
Meier, 1989; Wetmore, 1918) with red-footed boobies (S.
sula), brown noddies (Anous stolidus), and bridled terns
(Sterna anaethetus) accounting for another 12–14,000
birds. Humans shooting birds and harvesting eggs, habitat
destruction through farming, ranching and military
munitions training, and introduced feral goats (Capra
hircus) and black rats (Rattus rattus) reduced populations
of most seabird species and restricted many species to less
accessible areas of the island (Wetmore, 1918; Struthers,
1927; Evans, 1989; Meier, et al., 1989). Feral goats were
recently eradicated (2009; Hanson, unpublished data)
while black rats were eradicated in 2016 after an initial
attempt failed in 2012 (Will, et al., 2019). However,
predation by rhesus macaques (macaques), introduced in
1966 for research purposes, resulted in the complete loss
of seabird breeding on the island and was considered the
most significant threat to wildlife on Desecheo (Evans,
1989; Meier, et al., 1989; Noble & Meier, 1989). In 1969,
massive raids by macaques on booby nests were reported,
with macaques pushing boobies off their nests and
consuming an estimated 200–300 eggs per week (Noble
& Meier, 1989). In 1987, although nests were built and
eggs laid, brown and red-footed booby nesting success
was zero (Noble & Meier, 1989). Macaques contributed
to the extirpation of at least five seabird species, one
land bird species, and led to the depauperate state of
resident land birds on Desecheo (Noble & Meier, 1989;
Island Conservation, 2007). Macaques on Desecheo have
also been implicated in modifying vegetation structure,
contributing to the extirpation of several plant species, and
preying on native reptiles including three island-endemic
lizards (Evans, 1989; Breckon, 2000; Island Conservation,
2007).
In 1976, Desecheo was designated a National Wildlife
Refuge and the island was transferred from the Department
of Health, Education, and Welfare to the U.S. Fish and
Wildlife Service (USFWS). At this time the removal of
macaques was identified as an objective to restore the
island’s ecological integrity (Island Conservation, 2007).
Between 1976 and 1988, three eradication attempts took
place with a total of 155 animals removed (Herbert, 1987;
Evans, 1989; Breckon, 2000; USFWS, 2007). An initial
attempt was reported to have insufficient funding to proceed
(USFWS, 2007). The second eradication attempt required
multiple removal methods to target wary individuals.
After 246 days of effort it was assumed all individuals had
been removed, but less than a year later 15 individuals
were confirmed on the island (Evans, 1989). The third
eradication attempt ended prematurely in 1988 due to
a lack of resources; it was believed at that time that two
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 249–255. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
249
Island invasives: scaling up to meet the challenge. Ch 2A Other taxa: Mammals
slopes, which face the prevailing winds. The woodlands
are typically found covering coastal slopes, upper eastand south-facing slopes, along drainages, and within
valley floors. The floral community of Desecheo is dry
forest habitat. The island is composed primarily of Tertiary
volcanic sandstones and rises to 218 m. Steep slopes fall
away from five ridges interconnected by a perpendicular
ridge which rises abruptly from the northeast coast (Fig.
1). There is no permanent surface water or spring on the
island.
METHODS
Macaques carry B-virus (Cercopithecine herpesvirus
1), which can be lethal to humans (Huff & Barry, 2003),
so animal handling was minimised where possible. The
Desecheo macaque population originated from a population
with a high occurrence of the disease (Shah & Morrison,
1969) and most likely had B-virus. When animals were
handled, strict protocols were followed (Holmes, et al.,
1995).
Fig. 1 Aerial images of Desecheo National Wildlife Refuge.
males and an unidentified juvenile were all that remained
(USFWS, 2007). However, Breckon (2000) reported 11
animals in a single troop in 1996. Lack of funding, animals
becoming educated to removal techniques, and unreliable
detection methods contributed to the lack of eradication
success. In April 2007, Island Conservation in partnership
with USFWS developed a restoration plan (Island
Conservation, 2007) that outlined a strategy and methods
to eradicate macaques from the island. The planning effort
coincided with the development of an environmental
assessment covering the removal of non-hominid primates
from the Commonwealth of Puerto Rico and its offshore
islands (USDA, et al., 2008), including Desecheo. Here
we report on the 2008–2017 eradication of macaques from
Desecheo National Wildlife Refuge.
STUDY SITE
Desecheo is a small (117 ha) uninhabited hilly island
(18° 23’ N, 67° 29’ W) situated roughly 21 km off the west
coast of Puerto Rico. The vegetation is a mosaic of grassy
patches, shrublands, woodlands, and semi-deciduous forest.
The grassy patches and shrublands are on exposed ridges
and screes, especially on the northern and north-eastern
Several principles were employed to increase the
likelihood of success: 1) target whole groups where
possible, 2) limit opportunities to educate animals, 3)
first utilise methods that would not impact the efficacy
of other methods, 4) have sufficient methods to remove
animals faster than the rate of reproduction, and 5) provide
multiple detection methods that were independent of
removal techniques, capable of detecting animals at very
low densities. Variations of live-trapping and hunting were
selected after a suite of possible techniques, including
the use of toxicants, biological control, kill trapping, and
immunocontraception, were evaluated for use on Desecheo
(Island Conservation, 2007). The strategy to remove
macaques was structured around three general phases and
was adaptively managed from 2008 to 2017 (Fig. 2). The
initial phase relied on live-trapping to provide a population
reduction without educating animals to subsequent hunting
methods. Select individuals captured were radio collared
then released and tracked as sentinel (Judas) animals to
facilitate hunting of a social species. The second phase
aimed to remove remaining individuals through hunting
and transitioned to a third phase where monitoring was
anticipated to confirm eradication. A revised approach
was required when macaques could only be detected by
remote cameras. This involved specialised night hunting
technology paired with the use of Judas animals and
a distinct change in hunting strategy which primarily
occurred outside of daylight hours.
Fig. 2 Project timeline of events. (*) field work initiated by seasoning traps on site.
250
Hanson, et al.: Macaques eradicated from Desecheo Island
Phases relied on a team temporarily camped at one of
two sites on the island with all equipment and supplies
being delivered then removed each trip. The first campsite,
serviced by helicopter, was located near the peak of the
island. This site supported up to nine staff, was utilised from
project initiation through the duration of group hunting (see
Fig. 3) and allowed centralised access to the entire island.
A second site was later established along the coastline to
allow boat access and minimise logistic expenses for a
reduced field team to complete the eradication.
Phase I. Trapping and Judas animal release (2008–
2009)
Eighteen #208 dual-door cage traps (Tomahawk Live
Trap, Hazelhurst, WI) were placed in groups of three
across the island at sites of known macaque activity. Trap
dimensions of 107 × 38 × 38 cm were considered large
enough to capture multiple animals, based on mainland
Puerto Rico trapping efforts (López Ortiz, 2015).
Concurrently, a single, large 5 m wide group-style trap
(Day, 2004) was built upon a flat, ridgetop location. This
trap was constructed in an octagon shape with a wood
frame, cyclone fence sides and skirt. A 60 cm overhanging
eave and 60 cm vertical wall made from sheet metal faced
internally to prevent animals from exiting. A remotecontrolled drop-net was used in another site, comprising
an 11 × 11 m reinforced net elevated above the ground by
roughly 2 m around the edge with a tented peak of 5 m. Prebaiting took place across all trap sites for two weeks with
whole and sliced oranges. Oranges were chosen based on
successful results experienced during previous eradication
campaigns (Evans, 1989). Prior to departing the island, all
cage traps and a single side of the octagon trap were wired
open for animals to become familiar with their presence.
Seven months later all traps were activated and a network
of 48 padded leg-hold traps was installed along areas
suspected to have macaque activity. Traps were typically
set in groups of two or more and each were accompanied
by a magnetically triggered trap-monitor. Monitors were in
place to support near-real-time monitoring of each trap’s
status which was communicated by radio-transmitter to a
R-1000 telemetry receiver (Communications Specialist,
Orange, CA); traps were monitored several times daily. A
second pre-baiting effort took place during this time. To
supplement oranges and provide greater variety, additional
bait types including mangos, chicken eggs, and a water
drip pan were utilised and replaced regularly. A remotecontrolled audio lure (FoxPro Crossfire, Lewistown, PA)
programmed with macaque calls also was deployed in
association with baits at the drop-net location. Various
leg-hold traps were set with lures including mirrors, wind
chimes, streamers, feathers, or brightly coloured objects
suspended above the trap site. Traps that did not receive a
lure were set as a blind set with no distinguishing features
separating it from the original site.
Fig. 3 Number of macaques removed over time in relation
to project phase.
During this timeframe, a wild-caught adult male
macaque from mainland Puerto Rico was quarantined,
sterilised by vasectomy, radio-collared (Telenax, TXE311C, Playa del Carmen, Mexico), and transported to
Desecheo. This individual was released upon arrival and
monitored daily as a Judas animal.
Phase II. Hunting (2009–2013)
Trapping activities from Phase I overlapped with this
phase for one field trip. Hunting was intended to remove
remnant individuals that were avoiding trap sets. Timing
of this phase was based on the seasonally deciduous
dominant tree species (Bursera simaruba) which leafedout in response to rainfall. Field staff were selected from
within Island Conservation and from White Buffalo
Inc. (Connecticut, USA) based on their experience in
precision shooting and demonstration of eradication ethic.
In preparation, key vantage points were identified while
conducting a census of the population before any removals
took place. This assessment effort also was used to identify
concealed shooting hides that offered a wide field of
view for observation and clear shooting lanes. Hunting
was considered capable of placing all individuals at risk
of removal, particularly once population numbers were
reduced with a successful trapping phase.
Troop removal (2009–2010)
Hunting predominately relied on an ambush-then-stalk
strategy that collected troop characteristics (number of
individuals, body size of individuals) and movements at
dusk while macaques located a location to roost. In certain
circumstances, where specific trees were identified as a
roost site, field staff would proceed with hunting in the
middle of the night while utilising spotlights and closerange shooting. In most circumstances, staff would wait
until nightfall before returning to camp to develop a strategy
of engagement for the following day. Before first light,
field staff would be dispatched to pre-established hides or
to new locations thought to offer a better vantage point of a
troop’s roost location. Field staff were equipped with highcapacity centrefire semi-automatic .223 Remington or 6.5
Grendel rifles with telescopic sights ranging from 4.5× to
20× magnification and reticles matched to each firearm’s
ballistics. Other field staff were stationed along known
escape routes with high capacity 12-gauge semi-automatic
shotguns.
Shooters would communicate via 2-way radio to assess
the troop and attempt to identify the number of individuals,
their hierarchy, and body size. Body size class was estimated
based on body mass and ranked as one through five.
Groups would only be engaged if it was considered a high
likelihood that all individuals could be removed. Field staff
that had a visual on the dominant individual would engage
with the first shot, with other staff following by removing
individuals that presented a lethal shot opportunity. Adult
females (often dominant) were removed first, followed
by adult males and juveniles. Field staff would continue
to monitor the site while supporting shooters would be
redistributed to areas where escapes were thought to have
possibly occurred. Once macaque activity ceased in the
canopy, field staff equipped with close-range firearms
would enter the site to remove any remnant individuals.
Removals were tallied and the animals’ body size classes
would be recorded. Follow-up visual confirmation of
carcasses occurred whenever possible. Any known escapes
were recorded, along with their size class. Confirmed
removals and escapes would be cross-referenced with the
troop size estimate. To improve the detection of roosting
troops during this phase a commercial-grade handheld
thermal camera (FLIR, P620, Wilsonville, Oregon, USA)
was trialled.
251
Island invasives: scaling up to meet the challenge. Ch 2A Other taxa: Mammals
Remnant removal (2011–2013)
Phase III. Monitoring (2012–2017)
After the initial knock-down of the macaque population,
the project shifted focus to the detection and removal of lone
individuals and reconstituted groups created after troops
were fractured. Before dawn, field staff were stationed
across the island to conduct focused observations over as
much landscape as possible. Visual observation of canopy
movement and audible cracks of tree limbs and masticated
seeds were the primary cues of macaque presence prior to
direct observation. If a detection was made, the number
of animals was estimated and, if confidence existed that
the group or individual could be removed, field staff
would proceed by removing individuals through shooting.
When assistance was required, additional field staff would
be guided to the site offering the highest likelihood of
removing the entire group. If escapes were thought to be
probable, the team would reassess the opportunity and hold
off until another situation presented greater confidence in
removal.
Monitoring occurred simultaneously with the removal
of remnants and night hunting in Phase II. The presence
of macaques was assessed through active and passive
monitoring techniques; each independent of removal
methods. Active monitoring occurred through visual
observation of animals and the detection of fresh sign.
Passive monitoring trialled acoustic recording units
(Wildlife Acoustics, Maynard, MA), but relied primarily
on a network of 16–26 Hyperfire PC900 no-glow cameras
(Reconyx, Holmen WI). Cameras were placed in locations
known to have had previous macaque activity or at sites
which offered a clear field of view across a travel route.
Specific attention was given to rocky bluffs, exposed patches
of slope, or within tree canopies that were dominated by
horizontal tree branches. Tuning the camera field-of-view
used an integrated “walk-test” function which indicated
where the camera would be triggered by movement.
Lures made from cord passed through brightly coloured
balls were installed at sites which could accommodate a
swinging item without triggering the camera’s motion
sensor. Cameras were serviced every 3–9 months, where
memory cards (32GB, 95mb/s write speed) were switched
for empty ones and batteries were replaced if below 60%
charge. Camera operational settings were programmed to
operate from one hour prior to dawn to one hour after dusk,
provide the highest sensor sensitivity, take five photos in
succession, and reset immediately after a trigger event.
Phase II revision. Night hunting (2013–2015)
Remote cameras (see monitoring) continued to detect
macaques that were undetectable to field staff. Methods
employed were re-analysed, leading to detection dogs
and night hunting technology being considered. Dogs that
could effectively track animals traveling on the ground and
through forest canopy were considered necessary and a
breed of mountain cur that is used to pursue squirrels was
identified. Additionally, managers of NHPs on mainland
Puerto Rico had sourced effective thermal hunting optics
and began demonstrating success with night hunting.
In 2013, three macaques were selected from mainland
Puerto Rico to be used as Judas animals to support night
hunting. Young female macaques were chosen as they were
considered more likely to readily associate with remnant
animals on Desecheo. Replicating methods developed for
Judas goats, each macaque was sterilised via tubal ligation,
fitted with a radio-telemetry collar (ATS, M2950B, Isanti,
Minnesota, USA), and received a Compudose® 200 (25.7
mg estradiol; Elanco, Indianapolis, USA) implant to
induce prolonged oestrus (Zehr, et al., 1998; Campbell,
et al., 2005; Campbell, et al., 2007). Radio telemetry
collars had infrared (IR) reflective patches sewn in and
a solar powered light-emitting diode (LED) epoxied to
them to facilitate detection at night. Judas macaques were
transported to Desecheo via boat and released.
Hunting methodology changed to working strictly at
night, initially incorporating mainland Puerto Rico staff
and their equipment to train the project team. Based on the
success of these methods, thermal weapons scopes with a
built in adjustable reticle (BAE Systems, Inc. ATS-6000M,
Arlington, Virginia, USA), a 3rd generation night vision
clip-on unit (Knight Optics Ltd., Krystal 950, Harrietsham,
Kent, UK) used in combination with pre-existing telescopic
firearms optics, and an IR laser illuminator (Jager-Pro
LLC., JP-IR Laser, Fortson, Georgia, USA) were procured
to improve detection and facilitate removals. Night
operations continued with 2–3 field staff using the thermal
scope to detect heat signatures of macaques in conjunction
with telemetry scans for Judas animals (described in Phase
III). When no Judas animals were present in a group all were
targeted. When Judas animals were present night vision in
conjunction with infrared illuminators were used to detect
IR reflective patches sewn into collars to determine which
macaque in the group was the Judas, facilitating removal of
only uncollared macaques.
252
RESULTS
A total of 140 macaques were removed from Desecheo
Island between 2009 and 2015, excluding Judas animals
translocated from mainland Puerto Rico (Fig. 3). The cost
of the 2007–2017 campaign was US$ 1.229m. The majority
of costs (73%) were associated with implementation and
monitoring from 2009 to 2015 at US$ 893k. Planning and
preparation in 2007/8 utilised US$ 214k and US$ 121k
was spent on confirmation over 2015–2017.
Phase I. Trapping
Baiting to encourage macaques into traps was
ineffective. Additional lures such as a water drip and audio
lures also proved unsuccessful as evidenced by camera
traps. Non-target species, primarily black rats and hermit
crabs, would consume any bait not suspended from the
ground. Bait that did persist required regular replacement
due to the arid climate on the island; fruits quickly
desiccated and non-boiled eggs rapidly spoiled. After 26
days, unsuccessful traps that were located in remote sites
and not easily accessible by field staff were closed to
ration bait and improve the efficiency of trap monitoring.
Traps left open were outfitted with florescent flagging as a
visual lure and left open. A total of 546 trap nights accrued
between cage traps, the group-style octagon trap, and the
drop net with zero trap success.
Padded leg-hold traps were in place for 1,344 trap
nights and resulted in the capture of 13 macaques; 10.7%
of the population. Three received radio collars and were
released as Judas animals; all but one was sterilised. Traps
equipped with novel visual lures, particularly reflective
materials, demonstrated a higher catch rate than nonreflective items. Two blind sets established as a part of a
three-trap grouping, demonstrated success simultaneously.
Traps placed at the base of trees where macaques would
leap into the tree were particularly successful.
Hanson, et al.: Macaques eradicated from Desecheo Island
The adult male Judas animal transported to Desecheo
from mainland Puerto Rico was found dead 16 days after
its release for unknown reasons.
Phase II. Hunting
Hunting reduced the population of macaques to near
undetectable levels by removing 118 individuals (84%
of 140 macaques removed) over the span of two trips
(46 days where hunting took place) across two years.
Estimates of animals remaining at the end of each trip
significantly underestimated the population. Three animals
were known to be present at the end of the second trip,
one of which was a sterilised Judas macaque. Follow-up
hunting focused on the detection and removal of remnant
macaques, which removed four individuals in five field
trips (66 days where hunting took place) over three years.
At this time, camera monitoring indicated six individuals
remained and evidence of population recruitment, shown
by one newborn juvenile.
Phase II. Revision
The introduction of night hunting strategies supported
by thermal and night vision technology, field staff’s
intimate knowledge of the island’s terrain, and leveraging
Judas animal behaviour resulted in five macaque removals
over five trips (50 days where hunting took place) over 2.5
years.
Phase III. Monitoring
The single most effective monitoring tool proved to be
the remote camera network. Camera density ranged from
one camera per 4.5–7.25 ha. Roughly 450,000 images
were collected throughout the entire campaign. The
volume of images varied greatly depending on the length
of a monitoring period (2–9 months) with a mean of ~50K
images. More than 2K macaque detections were compiled.
Camera placement was impacted by vegetation growth
over time leading to the majority of images being false
captures. Once population numbers were reduced to five
individuals the entire group could be tracked with at least
one detection of each individual occurring per monitoring
period. Judas animals, with unique physical and collar
characteristics, could be identified within camera images
and were used to indicate camera network efficacy; all
Judas animals were detected per monitoring session and
were easily distinguishable from wild individuals due to
the presence of the collar.
Fig. 4 Number of removals in relation to effort expended
over time.
formed an independent pair. One Judas animal was later
removed to disrupt the social balance which resulted in a
more consistent interaction between the remaining Judas
animal and the two known remnant animals. Field staff
removed one remnant animal before the final Judas animal
was found dead in 2016. This Judas animal was associating
with the last wild macaque known to remain on island.
Twelve hunting trips were conducted totalling 5,280
detection hours (Fig. 4). A single wild adult matching
the description of the last known wild macaque was
detected on 15 occasions by the camera trap network over
approximately 41 months indicating that this was the only
wild individual that remained. Over the same timeframe, no
juveniles were shot or detected, reflecting no reproduction.
As a result, the project was closed in June 2017 with the
understanding that the population was functionally extinct.
DISCUSSION AND LESSONS LEARNT
The campaign to remove macaques from Desecheo
took 10 years, 17 field trips, variations of two primary
methods – trapping and hunting – and a network of remote
monitoring cameras to complete.
Acoustic recording units were trialled but did not
result in confident detections by monitoring macaque
vocalisation. This tool was quickly discounted as an
effective option to monitor animals at low density. The
lack of vocalisation was corroborated by field staff who
indicated macaques no longer vocalised with the same
frequency once the population was reduced to less than ten
individuals.
Pre-baiting attempts were unsuccessful, resulting
in the ineffectiveness of baited traps. Trapping efforts
may have benefited from trials and a longer pre-baiting
period which also would take into account timing to
allow learnt behaviours to transfer through the population.
Locally available food items including nuts and berries
were considered but discounted as they were found in
abundance across Desecheo. Having a diet with limited
exposure to novel food items on Desecheo is thought to
have contributed to their disinterest in baits provided. Once
baited trapping ceased, hunting and leg-hold trapping were
then relied upon as the sole methods. Trapping efforts on
mainland Puerto Rico that utilise a variety of fruits have
resulted in up to 50% of project removals (López Ortiz,
2015) suggesting that trap success is variable across sites
and should remain a management consideration.
Additional monitoring took place though the tracking
and assessment of Judas animals. Of the three animals
captured on-island, one unsterilised male was found dead
due to unknown causes, a second sterilised male was
inadvertently shot during the hunting phase of the project,
and a third sterilised female experienced a collar failure
and integrated back into the population. This individual
was one of the last macaques removed. Of the three
additional Judas animals later captured on mainland Puerto
Rico and released in Phase III, none experienced collar
failure although the installed LED lights did not function in
the field. After release, one was indistinguishable amongst
a group of wild macaques and shot while night hunting
with thermal optics. The remaining two Judas macaques
When hunting was initiated, only troops where all
individuals were thought to be at risk of removal were
targeted. Although this method proved to be effective and
efficient, it became apparent that escapes likely occurred
unbeknownst to field staff as macaques quickly adjusted
their behaviour in response to human presence and removal
methods. Macaques increasingly avoided detection during
the day, and if field staff were detected, would regularly
select a route of escape that placed an object between them
and the observer, limiting shot opportunities, as they fled to
an adjacent watershed. This behaviour eventually nullified
daylight hunting and required revised methods and tactics
to improve the probability that an individual would be
detected.
253
Island invasives: scaling up to meet the challenge. Ch 2A Other taxa: Mammals
Advanced night hunting equipment facilitated both
detections and removals. In many instances macaques
were detectable only with a thermal scope, even when
field staff knew the location of the animal. As a result,
an integrated shooting reticle with the ability to remove
and return the scope to the firearm without having any
shift in the scope’s point of aim was considered critical.
Furthermore, having the ability to de-couple the night
vision from traditional hunting scopes proved valuable as
the firearm’s point of aim did not need to be recalibrated
for daylight hunting; only the thermal scope could be used
in daylight without damaging the equipment or losing the
capability to continue hunting into dawn with the firearm
paired with night vision. A less sophisticated general-use
FLIR thermal camera was trialled early in the project
although low image resolution limited the unit’s detection
range. At ranges beyond ~150 m, individual pixels were
estimated to be larger than a macaque’s heat signature. The
camera’s limited range resulted in zero detections and thus
general-use thermal tools were abandoned.
If the project had been initiated with advanced night
hunting thermal equipment and Judas animals it is
estimated that its duration and cost would have been
significantly reduced. Hunting activities could have
taken place regardless of seasonal variation in vegetation,
detections would have been more frequent, and entire
groups could have been removed with greater confidence,
precision, and frequency. Furthermore, the incorporation of
suppressed firearms with subsonic ammunition would have
offered additional advantages. Suppressed firearms would
likely have reduced the flight response of any associated
macaques due to abated firearm report, projectile “crack,”
and identification of shot origin.
Camera traps provided high detection likelihood as
compared to other passive detection methods, particularly
once the density of animals was reduced to near-zero.
Camera placement, and the decision to increase the size
of the network, was guided by weeks of observation
before and after removals took place and is believed to
have significantly improved detection probability. Staff
were familiar with the use of the same cameras with feral
cats (Felis catus), however, a greater awareness of the
camera’s field-of-view and trigger window was necessary
when setting cameras to monitor a three-dimensional
environment. The presence of an accurate walk-test function
offered confidence that cameras were set to detect animals
at varying elevations and distances. In addition, a robust
camera design offered confidence that cameras would have
a low failure rate regardless of adverse field conditions
including hurricanes, intense heat, and sustained humidity.
Failures generally included screen malfunctions, walk-test
function ceasing to work, and component corrosion due to
termites burrowing into the camera case.
Ongoing commitment from the partners throughout a
dynamic project enabled the eradication to succeed. There
was a significant investment up front to start the project
and the initial projected methodologies and associated
tools did not result in eradication. As a result, the project
lasted longer than expected and overall costs were higher
than anticipated. These costs may have been reduced if
the funding was available in larger amounts rather than in
annual allocations, allowing higher intensity effort over a
shorter period of time. However, it is also believed that
the long periods between hunting trips was beneficial
because macaques became less agitated, resumed routine
behaviours, and were more likely to be detected.
254
Macaques becoming educated to removal techniques,
unreliable detection methods, and a lack of funding were
linked to previous failures on Desecheo. To reduce the
probability that similar issues would impact the success
of this attempt, the partnership routinely and transparently
reassessed all aspects of the project including the funding
required to proceed, equipment and field trips considered
necessary to achieve eradication, and how to interpret
results. These factors guided an adaptive management
strategy that supported principles outlined within original
project planning. This shared effort resulted in a robust
relationship that was capable of addressing a dynamic
project and uncertainty in a solution-oriented, step-bystep manner. As a result, a project with no precedent of
success – incentivised by a high conservation reward – was
completed in a conscious and calculated fashion.
CONCLUSION
Desecheo Island is the location of the first successful
removal of introduced non-hominid primates from an
island that we are aware of. The project was contingent
on the strength of the partnership, specialised equipment,
and commitment of an experienced field team with a strong
eradication ethic that followed a plan based on eradication
theory. These factors were all critical to the project’s
success after a protracted time-period. The challenges of
this eradication required several revisions to the original
methodologies and strategies, as well as continued funding
beyond the original budget projections.
Desecheo Island, and the unique species that are found
there are now safe from invasive mammals after nearly a
century. This restoration action should enable the island’s
return as the most important seabird colony within the
region.
ACKNOWLEDGEMENTS
This work was made possible by the generous support
of our philanthropic donors and funding from the United
States Fish and Wildlife Service. We also would like to
thank W. Wolfrom and B. Mancial, J. Padilla, F. Boyd, L.
Figueroa, M. Evans, J. Herbert, J. Schwagerl, E. Melendez,
A. Saunders, B. Tershy, J. Bonham, C. Bergman, L.
Bennett, W. Jolley, D. Will, E. Oberg, J.L. Herrera-Giraldo,
K. Swinnerton, C. Figuerola-Hernandez, R. Rodts, T.
Robinson, W. Shockley, M. Garcia, O. Acevedo, and many
other USFWS, DNER, Puerto Rico law enforcement, and
IC staff for assisting with planning, field work or logistical
support. Mayaguez Zoo and Puerto Rico DNER provided
support with sourcing and preparing Judas macaques.
USDA Wildlife Services and DNER staff assisted with two
field trips and graciously shared their skills and knowledge
acquired from their own night fieldwork.
REFERENCES
Bowdish, H.S. (1900). ‘A day on De Cicheo Island’. The Oölogist 17:
117–120.
Breckon, G.J. (2000). ‘Revision of the flora of Desecheo Island, Puerto
Rico’. Caribbean Journal of Science 36: 177–209.
Campbell, K.J., Baxter, G.S., Murray, P.J., Coblentz, B.E., Donlan, C.J.
and Carrion, V.G. (2005). ‘Increasing the efficacy of Judas goats by
sterilisation and pregnancy termination’. Wildlife Research 32: 737–
743.
Campbell, K.J., Baxter, G.S., Murray, P.J., Coblentz, B.E. and Donlan,
C.J. (2007). ‘Development of a prolonged estrus effect for use in Judas
goats’. Applied Animal Behaviour Science 102: 12–23.
Hanson, et al.: Macaques eradicated from Desecheo Island
Day, T.D. (2004). Feasibility of Pest Proof Fencing in Mauritius.
Cambridge, New Zealand: Xcluder Pest Proof Fencing Company.
Evans, M.A. (1989). ‘Ecology and removal of introduced rhesus monkeys:
Desecheo Island National Wildlife Refuge, Puerto Rico’. Puerto Rico
Health Science Journal 8: 139–156.
Feild, J.G., Henke, S.E. and McCoy, J.G. (1997). ‘Depredation on
Artificial Ground Nests by Japanese Macaques: The Unspoken Exotic in
Texas’. In: C.D. Lee and S.E. Hygnstrom (eds.) Thirteenth Great Plains
Wildlife Damage Control Workshop Proceedings, pp. 151–155. Kansas
State University Agricultural Experiment Station and Cooperative
Extension Service.
Will, D.J., Swinnerton, K., Silander, S., Keitt, B., Griffiths, R., Howald,
G.R., Figuerola-Hernandez, C.E. and Herrera-Giraldo, J.L. (2019).
‘Applying lessons learnt from tropical rodent eradications: a second
attempt to remove invasive rats from Desecheo National Wildlife
Refuge, Puerto Rico’. In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C.
Russell and C.J. West (eds.) Island invasives: scaling up to meet
the challenge, pp. 154–161. Occasional Paper SSC no. 62. Gland,
Switzerland: IUCN.
Zehr, J.L., Maestripieri, D. and Wallen, K. (1998). ‘Estradiol increases
female sexual initiation independent of male responsiveness in rhesus
monkeys’. Hormones and Behavior 33: 95–103.
Fooden, J. (2000). Systematic Review of the Rhesus Macaque, Macaca
mulatta (Zimmermann, 1780), Chicago: Field Museum of Natural
History.
Herbert, H.J. (1987). Unpublished Final Report on Desecheo Island
Monkey Removal Program, to Sean Furniss, Caribbean Islands
National Wildlife Refuge. Boqueron, Puerto Rico.
Holmes, G.P., Chapman, L.E., Stewart, J.A., Straus, S.E., Hilliard,
J.K. and Davenport, D.S. (1995). ‘Guidelines for the prevention and
treatment of B-virus infections in exposed persons. The B virus working
group’. Clinical Infectious Diseases 20: 421-439.
Huff, J.L. and Barry, P.A. (2003). ‘B-virus (Cercopithecine herpesvirus
1) infection in humans and macaques: Potential for zoonotic disease’.
Emerging Infectious Diseases 9: 246–250.
Island Conservation (2007). Restoring Desecheo Island National Wildlife
Refuge: Macaque Eradication Plan, December 2007. Santa Cruz:
Island Conservation.
Jones, H.P., Campbell, K.J., Burke, A., Baxter, G.S., Hanson, C.C. and
Mittermeier, R.A. (2018). ‘Introduced non-hominid primates impact
biodiversity and livelihoods: management priorities’. Biological
Invasions: <https://doi.org/10.1007/s10530-018-1704-5>.
Kemp, N.J. and Burnett, J.B. (2003). Final Report: A Biodiversity Risk
Assessment and Recommendations for Risk Management of LongTailed Macaques (Macaca fascicularis) in New Guinea. Washington,
DC: Indo-Pacific Conservation Alliance.
Kier, G., Kreft, H., Lee, T.M., Jetz, W., Ibisch, P.L., Nowicki, C., Mutke,
J. and Barthlott, W. (2009). ‘A global assessment of endemism and
species richness across island and mainland regions’. Proceedings of
the National Academy of Sciences 106: 9322–9327.
López Ortiz, R. (2015). T-10 Project for the Control of Wild Primates
in Southwestern Puerto Rico. Puerto Rico Department of Natural and
Environmental Resources. Terrestrial Resources Division. <http://
Drna.Pr.Gov/Programas-Y-Proyectos/Proyecto-Para-El-Control-DePrimates-En-El-Suroeste-De-Puerto-Rico/>. Accessed 26 June, 2017.
Meier, A.J., Noble, R.E. and Raffaele, H.A. (1989). ‘The birds of
Desecheo Island, Puerto Rico, including a new record for Puerto Rican
territory’. Caribbean Journal of Science 25: 24–29.
Noble, R.E. and Meier, A.J. (1989). Status of Boobies, Sula sula and Sula
leucogaster, on Desecheo Island, Puerto Rico, Unpublished report to
the U.S. Fish and Wildlife Service, Contract No. 14-16-009-1561.
Shah, K.V. and Morrison, J.A. (1969). ‘Comparison of three rhesus groups
for antibody patterns to some viruses: Absence of active simian virus 40
transmission in the free-ranging rhesus of Cayo Santiago’. American
Journal of Epidemiology 89: 308–315.
Strier, K.B. (2016). Primate behavioral ecology. London, England:
Routledge.
Struthers, P.H. (1927). ‘Notes on the bird-life of Mona and Desecheo
Islands’. Auk 44: 539–544.
Tershy, B.R., Shen, K.-W., Newton, K.M., Holmes, N.D. and Croll, D.A.
(2015). ‘The importance of islands for the protection of biological and
linguistic diversity’. BioScience 65: 592–597.
USDA, APHIS and WS (2008). Environmental Assessment (PreDecisional). Managing Damage and Threats Associated with Invasive
Patas and Rhesus Monkeys in the Commonwealth of Puerto Rico.
Washington, DC: US Department of Agriculture, Animal & Plant
Health Inspection Service, Wildlife Services.
USFWS (2007). Summary of Desecheo Trip Reports for Macaque
Eradication. Compiled by Joseph Schwagerl. Unpublished.
Wetmore, A. (1918). ‘The birds of Desecheo Island, Porto Rico’. Auk 35:
333–340.
255
N. Macdonald, G. Nugent, K-A. Edge and J.P. Parkes
Macdonald, N.; G. Nugent, K-A. Edge and J.P. Parkes. Eradication of red deer from Secretary Island, New Zealand: changing tactics to achieve success
Eradication of red deer from Secretary Island, New Zealand:
changing tactics to achieve success
N. Macdonald1, G. Nugent2, K-A. Edge3 and J.P. Parkes4
Department of Conservation, PO Box 743, Invercargill 9840 <nmacdonald@doc.govt.nz>. 2Landcare Research,
PO Box 69040, Lincoln 7640, NZ. 3Edge-Effect, 48 Bligh St., Te Anau, NZ. 4Kurahaupo Consulting, 2 Ashdale Lane,
Christchurch, NZ.
1
Abstract Red deer (Cervus elaphus) established on 8,140 ha Secretary Island after swimming from the mainland in the
early 1960s. Attempts to remove them began in the 1970s and after several starts and stops they were eradicated in late
2014. Since late 2006, 688 deer have been removed. Ground hunters killed 365 deer in 1,827 hunter-days, 320 deer were
shot from helicopters in 211 flying-hours, two deer were trapped and one was known to have been killed by a fisherman.
The campaign since 2006 was planned in three phases – an initial population reduction, a mop-up phase and a surveillance
and rapid response to any new immigration phase. An initial reduction of 80% of the population, between 530 and 550 in
2006, was planned and achieved in the first two years. The removal of surviving deer was planned to take a further four
years but despite 114 being shot and probably less than 14 deer remaining in 2013 eradication was not achieved using the
methods that succeeded in the initial phase. The change in tactics in 2014 that allowed for eradication was to (a) ground
survey the island and use camera traps to locate areas with deer, (b) identify individual deer from faecal DNA to estimate
numbers, know when they were shot or still alive, and to estimate potential new immigration from the mainland – which
was low, and (c) move from individual hunters seeking any deer within a widespread population, when about 10% of
hunter-deer encounters led to a kill, to re-train hunters as teams using GPS/radio systems and integrate them with aerial
hunting to seek individual deer at known locations, when 100% of encounters led to a kill. The change of tactics that led
to eradication success required about half the costs, i.e. $25,000 to $10,500 per deer direct operational costs, expected if
no change had been made.
Keywords: aerial hunting, catch-per-unit-effort, density, faecal DNA profiles, ground hunting, operational costs, World
Heritage Area
INTRODUCTION
There are few published examples of successful deer
eradication campaigns in the world. This is mostly because
deer are generally valued as resources rather than as pests
but, in New Zealand, red deer are an introduced species
so there is interest in completely removing deer from
some places in order to protect the native biota (Parkes
& Murphy, 2003). Here we document a prolonged but
ultimately successful campaign to remove deer from a
large island in south-western New Zealand.
Secretary Island covers 8,140 ha and rises to 1,196 m
a.s.l. at 45°14' S 166°55' E in Fiordland National Park,
part of Te Wahi Pounamu South-west New Zealand World
Heritage Area (Fig. 1). Red deer (Cervus elaphus) swam
to Secretary Island from the mainland in the early 1960s
(Mark & Baylis, 1975; Crouchley, et al., 2007) across a sea
gap of at least 630 m. A population established and their
impact on the pristine native forests was severe and rapid
(Mark & Baylis, 1975; Mark, et al., 1991) so in the 1970s,
New Zealand Forest Service attempted, unsuccessfully,
to remove the deer (Tustin, 1977). However, in the early
2000s, the New Zealand Department of Conservation
(DOC) initiated a new campaign (Brown, 2005; Crouchley,
et al., 2007) that began in earnest in late 2006. This second
eradication attempt was itself reassessed by DOC once
the population had been reduced to very low numbers
(estimated at 14 individuals) in 2012/13, resulting in
changes in strategy and operational tactics that eventually
led to successful eradication of the deer. In this paper,
we briefly reiterate the results presented in early reports
and in the second Island Invasives conference for the
first eradication attempt, and update the results from the
initial reduction phase (Crouchley, et al., 2011; Edge, et
al., 2011). We then focus on the new data to report on the
change in strategy and tactics to remove the last few deer
from the island and compare the predictions of a catchper-unit-effort (CPUE) model produced in 2012 (Nugent
& Arienti-Latham, 2012) with the actual outcomes of the
deer control during the final phase of the project.
MAIN FINDINGS
First eradication attempt: 1970–1989
Ground and aerial hunting began in the early 1970s
and although 250 deer were reported as killed by the New
Zealand Forest Service between 1970 and 1985 (Brown,
2005) the population, in the presence of abundant food
(Mark & Baylis, 1975), continued to increase. Tustin
(1977) guessed about 200 deer were present in 1975. A
poisoning technique (1080 gel smeared on the leaves of
deer-preferred plants; see Parkes, 1983) was trialled from
1975 to1987 (when 10% of the island was poisoned)
but informal track and pellet counts suggested efficacy
was moderate at best (Brown, 2005). The abundance of
preferred food species and a perception that the difficult
terrain on Secretary Island restricted ground access (later
disproved when hunters covered the whole island to survey
for surviving deer) were likely reasons this trial did not lead
to eradication of the deer. In contrast, in an area on Stewart
Island, where white-tailed deer (Odocoileus virgineanus)
had removed most palatable food plants and accessibility
to people was not difficult, the 1080-gel technique removed
close to 100% of the population of deer in the treated area
(Nugent, 1990). The best control methods depend on
context, showing that successful precedent does not supply
a recipe for new projects.
By the early 1980s it was concluded that neither
hunting nor the 1080-gel method could remove all deer, so
the policy shifted in 1985 to one of sustained control to low
residual densities (Sanson & von Tunzelman, 1985). By
1989, official deer control on the island was halted because
of budget constraints and the expectation that reinvasion
would always compromise the project (W. Chisholm, 1989,
unpubl. DOC Invercargill file ANI 4/6). Deer were still
shot on Secretary Island by commercial venison recovery
helicopter operators. However, the goals of restoring the
island’s ecosystems by controlling deer and stoats were
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
256
up to meet the challenge, pp. 256–260. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Macdonald, et al.: Red deer off Secretary Island, New Zealand
huts across the island, so each hunter covered different
areas in each hunting period (usually about nine days) with
hunters often swapping areas between hunting periods (see
Crouchley, et al. (2011) for a detailed description of the
hunting and other methods). Aerial and ground hunting
began in November 2006.
The ‘rapid knockdown’ aim was effectively met
as 84% of the total deer killed were shot within three
years (by the end of 2009). We estimated that hunters
operating individually killed only about 10% of deer they
‘encountered’, i.e. seen, heard or known to be in the area
being hunted from fresh sign. There was little motivation
to persist with hunting a particular deer that escaped when
there were plenty of other deer in the area being hunted.
However, the aim to eradicate the population by the end of
2012 was not met as deer were still present. In retrospect,
98% of the final tally had been killed by then, but not the
100% required for eradication.
Final push: January 2014–August 2014
Failure to eradicate by 2012 (Fig. 2) led to a hiatus in
activity while the strategy and tactics being used for the
‘mop-up’ phase were reconsidered. The surviving deer
were extremely wary and could detect and escape hunters
(with dogs) operating as individuals and were avoiding
the open grasslands where they would be most vulnerable
to aerial shooting. The Department of Conservation had
no novel control tools to add to the mix it had already
used so decided that they had to apply ground and aerial
hunting in a different way to counter these learnt avoidance
behaviours of the deer. A decision was made to shift
from individual hunting to team hunting informed by all
available information. To some extent this was informed
by the experience of the new project manager (the senior
author) who with a private company (Prohunt Ltd, now
Native Range Ltd) had recently achieved eradication of
feral pigs (Sus scrofa) from Santa Cruz Island (Parkes, et
al., 2010). Technological advances available in the final
phases of the Secretary Island project included the use of
hand-held GPS and radios that allowed immediate contact
and location details to be shared between hunters, high
definition remote trail cameras, and the ability to identify
individual deer from DNA in faecal pellets.
Fig. 1 Secretary Island, Fiordland National Park, New
Zealand (photo L. Wilson).
not forgotten (Munn, 2001) and in the early 2000s a new
eradication project was proposed (Brown, 2005) with a
formal operational plan developed in 2007 (Crouchley, et
al., 2007).
The first step, in February 2013, under the revised
strategy, was to use hunters with indicator dogs to search
the whole island from ridge tops to the sea along transects
about 200 m apart for sign of deer. Analysis of the DNA in
the mucus layer of fresh (i.e. moist with unbroken exterior
estimated to be only a few day’s old) faecal pellets (see
Ramón-Laca, et al., 2014 for details of the methods; such
Second eradication attempt: November 2006–April
2013
A new decision to attempt eradication of red deer (and
also stoats (Mustela erminea)) from Secretary Island and
nearby Resolution Island (21,000 ha) was proposed in
2004 and a budget of NZ$7.1 million was allocated (Edge,
et al., 2011). This second attempt adopted a more strategic
approach, aiming to reduce the population by 80% within
two years, then remove survivors within four years, and
subsequently detect and remove any new immigrants in
perpetuity (Crouchley, et al., 2011). It was expected that
the initial knockdown would rely on two main methods
(ground hunting with indicator dogs and helicopter
shooting) but that a variety of ‘niche’ control methods
(17 capture pens, fences, the use of telemetered deer)
would probably be required during the ‘mop-up’ phase
(Crouchley, et al., 2011). The ground hunting involved
hunters (and their dogs) operating individually from nine
Fig. 2 Monthly kills of red deer on Secretary Island between
the start of the second eradication campaign in late 2006
and the last deer killed in August 2014.
257
Island invasives: scaling up to meet the challenge. Ch 2A Other taxa: Mammals
analyses currently cost about NZ$90 per sample depending
on sample size) found during this survey allowed individual
deer to be identified and the area in which they lived to
be located. The whole-island sign survey suggested
that possibly 14 deer remained at the end of April 2013
(Macdonald, 2013). The second step, in late 2013, was
to select and train the hunters in the skills (and attitudes)
required for a team hunting and targeting individual deer.
The logic of this change in hunting method depended
on (a) identifying from ground surveys for sign roughly
where a targeted survivor was living, (b) using a helicopter
to place a team of hunters at key exit points around that
location, (c) then deploying the best hunter-and-dog teams
in the suspected range of the deer to attempt to find and kill
it, (d) and, where that failed and the deer also avoided the
perimeter hunters, to then use the dog to track the deer and
the helicopter to either relocate the perimeter ambushers or
to shoot the deer if it became visible.
As hunting under a new strategy proceeded and the
DNA taken from shot animals was compared with DNA
found in an ongoing collection of faecal pellets it was
estimated that only eight deer remained by the end of
2013. Deployment of 13 trail cameras at key sites around
the island combined with ongoing DNA sampling did not
identify any new ‘unknown’ deer at this stage of the mopup. All deer shot after the island-wide survey in 2013/14
were (apart from two fawns shot with their mothers)
identified with the DNA faecal pellet database, and all but
one had an image captured by a trail camera.
Nine deer were shot in in 2014 under the new strategy.
Seven were adults (3F, 4M) and two were fawns. Three
deer were shot by ground hunters, two from helicopters,
and four from helicopters after the deer had been flushed
out of the forest by ground hunters and their dogs. The last
known animals were shot during August 2014 – a pregnant
female which was flushed out of the forest by hunters and
their dogs and shot from a helicopter, and an adult male
shot by the ground hunters.
Between November 2006 and August 2014, a total of
688 deer were killed, of which at least between 530 and 550
would have been alive at the start of the eradication project
in late 2006: an MNA 2006 density of 6.7 deer/km2. The
actual number was probably slightly higher as our estimate
is based on known deaths and does not include animals that
may have been wounded and died, died naturally, or were
shot by other hunters and not reported.
Costs
Assuming direct operational costs of NZ$950 per
flying-hour and $330 per hunter-day (the hunters were
contracted for set periods but paid whether they actually
hunted on a particular day or not) and using a population
reconstruction model with a starting population size of 530
animals and an annual recruitment rate of 24%, the cost
per deer shot increased rapidly as deer density declined
for both aerial and ground hunting methods (Fig. 3). The
cumulative 2006–2014 direct operational costs totalled
$732,830 plus unknown management overheads that are
likely to be roughly similar across years as they are less
related to hunting effort.
We fitted a negative power function to the cost per
deer versus density data from 2006 to 2012 (Fig. 3).
Extrapolation from that curve suggested that expenditure
of > $200,000 in direct costs would be required to remove
the estimated residual population of eight deer within one
year if there were no change in tactics. However, with
the change in tactics in 2014, the actual direct costs were
only about $84,000, indicating that the change in tactics
was not only successful but much more cost-efficient. This
of course ignores the significant factor of good luck (or
bad luck from the deer’s point of view) at the end of such
Initial population size
The careful collection of hunting statistics – numbers
of deer killed, their age and sex and hunting effort – allows
us to construct models of the population size and structure
at any point during the project since 2006. The ages of 78
females shot on the island in 2006/07 and classed as adults
by the hunters were determined from tooth cementum layers
(Fraser & Sweetapple, 1993). All other animals were aged
into three classes (young of the year, yearling, and older)
by the hunters in the field. The population size in 2006 can
be estimated using a form of the ‘minimum number known
to be alive’ (MNA) analysis of McCullough, et al. (1990).
Simply, the age of each animal shot was used to determine
if it was alive in 2006 and the pre-fawning MNA population
size in December 2006 (fawns are assumed all born at this
time of year) is all animals shot after December 2006 that
had been born before December 2006, plus all deer killed
in 2007 other than fawns born in December 2006, plus all
deer shot in 2008 other than fawns born in 2008 and subadults born in 2007, and so on. After 2009 an unknown
number of deer in the oldest age class may have been born
after 2006. To subtract these from our estimate of the initial
population we used the age-class distribution of the 78 deer
accurately aged and assumed the proportions remained the
same across the post-2009 deer that were killed. Given
84% of the estimate of initial population size accumulates
in the first three years, the potential errors in using this age
distribution for older deer born after 2006 are minor. We
assumed all deer were accurately aged, particularly when
allocated an age class in the field, there was no immigration
from the mainland and hunting by the official hunters was
the major cause of mortality.
258
Fig. 3 Direct costs (hunter-days and flying hours) per deer
killed with decreasing density, for aerial hunting, ground
hunting and overall, for (a) the data from the first five
years, and (b) for the whole campaign. The power curves
in (b) are extrapolations from the 2007–2011 data and
show that costs in the final stages (after adoption of new
tactics) were lower than predicted from the initial data.
Macdonald, et al.: Red deer off Secretary Island, New Zealand
eradication projects, e.g. see the last pig from Santa Cruz
Island which was shot incidentally to another task (Parkes,
et al., 2010).
DISCUSSION
Eradication projects that rely on a succession of control
events to eventually remove the population have one
advantage over single-event projects, such as aerial baiting
for rodents, in that information on progress and problems
accrues as data are collected from each event. This allows
managers to change tactics as the population is reduced
and especially when surviving animals are less accessible
or have learnt to avoid the control methods deployed at
the start of the project. Previous successful and efficient
eradication projects of this type have developed some
practices (e.g. Ramsey, et al., 2009; Parkes, et al., 2010)
that were, in part, used in the Secretary Island project.
The first success factor in such projects is that they
reduce the population to very low densities as quickly
as possible using control techniques that maximize the
probability that every animal is killed at first encounter
and thus minimize the possibility that surviving animals
learn to avoid all control methods. It might be argued that
live trapping in capture pens or 1080-gel on natural bait
poisoning does not make surviving deer more wary, at
least to subsequent hunting if not to the danger of traps,
and should be used first. However, trapping is capital
and labour intensive, unlikely to achieve rapid reduction
in deer populations, while the earlier attempts at natural
bait poisoning in Secretary Island were thought to be
unsuccessful in achieving a large reduction. This left
aerial and ground-based hunting as the only practical
tools to achieve the initial population reduction, but which
inevitably do not kill all deer at first encounter and so leave
wary survivors. It is unknown whether the same successful
initial reduction could have been achieved, and without
creating wary survivors, by starting with the approach
(team hunting with the additional improved GPS/radio and
DNA technologies, and closer integration between ground
and aerial hunting) deployed in the mop-up phase after
2013.
Many of the estimates of the number of deer left
at various points across the campaign were essentially
informed guesses. However, three tools were used to
improve confidence in estimates of the number and identity
of deer surviving on the island – a model based on catch
per unit effort data, camera traps and the use of DNA
from faecal pellets and aged and sexed shot individuals to
determine presence of un-shot deer (pellets present for an
individual not yet shot) and familial relationships (younger
animal shot but not yet its parents) and potentially whether
the DNA is from a resident survivor or an immigrant from
the South Island.
The DNA from the deer shot during the campaign
suggested they were all closely related (Crouchley, et al.,
2011). This precluded trying to use the DNA in young
animals (which were easier to shoot than adults) as a
marker to see if their parents are eventually shot (e.g.
see Nugent, et al., 2005). However, this is good news as
the island deer had few of the rarer alleles present on the
mainland. This suggests that the initial immigration in
the 1960s had not been repeated, probably because deer
populations throughout Fiordland were greatly reduced by
commercial aerial hunting after that time (Nugent, et al.,
1987). Therefore, the extirpation of the resident population
on Secretary Island might indeed be eradication sensu
stricto – still, a precautionary approach of surveillance and
rapid response to any new incursions is intended.
Some general observations to ensure surviving deer did
not escape are:
(a) to deploy hunters at optimal times/weather rather than
on a set schedule,
(b) to know the general areas on the island where the
surviving deer are living by extensive ground
searches and use of camera traps,
(c) to know which individual deer have escaped the
hunters by comparing DNA in faeces with DNA in
animals shot and,
(d) to change the mindset of the hunters from ‘control’ to
‘eradication’, i.e. from acting as individuals, however
skilled, each hunting any deer in their hunting block,
to team hunters with appropriate technologies to act
as a team and target individual deer.
The success on Secretary Island, and other smaller
islands in Fiordland National Park, provides some templates
for the proposed projects against red deer on similar islands.
Eradication of red deer has been attempted on Resolution
Island (21,000 ha), which is also in Fiordland National Park
(Edge, et al., 2011). This project has not succeeded and is
currently being reviewed (N. Macdonald, pers. comm.).
The Government of Argentina is also considering whether
to attempt to eradicate red deer and feral goats (Capra
hircus) from Isla los Estados (Staten) Island (53,400 ha)
in Tierra del Fuego – another remote, mountainous island
dominated by southern beech forests (A. Schiavini, pers.
comm.). New technologies to locate cryptic deer are also
becoming available with improvements in infrared systems
(FLIR) currently being deployed against black-tailed deer
(Odocoileus hemionus) that have survived an eradication
attempt on 1,637 ha Ramsay Island in British Columbia
(N. Macdonald, unpubl. data).
The general strategy used on Secretary Island, of an
initial rapid reduction in the deer population followed by
removal of survivors, succeeded in its aim of eradication.
However, in retrospect there is always going to be a
difficult decision for managers when deciding when to
deploy different control tactics across such a campaign. An
ideal approach would be to begin with control methods that
do not teach surviving animals to avoid later control, and
then to apply control methods in a way that minimises the
chance of animals escaping each encounter. On Secretary
Island, and potentially for other deer eradication projects,
we suggest that the team hunting system and coordination
between ground and aerial hunting may have been better
applied from the start of the 2006 hunting campaign rather
than towards the end of the eradication.
ACKNOWLEDGEMENTS
We thank the many hunters without whose skills
such projects could not succeed. The DNA profiling was
conducted by Ecogene of Landcare Research, Cecilia
Arienti-Latham contributed to development of the
population model in 2012, and the whole project was
funded by the New Zealand Department of Conservation.
We also thank Dave Forsyth and Dick Veitch for useful
comments on drafts of this paper.
REFERENCES
Brown, D. (2005). Secretary Island Deer Eradication Scoping Document.
Invercargill: Department of Conservation.
Crouchley, D., Brown, D., Edge, K-A. and McMurtrie, P. (2007). Secretary
Island Operational Plan: Deer Eradication. Te Anau: Department of
Conservation.
259
Island invasives: scaling up to meet the challenge. Ch 2A Other taxa: Mammals
Crouchley, D., Nugent, G. and Edge, K-A. (2011). ‘Removal of red deer
(Cervus elaphus) from Anchor and Secretary Islands, Fiordland, New
Zealand’. In: C.R. Veitch; M.N. Clout and D.R. Towns (eds.). Island
invasives: eradication and management, pp. 422–425. Occassional
Paper SSC no. 42. Gland, Switzerland: IUCN and Auckland, New
Zealand: CBB.
Edge, K-A., Crouchley, D., McMurtrie, P., Willans, M.J. and Byrom, A.
(2011). ‘Eradicating stoats (Mustela erminea) and red deer (Cervus
elaphus) off islands in Fiordland’. In: C.R. Veitch; M.N. Clout and
D.R. Towns (eds.) Island invasives: eradication and management, pp.
166–171. Occassional Paper SSC no. 42. Gland, Switzerland: IUCN
and Auckland, New Zealand: CBB.
Fraser, K.W. and Sweetapple, P.J. (1993). Assessing the Age and
Condition of Deer from Jawbones. Christchurch, New Zealand:
Landcare Research.
McCullough, D.R., Pine, D.S., Whitmore, D.L., Mansfield, T.M. and
Decker, R.H. (1990). ‘Sex linked harvest strategy for big game
management with a test case on black-tailed deer’. Wildlife Monographs
11: 1–41.
Macdonald, N. (2013). Removal of Red Deer from Secretary Island,
Fiordland National Park 2013–2014. Invercargill: Department of
Conservation. DOCDM-1294828.
Mark, A.F. and Baylis, G.T.S. (1975). ‘Impact of deer on Secretary Island,
Fiordland, New Zealand’. Proceedings of the New Zealand Ecological
Society 22: 19–24.
Mark, A.F., Baylis, G.T.S. and Dickinson, K.J.M. (1991). ‘Monitoring the
impact of deer on vegetation condition of Secretary Island, Fiordland
National Park, New Zealand: A clear case for deer control and ecological
restoration’. Journal of the Royal Society of New Zealand 21: 43–54.
Munn, A. (2001). Secretary Island. A Consideration of the Ecological
Values of the Island and the Potential for Restoration. Invercargill:
Department of Conservation.
Nugent, G.; Parkes, J.P. and Tustin, K.G. (1987). ‘Changes in the density
and distribution of red deer and wapiti in northern Fiordland’. New
Zealand Journal of Ecology 10: 11–21.
Nugent, G. (1990). A White-tailed Deer Poisoning Trial on Stewart
Island. Christchurch, New Zealand: Forest Research Institute Contract
Report FEW 90/4.
Nugent, G.; Whitford, J. and McEwan, M. (2005). ‘DNA-based Markrecapture of Wild Deer: Using Fawns to Capture their Mothers’. In:
J. Parkes, M. Statham and G. Edwards (eds.) Proceedings of the 13th
Australasian Vertebrate Pest Conference, pp. 91–95. Lincoln, New
Zealand: Landcare Research.
Nugent, G. and Arienti-Latham, C. (2012). Simple Models of the Secretary
Island Deer Eradication Programme. Lincoln, New Zealand: Landcare
Research Contract Report LC1057.
Parkes, J.P. (1983). ‘Control of feral goats by poisoning with compound
1080 on natural vegetation baits and by shooting’. New Zealand Journal
of Forestry Science 13: 266–274.
Parkes, J.P.; Ramsey, D.S.L.; Macdonald, N.; Walker, K.; McKnight,
S.; Cohen, B.S. and Morrison, S.A. (2010). ‘Rapid eradication of
feral pigs (Sus scrofa) from Santa Cruz Island, California’. Biological
Conservation 143: 634–641.
Parkes, J. and Murphy, E. (2003). ‘Management of introduced mammals
in New Zealand’. New Zealand Journal of Zoology 30: 335–359.
Ramón-Laca, A.; Gleeson, D.; Yockney, I.; Perry, M.; Nugent, G. and
Forsyth, D.M. (2014). ‘Reliable discrimination of 10 ungulate species
using high resolution melting analysis of faecal DNA’. PLOS ONE 9(3):
e92043.
Ramsey, D.S.L.; Parkes, J.P. and Morrison, S.A. (2009). ‘Quantifying
eradication success: The removal of feral pigs from Santa Cruz Island,
California’. Conservation Biology 23: 449–459.
Sanson, L.V. and von Tunzelman, J.R. (1985). Wild Animal Control
Plan for Secretary Island, Fiordland National Park. Invercargill: New
Zealand Forest Service.
Tustin, K.G. (1977). ‘Secretary Island – a problem in deer control’.
What’s New in Forest Research No. 50. Rotorua, New Zealand: Forest
Research Institute.
260
I.A. Macleod, D. Maclennan, R. Raynor, D.B.A. Thompson and S. Whitaker
Macleod, I.A.; D. Maclennan, R. Raynor, D.B.A. Thompson and S. Whitaker. Large scale eradication of
non-native invasive American mink (Neovison vison) from the Outer Hebrides of Scotland
Large scale eradication of non-native invasive American mink
(Neovison vison) from the Outer Hebrides of Scotland
I.A. Macleod, D. Maclennan, R. Raynor, D.B.A. Thompson and S. Whitaker
Scottish Natural Heritage, Great Glen House, Leachkin Road, Inverness, IV3 8NW, UK. <iain.macleod@nature.scot>.
Abstract The Hebridean Mink Project was tasked with eradicating American mink (Neovison vison) from the Outer
Hebrides, an extensive, complex island archipelago, amounting to 3,050 km2. Hundreds of islands contribute to a coastline
of approximately 2,500 km, 15% of Scotland’s total. The geographical complexity continues inland with over 7,500
freshwater lochs, ~24% of Scotland’s total, which enables invasive American mink, in suitable habitats, to reach densities
seldom encountered elsewhere. With major funding from the EU LIFE programme, removal from the Uists began in
2001. By 2006 eradication was declared there, as no captures had occurred for 16 months. In 2007 the project extended
into Harris and Lewis, adopting a systematic network of live capture traps (7,039 spaced at 450–500 m intervals utilising
prominent features of the riparian network and coastline). The traps were checked in rotation until at least a 95% reduction
in population had been achieved. An incremental, strategic change from systematic trapping to detection; by means of
footprint monitoring, cameras and dog searching, followed by responsive trapping then occurred from 2011 onwards. By
2013 a lethal monitoring system utilising ‘kill traps’ was employed alongside remote alert systems which allowed the
project to remove the remaining population of mink from Lewis and Harris, with a reduced staff resource, and increase
the trap night total to in excess of 500,000. To date, 2,198 mink have been caught, but only two non-breeding females and
associated males have been caught in Lewis and Harris in the last 18 months (no juveniles captured). The challenges of
geographical scale, terrain, climatic conditions and a continuously reducing staff complement have required an adaptive
management approach to achieve the project goal of a mink-free Outer Hebrides that benefits ground nesting birds and
migratory fisheries. This is viewed as a highly effective eradication project, and lessons learnt can be put into place for
other ambitious control programmes.
Keywords: adaptive management, anal gland lure, dog searching, monitoring techniques, remote alert systems, trapping
INTRODUCTION
The Outer Hebrides of Scotland support some of
the most important breeding populations of waders in
Europe. Species include redshank (Tringa totanus), snipe
(Gallinago gallinago), lapwing (Vanellus vanellus) and
oystercatcher (Haematopus ostralegus); with dunlin
(Calidris alpina) and ringed plover (Charadrius hiaticula)
nesting at the highest densities recorded anywhere in the
world (Stroud, et al., 2001). In recognition of this, many
of the nesting areas have been notified as Sites of Special
Scientific Interest (SSSI) under the Nature Conservation
(Scotland) Act 2004, and classified as Special Protection
Areas (SPA) under the EC Birds Directive, covering an
area of about 37,596 ha and 87,158 ha respectively.
At the international level, there are many more species
of birds that are represented by important populations
on these sites. Species include red-throated diver (Gavia
stellata), black-throated diver (Gavia arctica), great
northern diver (Gavia immer), hen harrier (Circus
cyaneus), merlin (Falco columbarius), short-eared owl
(Asio flammeus), greylag goose (Anser anser), mallard
(Anas platyrhynchos), teal (Anas crecca), wigeon (Anas
penelope), gadwall (Anas strepera), shoveler (Anas
clypeata), tufted duck (Aythya fuligula), eider (Somateria
mollissima), shelduck (Tadorna tadorna), red-breasted
merganser (Mergus serrator), golden plover (Pluvialis
apricaria), common sandpiper (Actitis hypoleucos), curlew
(Numenius arquata), corncrake (Crex crex), common tern
(Sterna hirundo). Ground nesting seabirds such as little
tern (Sternula albifrons), arctic tern (Sterna paradisaea)
(Clode & Macdonald, 2002) and arctic skua (Stercorarius
parasiticus) also occur in significant numbers.
Historically, the introduction of mink in Scotland has
been directly connected to the fur farming industry which
was established in the 1950s (Dunstone, 1993; Bonesi &
Palazon, 2007). In the Outer Hebrides this was mirrored
when two fur farms on the Isle of Lewis went out of
business in the 1960s resulting in a feral mink population
becoming established (Angus, 1993). Small scale control
operations carried out by sporting estates and an attempt by
SNH to prevent the mink population spreading south had
little effect. By 1999, breeding populations of mink were
established on North Uist and Benbecula (Harrington, et
al., 1999).
Invasive non-native species are one of the main causes
of biodiversity loss worldwide (Genovesi, 2009) and
predatory species, such as mink, can have a devastating
impact on native species (Macdonald, et al., 2007). The need
to manage non-native species is increasingly recognised as
a necessity to minimise these impacts (Bryce, et al., 2011).
In particular the impact of mink predation on ground
nesting colonial seabirds can have a significant effect, on
not only the breeding success of the species concerned but
also the long term viability of the population (Craik, 1997;
Craik, 1998). It is documented that mink at relatively low
densities can also seriously affect salmonids (Areal & Roy
2006). Atlantic salmon (Salmo salar) is a species in decline,
for which two Special Areas of Conservation have been
established in the Outer Hebrides. The removal of mink
can have significant beneficial consequences to a range of
species, especially in island ecosystems (Nordström, et al.,
2003).
In the Outer Hebrides the impacts of invasive mink over
decades had become a significant concern and the most
immediate effects were on the colonial nesting species
such as tern which were being severely impacted both in
terms of their productivity and also the loss of significant
numbers of adult birds.
MATERIALS AND METHODS
The Outer Hebrides of Scotland are a highly complex
archipelago of hundreds of islands which also includes
the third biggest island in the UK, Lewis and Harris. It
is characterised by vast expanses of moorland dissected
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 261–266. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
261
Island invasives: scaling up to meet the challenge. Ch 2A Other taxa: Mammals
by numerous convoluted freshwater lochs that amount
to approximately 24% of the freshwater linear edge of
Scotland’s total. Due to the remoteness of some areas,
and the general coastal nature of the American mink’s
behaviour, much of the work required the use of rigid hull
inflatable sea-going boats that were used extensively, as
well as Canadian open canoes in the complex freshwater
habitats.
The project design was first established during the
application process for LIFE funding but from its earliest
conception it was regarded as an innovative trial of
eradication techniques and an experimental project that
required continuous critical appraisal of the progress being
made.
Phase I of the Hebridean Mink Project was to remove
all mink from the southern isles of the Outer Hebrides;
South Uist, Benbecula and North Uist. The plan was also to
reduce the mink density on South Harris to create a buffer
zone between North Harris and Lewis (Helyar, 2005),
minimising re-immigration, see Fig. 1.
Live capture traps were chosen as the core removal
method due to the perception that kill traps were too
much of a risk in terms of by-catch. Later in the project it
was recognised that with experience, training and robust
protocols these risks could be reduced to extremely low
levels. Traps were made using 3 mm gauge wire mesh 18
× 15 × 60 cm and had galvanised steel doors. Caught mink
were despatched using a .22 calibre air pistol.
From November 2001, for a period of three months, a
total of 2,545 traps were dug into the ground and dressed
in order that they became part of the landscape, although
no more than 10% were open at any time. This provided a
large number of pre-located traps, which could be used in
rotation. The most efficient spacing of traps was established
to be approximately 500 m apart, but with a higher trap
density at individual den sites. Traps were initially baited
with horse mackerel (Trachurus trachurus), that was later
replaced or accompanied with anal gland lure which was
more efficient (Roy, et al., 2006).
The team was comprised of a project co-ordinator,
two trapper supervisors (one each on Harris and Uist),
six permanent trappers, and seasonal/casual workers who
assisted when required. The trappers worked a defined 37
hours per week, setting traps on a given route on a Monday
and closing them on a Friday. This gave a weekly total of
four trap nights per trap opened on any individual route.
Traps were most efficient in the first few days of opening,
see Fig. 2, and were left open for two weeks initially,
reducing to one week in subsequent years.
During 2004 and 2005 the trapping was punctuated
with high intensity trapping regimes. This involved
co-ordinating a group of up to 25 individuals to trap
simultaneously for a period of two to three weeks. The
extra support was drawn from external organisational staff
from DEFRA and the State Veterinary Service. The aim
was to increase the likelihood of capturing any remaining,
highly mobile mink.
Throughout Phase I, the most difficult areas to trap were
the offshore islands. Two Rigid-Hulled Inflatable Boats
(RHIB) were purchased and the associated training was
given to trapping staff to enable them to reach all areas.
Dog searches were introduced as a technique during the
summer denning period, when trapping is less efficient.
The final mink caught during Phase I was on 23rd March
2005 (see Fig. 2). This was followed by a further 5,567
trap nights and a ‘summer’ of dog searches with no further
mink sighted or caught, bringing Phase I to an end in June
2006. In the interim between Phase I and II, two trappers
were employed to keep the mink population low across the
South Harris buffer zone.
Phase II of the project aimed to remove all mink from
Harris and Lewis, to complete a full eradication from
the entire Outer Hebrides. This project commenced in
February 2007 and was initially due to end in March 2014,
but at present is still ongoing.
Trap locations were pre-determined through the use
of a GIS system. Trap positions were chosen by placing
them at obvious intersections of linear riparian or coastal
features, with 500 m buffer zones to ensure there were no
geographical gaps. When in the field, staff were given a
leeway of 50 m from the pre-positioned point to allow the
Fig. 1 The Outer Hebrides of Scotland showing the
Hebridean Mink Project areas completed with
timeframes.
262
Fig. 2 The number of mink caught per length of time an
individual trap remains open (SNH, 2006).
Macleod, et al.: Eradication of mink from Outer Hebrides
best position to be chosen in relation to the habitat. Once
traps were installed, they were mapped on the GIS system
to confirm absolute coverage of an area.
RESULTS
From 2007 there were 12 full-time trappers working
37 hours per week, reduced to six full-time trappers
in 2012, and three in 2015. By 2008, 7,500 live capture
traps had been permanently placed approximately 350–
500 m apart, across Lewis and Harris. Traps within an
area of approximately 100 km2 were open at one time,
for a period of four days. From 2008 to 2014, systematic
trapping continued from the south-to-north, twice yearly.
An exception was made in 2012 when the direction of
trapping was altered to a north-to-south direction to ensure
that specific areas were not always being trapped at the
same time of year.
A total of 532 mink from approximately 200,000
trap nights were caught during Phase I, see Table 1.
Approximately half of those caught were on the Uists,
compared with a similar number being caught in just
the south of Harris (Fig. 3). This demonstrated that the
mink population in the Uists had not yet reached carrying
capacity, as south Harris has very similar terrain, and large
areas of available habitat on South Uist had few captures.
Between November 2004 and March 2005, only females
were caught. This is likely a result of the trap density and
the wider ranging behaviour of male mink. No mink were
caught while trapping on the Uists between March 2005
and March 2006.
The mink population had been reduced to much lower
densities by 2013 An assessment was carried out to ascertain
whether the number of trap nights per 2.5 km2 area was
comparable, ensuring effort was distributed evenly across
the entire project site. An extensive monitoring programme
was set up in areas where there had previously been the
highest mink densities, with 17 monitoring devices
placed within 10 km2 areas of interest. Monitoring devices
included the use of footprint monitoring tunnels (clay and
carbon plate), footprint monitoring rafts, camera traps and
dog searches. These monitoring techniques were replaced
with more efficient technology in the form of remote
monitoring alarms (RMAs) which are activated when a
trap is triggered. The monitoring devices are attached to a
magnet which is pulled off when a trap is triggered, sending
an SMS or email message to chosen team members. The
devices were placed on traps situated in areas of good
mobile phone coverage.
In 2014 the team reduced to three trappers. In order to
maintain good monitoring coverage the live capture traps
which had historically caught were replaced with 140 ×
140 mm ‘116 Magnum bodygrip’ spring traps contained
in a bespoke designed wire mesh cage to exclude all nontarget species. Over a period of two years almost 450
bodygrip traps were installed and 120 live capture traps
were fitted with remote monitoring alarms.
Meanwhile on the Uist’s, a few individual mink reemerged in North Uist, which were immediately captured.
In December 2014 another two mink were sighted in the
northern end of North Uist, initiating another trapping
project on the Uists. Staff from the Uist Wader project
installed kill traps in a small area to detect any further mink.
As more traps were installed, more mink were caught, and
the trapping area was widened. From 2014 to the present
there has been an increase in both trap nights and the
number of mink caught on the Uists, with the trapping area
now extending from North Uist down to Locheynort and
due to be expanded to cover the entire Uists.
Year
beginning
Trap nights
Sep-2001
Sep-2002
Sep-2003
Uist
22,155
26,357
30,064
Harris
15,350
13,213
10,325
Sep-2004
Sep-2005
Total
20,037
1,114
100,824
2755
76
41,674
Mink captured by
trapping
Uist
Harris
42
73
80
54
56
64
13
0
191
38
1
230
Phase I
During the initial stages of Phase I it was quickly
determined that the traps were most effective at catching
during the first four days of being open. When opened for
a further four days during the second week, the trap still
caught mink but in far fewer numbers (Fig. 2).
In South Harris, due to a much higher trapper resource
for the area available to trap, this number of trapping
cycles per year was much higher, up to five times per year
compared to just twice a year, and resulted in a very quick
collapse in the territorial mink population. Thereafter,
trapped animals were generally those immigrating, from
the north, into the area, as indicated by a higher proportion
of males caught during this period.
An important difference in the capture locations
between the Uist’s and South Harris became evident in the
first few months of the project with the Uist’s showing a
significantly higher proportion of captures inland compared
to coastal habitats. The difference was largely due to the
Fig. 3 Number of mink captured per month and year on
both South Harris and the Uists during Phase I (SNH,
2006).
Mink captured per
1000 trap nights
Uist
Harris
1.85
4.76
2.97
4.08
1.86
6.20
0.65
0
1.89
13.79
13.15
5.15
Mink captured by dog searches
(dependent young in brackets)
Uist
Harris
0
6
12 (18)
1 (2)
4 (2)
(3)
1
0
37
3 (1)
0
18
Table 1 The numbers of trap nights, mink captures and trap successes in the Uists and South Harris during Phase I (Roy,
et al., 2015).
263
Island invasives: scaling up to meet the challenge. Ch 2A Other taxa: Mammals
greater availability of food resources inland in the Uist’s,
including a large number of duck and wader species closely
associated with the freshwater edge but, importantly, the
presence of field voles in the moorland habitat that were
absent from Lewis and Harris.
Phase II
Trapping took place over large areas, only moving on
when a low mink density had been achieved. This resulted
in a significant drop to the overall mink population, with
51% of the final captures so far, being caught in the first
two years. The final total of mink captures by March 2012
was 91% of the current figure. From April 2012 a further
116 mink were caught over the next three years, equating
to a further 6% of the current final total.
Initially there was a two week live trapping cycle
carried out but this was reduced to a one week cycle to
increase the efficiency of the knock down phase. Whilst
initially unpopular with the trapping staff, as they felt they
were leaving animals behind, the speed with which the
project reduced the mink population over a wide area soon
became apparent and the staff bought into the techniques
employed.
Fig. 4 Trap captures from Feb 2007–Nov 2016. Black bars
are mink caught; grey area is the trapping effort.
From 2007 to the present a total of 1,666 mink have
been caught from 527,431 trap nights, across Lewis and
Harris.
The major result of moving from live traps to a kill
trapping regime was an increase in the total trapping effort,
despite being reduced to a trapping team of just three. This
can be seen in Table 2, where up to 14,000 trap nights per
month were being achieved compared to approximately
2,000–2,500 per month when 12 trappers were employed
for live capture trapping.
The captures per unit effort have declined over time
but reflect the seasonality related to the trapability of
more mobile mink during the rut and the naivety of young
animals during the dispersal period (Fig. 4). The striking
issue, however, is the extremely long tail to the graph
which describes the extreme difficulty in catching the final
animals over such a large geographical area with a declining
staff resource (see Fig. 5). Two modelling exercises were
completed, (Shirley, et al., 2012) and the modelling
exercise carried out by Aberdeen University (Lambin, et
al., 2014) did predict that this would be the case: 80% of
Table 2 Actual trap nights and captures for all years of the
project from 2007 onwards.
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
Total trap
nights
14,914
24,755
38,749
40,894
33,446
26,665
21,695
41,954
126,088
158,271
2017*
87,000
Year
Total
captures
280
527
367
212
137
56
31
26
23
7
*2017 figures to the end of June.
264
4
Male
Female
146
266
171
98
53
31
19
16
14
5
134
261
196
114
84
25
12
10
9
2
3
1
Fig. 5 Number of mink captures per 1,000 trap nights
between February 2007 and December 2016.
iterations predicting eradication by 2017, using the data up
to 2011 and a trapping regime based on live capture and a
trapping regime of 12 trappers see Table 2.
During the final monitoring phase of the project in
Lewis and Harris, the final 1.5% of the mink population
was caught and functionally the population was eradicated
with only isolated individuals, unable to find a mate and
breed, left to track down. No juveniles have been caught in
Lewis and Harris since August 2015
Through an increase in trapping effort and larger areas
being monitored, there has been an increase in the number
of mink being caught on the Uists since 2014 (Table 3).
These animals are finally reducing in number as the same
kill trapping regime used in Lewis and Harris takes effect.
Table 3 Number of trap nights and mink caught from the
re-emerged population in the Uists between 2014 and
present.
Year
Trap
beginning nights
Jan 2014
36
Jan 2015
507
Jan 2016
3,776
Jan 2017* 4,799
Total
9,118
Total
captures Male Female Unk
5
5
0
0
22
12
5
5
63
38
21
4
41
23
14
3
131
78
40
12
*2017 figures to the end of June.
Macleod, et al.: Eradication of mink from Outer Hebrides
DISCUSSION
In 2001 when this project was initiated, there were few,
if any, successful eradications that used trapping as the
main technique for the removal of an invasive non-native
mammal; the only UK example being the coypu eradication
in Norfolk, (Gosling & Baker, 1989). In addition, there was
a limited range of literature available providing examples
of wildlife management project design and best practice to
follow (IUCN, 2000). The EU LIFE fund recognised that
the project would need to adapt as it progressed and agreed
to provide funding based on the understanding that it was
innovative in its concept, scale and design.
During Phase I, one of the main lessons learnt was the
necessity to ensure trap distribution was coordinated by the
supervisor. Initially trappers were relied upon to distribute
traps in the field according to their own judgement, with
only a specific distance between traps to guide them.
This meant that traps were situated in ideal locations for
catching mink, but trappers on the ground were unable to
ensure that there were no gaps in the overall trap coverage,
leading to irregular densities. Over time, the emergence
of better GPS technology enabled trappers to be more
efficient in the field and able to provide more accurate trap
locations. Establishing the most effective trapping schedule
was important as it was not possible to trap the entire area
at once with the staff available. A twice yearly minimum
trapping cycle of the entire trap network was vital to
ensure that all areas maintained sufficient trap nights, while
removing animals in a timely manner to avoid successful
breeding.
Despite the ongoing learning process during the first
phase, the project managed to achieve the removal of the
majority of the mink from the project area in just under
three and a half years, followed by a summer of monitoring.
It was thought at this point that it was very unlikely that
any mink remained in the Uists and Benbecula and that
eradication from these islands had been achieved.
The second phase of the project was an absolute
requirement if the gains of the first phase were to be
secured over an even larger geographical area and the
investment in the previous five years was to be protected.
Scottish Natural Heritage demonstrated significant
commitment in proceeding with Phase II, helped with
funding from the Esmeè Fairbairn Foundation, but from
the outset the budgetary constraints on the project were
clear. The modelling work undertaken by the Central
Science Laboratory (now Animal and Plant Health
Agency) indicated that 16 trappers would be ideal (Moore,
et al., 2003) but due to budgetary constraints, the project
proceeded with just 12. Restricted resources continued
into the project extension and the monitoring phases and
required significant adaptive changes to strategy and
efficiency in order to give the project the greatest chance
of success. It is undoubtedly true that the project has taken
longer due to these budgetary constraints and that, if fully
funded for the entire requirement of 10 years plus two extra
years to ensure eradication, significant savings could have
accrued over this period. This type of consecutive longterm funding is simply not available in the UK, (Lambin,
et al., 2014), as it does not fit with the funder’s requirement
to demonstrate success, generally within five years, and
exceeds the acceptable commitment levels between
political administrations.
Throughout the project, different methods were
employed at various stages to overcome the challenges
of limited resources. The addition of the bodygrip traps
instead of solely live traps enabled a high level of trapping
effort to be maintained with limited staff. Bodygrip traps
meant that trappers did not have to respond to triggered
traps immediately as the mink would be dead upon capture.
The initial concern of accidental by-catch was reduced
to an acceptable level through very strict protocol in the
practical setting of the trap, including the bespoke tunnels
which excluded all non-target species, and camouflage
technique.
Monitoring such a huge geographical area with only six
trappers was challenging and several monitoring devices
and techniques were trialled. Footprint rafts were not able
to withstand the extreme weather of winter months either
through wind or high water spate events, the cameras had
slow triggers and reset times which led to missed targets,
while the clay/carbon footprint monitoring required careful
set-up and protection from the elements to provide useful
data. In addition, the time between detecting the mink and
being able to initiate the trapping was too long to catch a
highly mobile individual. The acquisition of trap RMAs
were particularly useful for the monitoring period, giving
a precise time stamp for when a trap caught and enabling
further traps to be installed in the area immediately. This
was immediately effective as the mink population had
begun to cluster in their distribution, not only during the
rutting period which would be expected, but animals would
also set up territories next to existing ones rather than be
isolated and alone. This helped greatly once an individual
was trapped, as a localised trapping campaign could be
mobilised to catch a few additional animals.
The Hebridean Mink Project is now into its 16th year,
and has cost a total of £5.26M. The learning process has
been difficult and expensive and these lessons should be
passed on to others. There is a requirement for simple tools
to be developed that will allow projects to recognise the
key stages of eradication from the data they collect. These
comprise: population crash completion (knock down),
identification of groups of target species (cluster effect)
and difficult to trap areas to allow targeted action (trap
everywhere at the same intensity), detection of individuals
and their rapid removal (find the right monitoring
technique), effective and efficient long-term monitoring
and biosecurity (ensure the last individuals are not left
behind or re-introduced).
Clearly there are vast amounts of data associated
with this project that could provide a lifetime of analysis
opportunities of which only a tiny fraction has been used
here. Some of the intuitive assumptions made within this
paper need to be statistically analysed to provide definitive
proof of behaviours such as clustering, which appear so
obvious from mapping the capture data geographically
over time.
CONCLUSIONS
Phase II of the Hebridean Mink Project commenced
with a wealth of knowledge, practical scientific information,
techniques and trapping scheme models, not to mention a
core of well-trained staff. This no doubt contributed to the
success in greatly reducing the population of American
mink to near eradication. With the re-emergence of mink in
the Uists, the main lesson that can be learnt from Phase I, is
the importance of ensuring a sufficiently long monitoring
period with a sustained level of effort is implemented once
the last mink is thought to have been captured. Maintaining
sufficient resources to continue monitoring during the
final years following eradication is crucial to ensuring the
project’s success (Rout, et al., 2009). Any lapse in funding
before eradication is declared could result in the mink
being able to breed successfully and repopulate, leading to
financial losses that are both immediate and exponential.
265
Island invasives: scaling up to meet the challenge. Ch 2A Other taxa: Mammals
If eradication can be achieved in the Outer Hebrides
this would represent the largest mammalian eradication
initiative worldwide using just trapping techniques.
ACKNOWLEDGEMENTS
We would like to thank Charles Thompson and Rachel
Cartwright for their help with initial drafts of this paper
and Alasdair Macleod of the University of the Highlands
and Islands for his help in producing and maintaining
the database software beyond the contract requirements.
Further thanks to the staff of Newcastle and Aberdeen
Universities for their modelling, especially Xavier Lambin
for his scientific input and constructive practical ideas.
Most of all, a huge thanks to all who have been
involved with the project over the past 17 years. Too many
to name individually, this has been a collective long-term
achievement due to the hard work of hundreds of people:
in particular the trapping staff, who have walked the
equivalent of several times round the world in order to
check the trapping network in the worst of weather that the
west coast of Scotland could throw at them. Lastly, but not
least, the people of the Outer Hebrides for their help and
co-operation throughout the project. Thank you all.
REFERENCES
Angus, S. (1993). ‘A mink control programme for Lewis and Harris’.
Hebridean Naturalist 11: 78–84.
Areal, F. and Roy, S.S. (2006). ‘A management decision tool for mink
(Mustela vison) control in the Western Isles of Scotland (UK)’.
International Journal of Design and Nature and Ecodynamics 4: 16–31.
Bonesi, L. and Palazon, S. (2007). ‘The American mink in Europe: Status,
impacts, and control’. Biological Conservation 134: 470–483.
Bryce, R., Oliver, M.K., Davies, L., Gray, H., Urquhart, J. and Lambin,
X. (2011). ‘Turning back the tide of American mink invasion at an
unprecedented scale through community participation and adaptive
management’. Biological Conservation 144: 575–583.
Clode, D. and Macdonald, D.W. (2002). ‘Invasive predators and the
conservation of island birds: The case of American mink Mustela vison
and terns Sterna spp. in the Western Isles, Scotland’. Bird Study 49:
118–123.
Craik, C. (1997). ‘Long-term effects of North American mink Mustela
vison on seabirds in Western Scotland’. Bird Study 44: 303–309.
Craik, J.C.A. (1998). ‘Recent mink-related declines in gulls and terns in
west Scotland and the beneficial effects of mink control’. Argyll Bird
Report 14: 99–110.
Dunstone, N. (1993). The Mink. London: T & AD Poyser.
Genovesi, P. (2009). ‘Invasive alien species in a changing world’.
Biodiversity 10: 3–4.
Gosling, L.M. and Baker, S.J. (1989). ‘The eradication of muskrats and
coypus from Britain’. Biological Journal of the Linnean Society 38:
39–51.
Harrington, L.A., Grogan, A.L., Bonesi, L. and Macdonald, D.W. (1999).
Survey of American Mink, Mustela vison, on North Uist. Unpublished
report to Scottish Natural Heritage. Inverness: Scottish Natural Heritage.
Helyar, A. (2005). ‘The Ecology of American Mink (Mustela vison):
Response to Control’. PhD thesis. York, UK: University of York.
IUCN. (2000). Guidelines for the Prevention of Biodiversity Loss Caused
by Alien Invasive Species. Information Paper for CBD COP 5. Gland,
Switzerland: IUCN.
Lambin, X., Cornulier, T., Oliver, M.K. and Fraser, E.J. (2014). Analysis
and Future Application of Hebridean Mink Project Data. Scottish
Natural Heritage Commissioned Report No. 522. Inverness: Scottish
Natural Heritage.
Macdonald, D.W., King, C.M. and Strachan, R. (2007). ‘Introduced
Species and the Line between Biodiversity Conservation and
Naturalistic Eugenics’. In: D.W. Macdonald and K. Service (eds.) Key
topics in Conservation Biology, pp. 186–205. Oxford, UK: Blackwell
Publishing.
266
Moore, N.P., Roy, S.S. and Helyar, A. (2003). ‘Mink (Mustela vison)
eradication to protect ground-nesting birds in the Western Isles,
Scotland, United Kingdom’. New Zealand Journal of Zoology 30:
443–452.
Nordström, M., Högmander, J., Laine, J., Nummelin, J., Laanetu, N.
and Korpimäki, E. (2003). ‘Effects of feral mink removal on seabirds,
waders and passerines on small islands in the Baltic Sea’. Biological
Conservation 109: 359–368.
Rout, T.M., Thompson, C.J. and McCarthy, M.A. (2009). ‘Robust
decisions for declaring eradication of invasive species’. Journal of
Applied Ecology 46: 782–786.
Roy, S.S., Chauvenet, A.L.M. and Robertson, P.A. (2015). ‘Removal
of American mink (Neovison vison) from the Uists, Outer Hebrides,
Scotland’. Biological Invasions 17: 2811–2820.
Roy, S.S., Macleod, I. and Moore, N.P. (2006). ‘The use of scent glands
to improve the efficiency of mink (Mustela vison) captures in the Outer
Hebrides’. New Zealand Journal of Zoology 33: 267–271.
SNH. (2006). Mink Control to Protect Important Birds in SPAs in the
Western Isles. Final Report to EU LIFE III–Nature, September 2006.
Scottish Natural Heritage. Inverness: Scottish Natural Heritage.
Shirley, M., Rushton, S. and Lurz, P. (2012). Simulation Model for the
Management of American Mink, Neovison vison, on Harris and Lewis.
Scottish Natural Heritage, Commissioned Report No.452. Inverness:
Scottish Natural Heritage.
Stroud, D.A., Chambers, D., Cook, S., Buxton, N., Fraser, B., Clement, P.,
Lewis, P., Maclean, I., Baker, H. and Whitehead, S. (2001). The UK SPA
Network: Its Scope and Content. Three vols. Peterborough, UK: Joint
Nature Conservation Committee.
A. Ortiz-Alcaraz, A. Aguirre-Muñoz, F. Méndez-Sánchez, E. Rojas-Mayoral, F. Solís-Carlos, B. Rojas-Mayoral, E. Benavides-Ríos, S. Hall, H. Nevins and A. Ortega-Rubio
Ortiz-Alcaraz, A.; A. Aguirre-Muñoz, F. Méndez-Sánchez, E. Rojas-Mayoral, F. Solís-Carlos, B. Rojas-Mayoral, E. Benavides-Ríos, S. Hall, H. Nevins
and A. Ortega-Rubio. Ecological restoration of Socorro Island, Revillagigedo Archipelago, Mexico: the eradication of feral sheep and cats
Ecological restoration of Socorro Island, Revillagigedo Archipelago,
Mexico: the eradication of feral sheep and cats
A. Ortiz-Alcaraz1, A. Aguirre-Muñoz1, F. Méndez-Sánchez1, E. Rojas-Mayoral1, F. Solís-Carlos1,
B. Rojas-Mayoral1, E. Benavides-Ríos1, S. Hall2, H. Nevins3 and A. Ortega-Rubio4
Grupo de Ecología y Conservación de Islas, A.C., Ensenada, México. <antonio.ortiz@islas.org.mx>. 2National Fish
and Wildlife Foundation. Washington, D.C., USA. 3American Bird Conservancy. The Plains, Virginia, USA. 4Centro de
Investigaciones Biológicas del Noroeste, S.C., La Paz, México.
1
Abstract Socorro Island is part of the Revillagigedo National Park, Mexico. At 132 km2, it is the Mexican island with
the highest level of endemism. It provides habitat for 117 vascular plant species, 26% of which are endemic. There is
also an endemic blue lizard (Urosaurus auriculatus) and eight endemic terrestrial birds. Socorro’s ecosystem had been
heavily degraded by invasive mammals for the past 140 years. Feral sheep (Ovis aries) destroyed one third of the island’s
habitat and feral cats (Felis catus) severely impacted the island’s avifauna and the Socorro blue lizard. Together, feral
sheep and cats are responsible for the extinction in the wild of the Socorro dove (Zenaida graysoni) and the Socorro elf
owl (Micrathene whitneyi graysoni) and have been a serious threat to other vulnerable species, particularly Townsend’s
shearwater (Puffinus auricularis). As such, the island’s restoration is a high priority. We conducted a feral sheep eradication
from 2009 to 2012, using aerial and terrestrial methods, aided by Judas sheep and trained dogs, to kill 1,762 animals. The
vegetation recovery has been remarkable, as well as the improvement of soil properties such as compaction, nitrogen,
organic carbon, phosphorus, and calcium. In 2011, we initiated a feral cat control programme, which soon became an
eradication project. The ongoing feral cat eradication has been a challenge, due to Socorro’s large size, vegetation and
topographical complexity. By December 2016, 502 cats had been dispatched, using soft leg-hold traps equipped with
telemetry transmitters and lethal traps: a total effort of 50,000 trap-nights. Cat abundance has decreased very significantly
and catch per unit of effort indicates that the eradication is nearing completion. The abundance of the Socorro blue lizard
and terrestrial birds has already increased. We estimate completing the feral cat eradication by the end of 2017, when we
will shift to a verification of eradication phase.
Keywords: exotic mammals, habitat recovery, outcomes of eradications
INTRODUCTION
Mexican islands are known for their high biodiversity
richness. They are home to many endemic species and
are important breeding grounds for a variety of birds
and marine mammals (Aguirre-Muñoz, et al., 2011).
Unfortunately, these ecosystems are suffering serious
impacts resulting from human activity. Exotic species are
among the main causes of biodiversity loss and ecological
disequilibrium in many environments (Courchamp, et
al., 2003). Herbivores, like feral sheep (Ovis aries), have
caused serious ecological impacts on insular ecosystems.
In 1869, 100 sheep were introduced to Socorro Island, in
Revillagigedo National Park, Mexico (Fig. 1) for ranching.
Over time, they became feral, successfully adapting
to island conditions (Levin & Moran, 1989; ÁlvarezCárdenas, et al., 1994; Brattstrom, 2015). In the absence
of natural predators, the sheep population grew to be about
5,000 individuals by 1960 (Villa, 1960). This reduced to
around 2,000 in 1988 as a result of increased hunting by the
Mexican Navy (Walter & Levin, 2008), but they became
the main cause of the island’s poor ecological condition
Fig. 1 Location of sampling sites of vegetation and soil.
(Richards & Brattstrom, 1959; Veitch, 1989). Since their
introduction, feral sheep have caused huge modifications
to the natural habitat. Erosion rates and loss of vegetation
caused by the presence of sheep were documented, along
the southern-central region of the island (León de la Luz,
et al., 1994; Maya-Delgado, et al., 1994; Rhea, 2000).
Nearly 30% of the original soil and vegetation on Socorro
Island was lost due to erosion caused by feral sheep
(Ortega-Rubio, et al., 1992). Among the most significant
changes to the original floral composition has been an
increase in the presence of grasses and shrub species, as
well as a reduction in the area covered by native flora.
Sheep aid the propagation of introduced plant species,
dispersing seeds in their coat and excrement. The change
in native vegetation has been observed in every habitat
that sheep occupied (SEMARNAT, 2004). Another serious
threat is the feral cat (Felis catus), which have severely
impacted the island’s bird communities and the endemic
Socorro tree lizard (Urosaurus auriculatus) (Arnaud, et
al., 1993; Arnaud, et al., 1994). Together, feral sheep and
cats are responsible for the extinction in the wild of the
Socorro dove (Zenaida graysoni) and the Socorro elf owl
(Micrathene whitneyi graysoni), and pose a serious threat
to other vulnerable species, such as Townsend’s shearwater
(Puffinus auricularis) (Martinez-Gomez & Jacobsen
2004). The eradication of feral cats represented another
serious challenge, as Socorro is a large and complex
island, and little baseline information was available on the
distribution and abundance of cats (Arnaud, et al., 1994).
Fortunately, technologies have been developed on other
islands of Mexico and the world to achieve the eradication
of these predators (Bester, et al., 2002; Wood, et al., 2002;
Algar, et al., 2010; Aguirre-Muñoz, et al., 2011; LunaMendoza, et al., 2011; Parkes, et al., 2014). The successful
implementation of an eradication campaign of this type is
essential to determine the basic aspects of the species, the
impact of the methods applied on the native fauna, and the
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 267–273. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
267
Island invasives: scaling up to meet the challenge. Ch 2A Other taxa: Mammals
development of an official eradication plan (Veitch 1989;
Arnaud, et al., 1994; Donlan, et al., 2003; Dowding, et al.,
2009).
MATERIALS AND METHODS
Study site
The volcanic Socorro Island is the largest and most
diverse portion of the Revillagigedo Archipelago, a World
Heritage site in Mexico that was listed by UNESCO
in July 2016. It is located in the tropical eastern Pacific,
460 km from the Baja California Peninsula and 700 km
from Manzanillo, Colima. It has an area of 132 km2 and a
maximum altitude of 1,040 m (18º 47´ N, 110º 58´ W). Due
to its remoteness, the island is a strategic point for Mexico’s
military personnel (40–50 people) situated at a naval base
located in the southernmost part of the island. Additionally,
Socorro Island has critical biodiversity significance through
a high level of endemism, due to its isolated position A
remarkable number of its native biota are exclusively found
in this part of the world. Approximately one third of the
118 species of native vascular plants inhabiting the island
are endemic. The native fauna is comprised of one endemic
reptile species and almost 101 species of birds, of which
eight of the terrestrial birds are endemic (SEMARNAT,
2004).
Feral sheep eradication
Monitoring: To identify the main areas where sheep
were distributed, several flights (on a Beechcraft Bonanza
aircraft) were made over the island during October 2005.
At the same time, Mexican Navy officers conducted land
surveys on foot and in motorised vehicles (ATV´s).
Aerial hunting: The aerial hunting stage was carried
out using a single turbine helicopter (model MD369D),
between April 20 and 29, 2009; supported by a GPS to
record the flight trajectories. Two hunters were shooting
simultaneously during the flight, using semi-automatic
rifles and shotguns. All flights took place between 07:00
and 11:00 h, and between 16:00 and 19:00 h; at an average
speed of 42 km/h and average height of 35 m.
Judas sheep: During the aerial hunting, 12 live animals
were captured, to be used as ‘Judas’ sheep (individuals that
serve to help locate remaining herds; Taylor & Katahira,
1988). These animals were neutered and fitted with radiotelemetry collars (Telenax, Mexico). These Judas sheep
were deployed back to the sites where they were captured.
Terrestrial hunting and trapping: From February
2010 to April 2012, 4–7 experienced hunters carried out
terrestrial hunting. Each one was equipped with a handheld
GPS to record their hunting tracks, rifles (calibre .222, .243
and .308) with telescopic sights, as well as a 12-gauge
shotgun with cartridges. Supported by the Judas sheep, it
was possible to identify sheep herds. Simultaneously, leghold traps (Oneida Victor Soft Catch # 3) and snare traps
were used on previously identified sheep trails; both types
of trap were checked daily.
Hunting dogs: for the last stage of the sheep eradication,
we used two hunting dogs (beagle and foxhound) to track
down the remaining sheep herds; the dogs were fitted
with GPS collars to record their locations and movements
(Ortiz, et al., 2016a).
Feral cat eradication
Trapping: The eradication method consisted mainly of
catching cats using leg-hold traps (#1 ½) and lethal traps
(Conibear Bodygrip traps 10”: Rauzon, 1985; Twyford,
et al., 2000; Phillips, et al., 2005; Rodríguez, et al., 2006;
268
Rauzon, et al., 2008; Luna-Mendoza, et al., 2011). Leghold traps with pads were used in 220 sites over the duration
of the expedition (21–51 days), baited with a commercial
cat bait made of seafood, tuna or fried sardine (Brothers,
1982). Traps were checked daily from 7:00 to 10:00 h.
Lethal and leg-hold traps were located in sites of difficult
access, equipped with telemetry systems (ATS, mammal
trap monitor Series M4000) to determine whether they had
been activated from a distance (Will, et al., 2010). Once
cats were captured, these were euthanised by intramuscular
injection of an anaesthetic and lethal intracardiac injection
(pentobarbital). As a secondary method, night hunting was
conducted using .222 calibre rifles with telescopic sights,
and lamps (Kohree 80,000 lux: Ortiz, et al., 2016a).
Soil quality assessment
Soil compaction: In 2013, soil penetration resistance
measurements were taken using a penetrometer (Soil
Compaction Tester Dickey-john®) within 20 vegetation
transects (Fig. 1). Sites were categorised as: bare soil
sites, those with 50% recovered vegetation and those with
100% recovered vegetation. Additionally, soil compaction
measures were taken in sites with 100% vegetation
coverage, not previously disturbed by the sheep (ND = not
disturbed). Fifty replicates were obtained in each category,
resulting in a total of 200 measurements. An analysis
of variance and Tukey’s honest significance test were
performed to analyse the differences among the different
categories of vegetation cover.
Physicochemical soil parameters: soil samples of
approximately 1 kg each, were collected from each transect
at a depth of 0–10 cm: 16 samples were obtained in each one
of the soil categories (N= 64). Subsequently, the following
physicochemical parameters were determined: pH and
electrical conductivity, total nitrogen by the Dumas method
in a LECO nitrogen analyser; organic matter by the method
of Walkley-Black; phosphorus by colorimetric reading of
a spectrophotometer, and calcium and magnesium by the
EDTA method.
Vegetation recovery assessment
Field assessment: Prior to eradication, in 2009,
vegetation data collection was started to obtain a baseline
scenario of the degraded environment, with the aim of
making subsequent comparisons possible, and to detect
signs of recovery after sheep removal. The estimation
of sheep overgrazing consequences on the island was
determined by selecting 20 plot sites. Transects of 10 m
× 100 m were established in the more disturbed areas, to
identify pioneer species on eroded soils; all plants were
identified and counted. Plot sites were categorised in:
forest (six replicates), mixed scrub (six replicates) and
eroded surface (seven replicates) (León de la Luz, et al.,
1994). The vegetation monitoring continued from 2011 to
2016. Analysis of variance was performed (rANOVA) to
analyse differences in vegetation cover and in the number
of species over the years of the study (Ortiz, et al., 2016b).
Normalised Difference Vegetation Index (NDVI): To
identify changes in vegetation cover the photosynthetic
vegetation vigour of the island was obtained, quantified
with the Normalized Difference Vegetation Index (NDVI).
Supported with QGIS software, two maps were generated. A
“pre-eradication” map, created using a QuickBird satellite
image, dated on May 11, 2008; and a “post-eradication”
map, generated with a WorldView 2 satellite image, dated
on May 9, 2013. Finally, the change in vegetation cover
between the two dates was determined by subtracting
the 2008 image NDVI raster pixel image values from the
image of 2013, considering only differences exceeding 0.2
(bare soil).
Ortiz-Alcarez, et al.: Restoration of Socorro Island, Mexico
observer moved to the next counting point, located 250
m away, with a five-minute break before starting the next
count. The statistical test rANOVA was run to determine
the effect of season and habitat type on the total number
of birds, plus Student t-tests for paired samples with a
Bonferroni adjustment, to compare sightings during the
different seasons.
RESULTS
Sheep eradication
Aerial hunting: a total of 35 flight hours was achieved
in one week during the aerial hunting stage, in which most
of the island was covered, with an average flight time of
1h 20 min per event; this resulted in removal of 1,257
individuals. The aerial hunting ceased when sheep became
difficult to locate, and relatively few animals were being
shot within a flight event. This method was selected due
to its proven effectiveness in achieving rapid eradication
(Campbell & Donlan, 2005) and was ideal on this island
owing to its tropical conditions, which allowed the
carcasses to decompose rapidly.
Fig. 2 Location of transects of Socorro tree lizard (lines)
and count-points for birds (quadrats).
Monitoring of native fauna
Socorro Island tree lizard: To evaluate if the vegetation
recovery was promoting any improvement of native animal
populations, we monitored the Socorro Island tree lizard,
during April and October, from 2012 to 2017. Twenty-four
transects were set up in three different types of habitats
(eight replicates per habitat type): forest, deciduous
scrubland and eroded surfaces or areas impacted by sheep
(León de la Luz, et al., 1994), each measuring 6 m × 100 m
(Fig. 2). Transects were each visited on three consecutive
days, between 10:00 and 12:00 h, during two different
seasons (dry and rainy). Density (D) was estimated
using the formula: D = (n/2wL), where n is the number
of individuals recorded, L is the total transect length,
and w is the width of the transect on each side of the line
(Gallina & López-González, 2011). A one-way repeated
measure analysis of variance (rANOVA) was conducted to
determine differences in seasonality and in habitat type on
tree lizard density, the statistical software R (Version 3.2.2)
was used for the analysis.
Terrestrial birds: Terrestrial birds were also monitored,
using the point-count technique. Six transects were
established during April and October (two seasons per year)
from 2012 to 2017 (Fig. 2). The monitoring was carried out
from 6:30 to 9:30 h and was repeated on three consecutive
days, during the dry and rainy season, respectively. At
each site, all birds observed within a radius of 25 m in a
time span of five minutes were counted. Subsequently, the
Ground hunting and trapping: 505 sheep were
dispatched during the ground hunting stage, which
comprised a nine-month period of hunting, over two
years (March 2010–April 2012). Judas sheep were mostly
effective when there were more remaining sheep, due to
an increased probability of aggregation. Hunting dogs
were used only at the final stage of eradication to locate
the last ten remaining animals, which were difficult to
locate for hunters. A total of 1,762 sheep were dispatched
from Socorro Island in a three-year eradication campaign
(Table 1).
Feral cat eradication
By December 2016, 502 cats had been removed, using
soft leg-hold traps equipped with telemetry transmitters
and lethal traps (body grip). Traps were placed in more than
250 sites on the island. Up to that date, there was an effort
of more than 50,000 trap-nights. The success of cat capture
during the trapping period fluctuated throughout the year
(greater catch in January–May, dry season; and lower catch
in June–October, rainy season). However, a clear trend to
a smaller population was noted at a multi-year timescale
(Fig. 3). In general terms, the success of capture is greater
in the dry season and decreases during the rainy season. It
is expected that cat eradication will cease in 2018; if that is
the case, absence confirmation monitoring will be carried
out in 2019.
Soil quality assessment
The results of the soil compaction assessment showed
that eroded soils were the most compacted and a trend
towards compaction reduction on areas with recovered
Table 1 Feral sheep dispatched on Socorro Island.
Year
Months
2009
2010
May
Mar–Apr
Jul
Apr
Aug–Sep
Nov–Dec
Apr
2011
2012
Total
Personnel Hunter hours Distance
(km)
7
6
5
4
4
4
35 (helicopter)
1,323
588
512
728
420
240
3,811
815
460
433
644
385
216
2,953
Judas
sheep
Trap
nights
53
18
11
4
86
900
650
1,550
Captured Dog
sheep
hours
41
8
49
49
49
Sheep
removed
1,257
355
48
67
25
8
2
1,762
269
Island invasives: scaling up to meet the challenge. Ch 2A Other taxa: Mammals
Fig. 3 Success of capture of feral cats.
vegetation was observed (50% and 100% recovered
vegetation cover). Transects that retained eroded soils (0%
vegetation) because of the sheep trampling, showed greater
penetration resistance (>300 pounds-force per square inch,
or psi, to 12 inches deep); at sites with 50% and 100%
recovered vegetation cover, soils were also compacted and
shallow (100–120 psi to three inches deep) and became
more compacted at greater depths (300 psi to 24 inches
deep); at sites with 100% coverage without disturbance
(ND), the soil showed little variation (230–300 psi until
21 inches depth), which was in the optimal range for the
growth of most plants (from 200–400 psi to 24 inches),
which could be due to the constant, stable conditions.
Significant differences (p < 0.001) were observed among
sites with 0% and those with 50% and 100% recovered
vegetation cover. The results of physicochemical analyses
of soil samples showed increased nutrients: pH values
remained close to neutral, showing a significant difference
(p < 0.021) between sites without vegetation and 50%
vegetation cover (results in Ortiz, et al., 2016b). Electrical
conductivity, which is an indicator of salt presence in soil,
was also significantly different (p < 0.013) between the
eroded and 100% vegetation covered sites, although no
difference was observed between eroded soils and those
that were not disturbed. In the case of total nitrogen, organic
carbon, phosphorus and calcium, sites with recovered
vegetation were significantly different (p < 0.001) to those
with erosion. Both nitrogen and organic carbon doubled,
while phosphorus and calcium values almost tripled in
places with increased vegetation cover compared to the
eroded sites. Meanwhile, magnesium showed significant
differences among the eroded sites (0%, 50%, and 100%
recovered vegetation cover) and undisturbed sites (100%
ND; Ortiz, et al. 2016b). The sites that were never
altered by the presence of sheep exhibited a magnesium
concentration twice that of disturbed sites.
Fig. 4 Increase in vegetation cover (a) and species richness
(b) on Socorro Island.
Monitoring of native fauna
The results of tree lizard monitoring reveal that the
population is increasing, taking into consideration both the
dry and rainy seasons (Fig. 5). Lizard density fluctuated
significantly between seasons since the trapping of cats
started on Socorro Island (p = 0.014). The number of
birds sighted from 2012 to 2015 also showed significant
differences between seasons (p = 2.2 × 10-4). Although
population fluctuation is evident over the years of the
study, there is an increase of birds in the dry seasons of
2014 and 2015 (Fig. 6). Significant differences were found
between November 2012 and the rest of the monitoring
time points (except November 2014). No significant
differences were found between dry seasons during the
years 2013 to 2015. The most abundant species was the
Socorro tropical warbler (Setophaga pitiayumi graysoni),
followed by the Socorro wren (Troglodytes sissonii), and
the towhee (Pipilo maculatus socorrensis), all of them
endemic to the island.
Vegetation recovery assessment
Calculations (comparison of the images from 2008 and
2013) showed a difference of 1452 ha, which is equivalent
to vegetation recovery of 11% of the island surface. The
eastern part of the island was the area with the greatest
habitat disturbance (Álvarez-Cárdenas, et al., 1994), and
where the greatest vegetation recovery seemed to have
occurred within the analysed period. Due to the presence
of sheep, most of the evaluated sites lacked vegetation, and
few species were present in 2009 (Fig. 4). Additionally,
trails made by the sheep were observed to have compacted
soils. Statistical tests showed significant differences from
2009 to 2013 in the number of species present in the
eroded sites as well as in percentage cover. It was possible
to record obvious recovery in all the habitats in 2013, i.e.
the forest habitat with the highest values, followed by the
mixed scrub, and then the eroded surface.
270
Fig. 5 Density of the Socorro tree lizard.
Ortiz-Alcarez, et al.: Restoration of Socorro Island, Mexico
rainfall, could be causing leaching and replacing cations
with H+ ions, acidifying the soil.
Fig. 6 Numbers of birds sighted.
DISCUSSION
Compared with other islands where goats or sheep have
been eradicated (Van Vuren, 1992), the Socorro Island feral
sheep eradication can be considered highly effective as it
was completed in three years, when similar projects have
taken 3–5 years, and even decades, to conclude (Campbell
& Donlan, et al., 2005). Moreover, the methods used
reduced project cost, which was US$ 38/ha, while other
sheep eradication projects, such as that developed on Santa
Cruz Island, California, cost US$ 80/ha, due to capture and
transport of sheep to the continent (Faulkner & Kessler,
2001).
The capture of cats increases during the dry season
(January–May) and decreases in the rainy season (June–
October). The results differ between wet and dry seasons,
since moist land interferes with the installation of leg-hold
traps, and dry substrate is unavailable to cover them. Rain
also compacts and hardens the substrate covering the traps,
hence restraining their activation. At the same time, another
key factor that affects trapping in the rainy season is the
higher abundance of land crabs (Gecarcinus planatus),
which consume the bait placed in traps, or activate traps
when attempting to reach the bait. The combination of
lethal traps and telemetry devices is essential during
trapping in the most remote areas of the island. In this
way, traps do not have to be checked daily but every five to
seven days any bait lost to insects (mainly ants) and crabs
is replenished (Parkes, et al., 2012).
The changes in soil physicochemical properties on
Socorro Island seem to be related to the gradual recovery
of vegetation after the eradication of feral sheep. Prostrate
Chamaesyce sp. and Erigeron socorrensis have been
observed to have a great capacity to retain soil. Hyptis
pectinata and Pteridium caudatum established in high
densities; in addition to retaining soil, they have generated
much organic matter. Possibly the most successful species
to colonise disturbed areas has been Dodonaea viscosa,
which has a great ability to germinate in eroded soil (CampaMolina, 1989), generating organic matter and preventing
the germination of other species (Castellano & Valone,
2007). In the absence of trampling, soil aggregate stability
increases, which enhances permeation, reduces erosion,
and may promote nutrient accumulation and soil retention
(Allington & Valone, 2010). As pioneer plants began to
establish, the ground became less compacted because the
roots of plants, particularly annual grasses in this instance,
act as biological perforators, also incorporating organic
matter into the soil. Once the roots die and shrink, these
pores are large enough to allow the roots of perennial
shrubs to penetrate (Sellés, et al., 2012). Greater ease of
water movement in the soil matrix, coupled with heavy
Both the results obtained with the NDVI calculation
and field observations suggested that some pioneer plants
had the ability to germinate on eroded soils and were
instrumental in the succession process by providing the
right conditions for seeds of tree species to germinate.
The progressive increase in vegetation cover reduces
soil compaction and restores the biogeochemical cycles
of essential nutrients, such as nitrogen, phosphorus,
and calcium, which are essential for the recovery of
communities and the ecosystem in general, as well as the
incorporation of carbon on the ground, which is essential
for the proper functioning of important microbiological
components. Any change in the habitat that produces
changes in litter production, soil aeration, or any other
factor affecting microorganisms will be reflected in
changes in biogeochemical cycles, such as those of carbon
and nitrogen (Hartmann, et al., 1997).
We found differences in the number of species and
vegetation cover in the sampling area between 2009 and
2013. The forests and mixed scrub areas showed the
greatest recovery, probably favoured by their vegetal
components and the permanence of seed banks, due to a
more stable landscape, water availability, and precipitation
patterns. The endemic tree species recovering were
Guettarda insularis and Psidium socorrense. The smaller
number of plant species found in the isolated patches of
mixed scrub included in large expanses of erosion could
be due to the steepness of slopes and wind exposure.
Gravity also makes the permanence of naturally occurring
soil seed banks difficult. Some species of exotic grasses
have increased with sheep eradication because they are no
longer grazed.
The Socorro Island tree lizard was found at higher
densities in the deciduous scrubland, being less abundant
in forests at higher altitudes. The results of this particular
study show that the density of lizards on eroded surfaces
was as high as 43 individuals/ha after cat abundance was
reduced, however Galina et al. (1994) reported not having
observed lizards in these areas. This may be due to a
gradual recovery of the vegetation resulting from the recent
eradication of sheep (Ortiz-Alcaraz, et al., 2016a; OrtizAlcaraz, et al., 2016b) and to the sustained trapping of cats
in these areas. Lizard abundance was slightly higher during
the rainy season, likely due to greater food availability. As
the cat eradication programme in the eastern area of the
island has progressed, the predation pressure of cats on
the lizard population has decreased. Lizards are a major
component of the cats’ diet (50% of faecal samples of cats
analysed contained lizard remains; Arnaud, et al., 1993).
The vegetation type where the highest number of birds was
observed was the forest (Ficus-Guettarda-Ilex), especially
in the highest parts of the island, where the recovery of
vegetation resulting from the absence of grazing sheep
has led to greater availability of food and shelter against
predators (Rodríguez-Estrella, et al., 1994). On the other
hand, special efforts have also been made to eradicate cats
in the forest, aiming to protect the native bird species, such
as the Townsend´s shearwater (Ratcliffe, et al., 2009).
The plans developed by Veitch (1989), Arnaud, et al.
(1993) and Parkes, et al. (2012) are all in agreement that
feral cats can be eradicated using traditional techniques:
trapping and night hunting. However, the experience on
the island has shown the importance of using detection
dogs to locate the remaining cats, either during the day, in
their dens (placing traps to catch them), or at night, killed
by hunting (Tortora, 1982; Veitch, 2001), as well as the
statistical confirmation of absence (Ramsey, et al., 2011).
271
Island invasives: scaling up to meet the challenge. Ch 2A Other taxa: Mammals
CONCLUSION
The aerial hunting method proved to be an ideal
technique for the eradication of sheep from Socorro Island.
It enabled the eradication team to dispatch a high number of
animals in a few days of work, while allowing the hunters
to access difficult terrain. The use of Judas sheep and
hunting dogs was crucial for completing the eradication.
Removing the exotic herbivorous species from the
island is a conservation tool, which is evident in recovery
of the natural environment. Habitat fragmentation and
degradation caused by the presence of sheep was evident
on the island, where the main impact was on vegetation.
The resistance of native species has been important, not
only in the relatively rapid recovery of the vegetation
cover, but also in offering the possibility of recovering the
former island vegetation. The results reflect the important
role of vegetation in erosion control, both for establishing
mechanical support due to plant roots in the soil structure
and in capturing water flow and nutrients, providing fresh
organic matter to the soil and restoring biogeochemical
cycles and ecosystem processes.
With habitat recovery and progress in the feral
cat eradication, wildlife recovery is expected as food
availability and resources for the native species of the
island gradually increase and predation decreases. Socorro
Island seems resilient enough to recover over a relatively
small-time scale, after the removal of the pressures caused
by exotic mammals.
REFERENCES
Aguirre-Muñoz, A., Samaniego-Herrera, A., Luna-Mendoza, L., OrtizAlcaraz, A., Rodríguez-Malagón, M., Méndez-Sánchez, F., FélixLizárraga, M., Hernández-Montoya, J.C., González-Gómez, R., TorresGarcía, F., Barredo-Barberena, J.M. and Latofski-Robles, M. (2011).
‘Island Restoration in Mexico: Ecological outcomes after systematic
eradications of invasive mammals’. In: C.R. Veitch, M.N. Clout and
D.R. Towns (eds.) Island invasives: eradication and management, pp.
250–258. Occassional Paper SSC no. 42. Gland, Switzerland: IUCN
and Auckland, New Zealand: CBB.
Algar, D., Angus, G.J., Brazell, R.I., Gilbert, C. and Withnell, G.B.
(2010). ‘Eradication of feral cats on Faure Island, Western Australia’.
Journal of the Royal Society of Western Australia 93: 133–140.
Campbell, K. and Donlan, J. (2005). ‘A review of feral goat eradication
on islands’. Conservation Biology 198: 1362–1374.
Castellano, M.J. and Valone, T.J. (2007). ‘Livestock, soil compaction and
water infiltration rate: Evaluating a potential desertification recovery
mechanism’. Journal of Arid Environments 71: 97–108.
Courchamp, F., Chapuis, J.L. and Pascal, M. (2003). ‘Mammal invaders
on islands: Impact, control, and control impact’. Biological Reviews 78:
347–383.
Donlan, C.J., Tershy, B.R., Campbell, K.J. and Cruz, F. (2003). ‘Research
for requiems: The need for more collaborative action in eradication of
invasive species’. Conservation Biology 17: 1–2.
Dowding, J.E., Murphy, E.C., Springer, K., Peacock, A.J. and Krebs, C.J.
(2009). ‘Cats, rabbits, myxoma virus, and vegetation on Macquarie
Island: A comment on Bergstrom et al. (2009)’. Journal of Ecology 46:
1129–1132.
Faulkner, K.R. and Kessler, C.C. (2011). ‘Live capture and removal
of feral sheep from Eastern Santa Cruz’. In: Veitch, C. R.; Clout, M.
N. and Towns, D. R. (eds.). 2011. Island invasives: eradication and
management, pp. 295–299. Occassional Paper SSC no. 42. Gland,
Switzerland: IUCN and Auckland, New Zealand: CBB.
Gallina, S. and C. López-González (eds.) (2011). Manual de Técnicas
para el Estudio de la Fauna. Volumen I. Querétaro, México: Universidad
Autónoma de Querétaro–Instituto de Ecología, Civil Association.
Hartmann, A., Assmus, B., Kirchhof, G. and Schloter, M. (1997). ‘Direct
Approaches for Studying Soil Microbes’. In: J.D. van Elsas, J.T.
Trevors and E.M.H. Wellington (eds.) Modern Soil Microbiology, p.
172. New York: Marcel Dekker Inc.
León de la Luz, J.L., Breceda, A., Coria-Benet, R. and CancinoHernández, J. (1994). ‘Asociaciones Vegetales’. In: A. Ortega-Rubio
and V. Castellanos (eds.) La Isla Socorro, Reserva de la Biosfera
Archipiélago de Revillagigedo, México, pp. 115–141. Mexico: Centro
de Investigaciones Biológicas del Noroeste.
Levin, J.A. and Moran, R. (1989). The Vascular Flora of Isla Socorro,
México. University of Michigan, USA: San Diego Society of Natural
History.
Luna-Mendoza, L., Barredo-Barberena, J.M., Hernández-Montoya,
J.C., Aguirre-Muñoz, A., Méndez-Sánchez, F., Ortiz-Alcaraz, A. and
Félix-Lizárraga, M. (2011). ‘Planning for the eradication of feral cats
on Guadalupe Island, México: home range, diet, and bait acceptance’.
In: C.R. Veitch, M.N. Clout and D.R. Towns (eds.) Island invasives:
eradication and management, pp. 192–197. Occassional Paper SSC no.
42. Gland, Switzerland: IUCN and Auckland, New Zealand: CBB.
Allington, G.R.H. and Valone, T.J. (2010). ‘Reversal of desertification:
The role of physical and chemical soil properties’. Journal of Arid
Environments 74: 973–977.
Maya-Delgado, Y., Salinas-Zavala, F. and Troyo-Diéguez, E. (1994).
‘Estado Actual del Suelo y Propuestas para su Conservación’. In: A.
Ortega-Rubio and V. Castellanos (eds.) La Isla Socorro, Reserva de la
Biosfera Archipiélago de Revillagigedo, México, pp. 63–75. Mexico:
Centro de Investigaciones Biológicas del Noroeste.
Álvarez-Cárdenas, S., Castellanos-Vera, A., Galina-Tessaro, P. and
Arnaud, G. (1994). ‘Aspectos de la Población y el Hábitat del Borrego
Doméstico (Ovis aries)’. In: A. Ortega-Rubio and V. Castellanos (eds.)
La Isla Socorro, Reserva de la Biosfera Archipiélago de Revillagigedo,
México, pp. 301–317. México: Centrode Investi-gaciones Biológicas
del Noroeste.
Ortega-Rubio, A., Castellanos-Vera, A., Arnaud, G., Maya-Delgado, Y.,
Rodríguez, R., León de la Luz, J.L., Cancino, J., Jiménez, C., Llianas,
J., Álvarez, S., Galina, P., Breceda, A., Troyo, E., Salina,s F., Díaz, S.,
Servín, R., Romero, H., Rodríguez, A. and Coria, R. (1992). ‘Estudio
de los recursos naturales de la Isla Socorro, Revillagigedo’. Ciencia 43:
175–184.
Arnaud, G., Rodríguez, A., Ortega-Rubio, A. and Álvarez, C.S. (1993).
‘Predation by cats on the unique endemic lizard of Socorro Island
(Urosaurus auriculatus), Revillagigedo, México’. The Ohio Journal of
Science 93(4): 101–104.
Ortiz-Alcaraz, A., Aguirre-Muñoz, A., Méndez-Sánchez, F. and OrtegaRubio, A. (2016a). ‘Feral sheep eradication at Socorro Island, Mexico:
A mandatory step to ensure ecological restoration’. Interciencia 41:
184–189.
Arnaud, G., Rodríguez, A. and Álvarez, S. (1994). ‘El Gato Doméstico
(Felis catus), Implicaciones de su Presencia y Alternativas para su
Eradicación’. In: A. Ortega-Rubio and A. Castellanos-Vera (eds.) La
Isla Socorro, Reserva de la BiosferaArchipiélago de Revillagigedo,
México, pp. 341–359. S. C. La Paz, Mexico: Centro de Investigaciones
Biológicas del Noroeste.
Ortiz-Alcaraz, A., Maya-Delgado, Y., Cortés-Calva, P., Aguirre-Muñoz,
A., Rojas-Mayoral, E., Cordoba -Matson, M. and Ortega-Rubio, A.
(2016b). ‘Recovery of vegetation cover and soil after the removal of
sheep in Socorro Island, Mexico’. Forests 7: 91.
Bester, M.N., Bloomer, J.P., Van Aarde, R.J., Erasmus, B.H., Van
Rensburg, P.J., Skinner, J.D. Howell, P.G. and Naude, T.W. (2002). ‘A
review of the successful eradication of feral cats from sub-Antarctic
Marion Island, Southern Indian Ocean’. South African Journal of
Wildlife Research 32: 65–73.
Brattstrom, B. (2015). ‘Food webs and feeding habits on the Revillagigedo
Islands, Mexico’. Pacific Science 69: 181–195.
Parkes, J., Fisher, P. and Robinson, S. (2012). Eradication of Feral Cats
on Large Mexican Islands: A Discussion of Options and Feasibility.
Lincoln, New Zealand: Invasive Species International, Landcare
Research New Zealand.
Parkes, J., Fisher, P., Robinson, S. and Aguirre-Muñoz, A. (2014).
‘Eradication of feral cats from large islands: An assessment of the effort
required for success’. New Zealand Journal of Ecology 38: 307–314.
Brothers, N.P. (1982). ‘Feral cat control on Tasman Island’. Australian
Ranger Bulletin 2: 9.
Phillips, R.B., Cooke, B.D., Campbell, K., Carrion, V., Marquez, C. and
Snell, H.L. (2005). ‘Eradicating feral cats to protect Galápagos land
iguanas: Methods and strategies’. Pacific Conservation Biology 11:
57–66.
Campa-Molina, M.A. (1989). ‘Flora y Potencial Apícola de Isla Socorro,
Archipiélago de las Revillagigedo’. MSc thesis. Jalisco, México:
Universidad de Guadalajara.
Ramsey, D.S.L., Parkes, J.P., Will, D., Hanson, C.C. and Campbell, K.J.
(2011). ‘Quantifying the success of feral cat eradication, San Nicolas
Island, California’. New Zealand Journal of Ecology 35: 163–173.
272
Ortiz-Alcarez, et al.: Restoration of Socorro Island, Mexico
Ratcliffe, N., Bell, M., Pelembe, T., Boyle, D., Benjamin, R., White, R.,
Godley, B., Stevenson, J. and Sanders, S. (2009). ‘The eradication of
feral cats from Ascension Island and its subsequent recolonization by
seabirds’. Oryx 44: 20–29.
Rauzon, M.J. (1985). ‘Feral cats on Jarvis Island: Their effects and their
eradication’. Atoll Research Bulletin 282: 1–30.
Rauzon, M.J., Everett, W.T., Boyle, D., Bell, L. and Gilardi, J. (2008).
‘Eradication of feral cats at Wake Atoll’. Atoll Research Bulletin 560:
1–21.
Rhea, M. (2000). ‘Distribution of the Endangered Socorro Mockingbird,
Mimodes graysoni, in Relation to Habitat Availability’. University
Thesis. Villanova, Pennsylvania: Biology Department, Villanova
University.
Richards, A.F. and Brattstrom, B.H. (1959). ‘Bibliography, cartography,
discovery and exploration of the Islas Revillagigedo’. Proceedings of
the California Academy Sciences 29: 315–360.
Rodríguez, C., Torres, R. and Drummond, H. (2006). ‘Eradicating
introduced mammals from a forested tropical island’. Biological
Conservation 130: 98–105.
Rodriguez-Estrella, R., Rivera, L. and Mata, E. (1994). ‘Avifauna
Terrestre’. In: A. Ortega-Rubio and A. Castellanos-Vera (eds.). La Isla
Socorro, Reserva de la Biosfera Archipiélago de Revillagigedo, México,
pp. 199–224. La Paz, Mexico: Centro de Investigaciones Biológicas del
Noroeste S.C.
Sellés, G., Fereyra, R., Ruíz, R., Ferreyra, R. and Ahumada, R. (2012).
Compactación de Suelos y su Control: Estudio de Casos en el Valle
de Aconcagua. Boletín INIA 234. Santiago, Chile: Instituto de
Investigaciones Agropecuarias.
SEMARNAT. (2004). Programa de Conservación y Manejo, Reserva
de la Biósfera Archipiélago de Revillagigedo. México: Secretaría de
Medio Ambiente y Recursos Naturales.
Taylor, D. and Katahira, L. (1988). ‘Radiotelemetry as an aid in
eradicating remnant feral goats’. Wildlife Society Bulletin 16: 297–299.
Tortora, D.F. (1982). ‘Understanding electronic dog-training. 1’. Canine
Practice 9: 17–22.
Twyford, K.L., Humphrey, P.G., Nunn, R.P. and Willoughby, L. (2000).
‘Eradication of feral cats (Felis catus) from Gabo Island, south-east
Victoria’. Ecological Management and Restoration 1: 42–49.
Van Vuren, D. (1992). ‘Eradication of Feral Goats and Sheep from Island
Ecosystems’. In: J.E. Borrecco and R.E. Marsh (eds.) Proceedings of the
15th Vertebrate Pest Conference, pp. 377–381. Davis, USA: University
of California.
Veitch, C.R. (1989). The Eradication of Cats and Sheep from Socorro
Island: A Report to the Socorro Island Project. Northern Region
Technical Report Series No. 11. Auckland, New Zealand: Department
of Conservation.
Veitch, C.R. (2001). ‘The eradication of feral cats (Felis catus) from
Little Barrier Island, New Zealand’. New Zealand Journal of Zoology
28: 1–12.
Villa, B. (1960). ‘La Isla Socorro. Vertebrados terrestres’. Monografías
Instituto de Geofísica 2: 203–206.
Walter, H.S. and Levin, G.A. (2008). ‘Feral sheep on Socorro Island:
Facilitators of alien plant colonization and ecosystem decay’. Diversity
and Distributions 14: 422–431.
Will, D., Hanson, C.C., Campbell, K.J. Garcelon, D.K. and Keitt, B.S.
(2010). ‘A Trap Monitoring System to Enhance Efficiency of Feral
Cat Eradication and Minimize Adverse Effects on Non-target Endemic
Species on San Nicolas Island’. In: R. M. Timm and K. A. Fagerstone
(eds.) Proceedings of the 24th Vertebrate Pest Conference, pp. 79–85.
Davis, CA: University of California.
Wood, B., Tershy, B.R., Hermosillo, M.A., Donlan, C.J., Sánchez, J.A.,
Keitt, B.S., Croll, D.A., Howard, G.R. and Biavaschi. N. (2002).
‘Removing cats from islands in North-west Mexico’. In: C.R. Veitch
and M.N. Clout (eds.) Turning the tide: the eradication of invasive
species, pp. 374–380. Occasional Paper SSC no. 28. IUCN SSC
Invasive Species Specialist Group. IUCN, Gland, Switzerland and
Cambridge, UK.
273
Thompson, R.C. and J.M. Ferguson
Thompson, R.C. and J.M. Ferguson. Removing introduced hedgehogs from the Uists
Removing introduced hedgehogs from the Uists
R.C. Thompson and J.M. Ferguson
Scottish Natural Heritage, Stilligarry, South Uist, HS8 5RS. <Charles.Thompson@nature.scot>.
Abstract Hedgehogs (Erinaceus europaeus) are native to Great Britain but were introduced to the island of South Uist
in 1974 and gradually colonised South Uist and Benbecula. In 1999 hedgehogs were confirmed in southern areas of North
Uist. Hedgehogs eat the eggs and occasionally the chicks of waders, which breed at high densities in the Uists. Initial
research by Royal Society for the Protection of Birds (RSPB) in 1998 suggested that predation by hedgehogs was having
a significant effect on the wader populations in South Uist. In 2014, remote cameras were used on a sample of wader nests
and found hedgehogs responsible for 52% of all predation in South Uist. The Uist Wader Project was set up in 2000 to
remove hedgehogs from North Uist initially, but with a long-term aim to remove hedgehogs completely from the Uists.
Various methods including lamping, trapping and the use of sniffer dogs were developed, trialled, and improved. We
developed an Index of Abundance (IOA) of hedgehogs, using footprint monitoring tunnels. This IOA provides a means
of confirming the impact of removal activities on the hedgehog population. In anticipation of scaling up, we carried
out a removal trial on a two km² area at Drimore in South Uist. The trial demonstrated the effort required to reduce the
abundance of hedgehogs from high density, 30 animals/km2, to zero and enabled the project team to estimate the resources
required to eradicate hedgehogs from Benbecula, North and South Uist. The North Uist phase should be complete by the
beginning of 2018, with only eight hedgehogs caught in 2016 and just one in 2017. Two years of monitoring are planned
between 2018 and 2020, to confirm eradication.
Keywords: dunlin, IOA, redshank, ringed plover, sniffer-dogs, translocation, trap, wader
INTRODUCTION
Wader surveys in the early 1980s showed that the
Uists, off the west coast of Scotland, held high densities
of breeding redshank (Tringa totanus), ringed plover
(Charadrius hiaticula) and dunlin (Calidris alpina)
(Fuller, et al., 1986). In recognition of the importance of
the Uists, 14 Sites of Special Scientific Interest (SSSIs)
and two Special Protection Areas (SPAs) for Birds were
designated in the late 1990s. Shortly afterwards a decline
was found in wader populations on the islands of South
Uist and Benbecula that was largely due to egg predation
by hedgehogs (Jackson, 2001; Jackson & Green, 2000;
Jackson, et al., 2004). Hedgehogs (Erinaceus europaeus)
are native to Great Britain but were introduced to South
Uist in 1974–75 (Angus, 1993). In 1999, hedgehogs were
starting to colonise southern areas of North Uist (Jackson
& Green, 2000; Jackson, et al., 2004). Declines of waders
recorded in South Uist between 1983 and 1998 were:
ringed plover by -58%; dunlin by -65%; and redshank by
-43% (Fuller & Jackson, 1999). In 2014, remote cameras
were used on a sample of wader nests and found hedgehogs
responsible for 52% of all predation in South Uist.
MATERIALS AND METHODS
Study area
The Uists are part of the Outer Hebrides, located off the
north-west coast of Scotland (Fig. 1). The Uists include six
inhabited, low-lying islands, connected by causeways. The
three main islands are North Uist (333 km²), Benbecula
(81 km²) and South Uist (315 km²). The climate is wet
and windy. Wind-blown shell sand has formed extensive
machair habitats on the west side of these islands. These
lime-rich coastal grasslands are grazed by livestock and
cultivated with arable crops (oats, rye, barley and potatoes)
on a traditional rotation (Angus, 2006). There are a few
farms but most of the agricultural land is divided into small
tenanted units, known as crofts, each with shares in larger
common grazings. The other predominant habitat types in
the Uists are moorland and blackland (an intermediate zone
of mesotrophic grassland between machair and moorland).
The hedgehog population on the United Kingdom
mainland has been in decline since the1960s (Noble, et
al., 2012). Hedgehogs are protected under Schedule 6 of
the Wildlife and Countryside Act 1981 throughout the UK,
but are classified as invasive non-natives in the Uists, as
they are classified outwith their native range under section
14. Hedgehogs have no natural predators in the Uists, can
breed five months out of the year, and can produce at least
as many young per year as their population, as measured
in the spring (Jackson, 2007). Initial research in South
Uist on hedgehog behaviour and methods of locating them
was carried out between 1997 and 2001 (Jackson, 2007).
This work estimated the density of hedgehogs in different
habitats in South Uist at 31.8 animals/km2 for machair,
15.4 animals/km2 for blackland and two animals/km2 for
moorland.
The Uist Wader Project was launched in 2000 as a
partnership of Scottish Natural Heritage (SNH), Royal
Society for the Protection of Birds (RSPB) and the Scottish
Executive. The Project’s objective was to safeguard the
waders of the Uists from introduced hedgehogs. In order
to achieve this it would be necessary to remove all the
hedgehogs from the Uists, starting in North Uist.
Fig. 1 Location map of the Uists in conjunction with
mainland Scotland.
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
274
up to meet the challenge, pp. 274–281. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Thompson & Ferguson: Removing hedgehogs from the Uists
In order to carry out any type of fieldwork in the
Uists it was essential to have the full support of the local
community, land owners, crofters and residents. Project
staff spent time working with these groups to secure access
to land and receive information relating to hedgehogs.
Although access permission was always granted, there
were constraints placed on some of the removal methods
described below.
Population model
Based on initial research by Jackson (2007), modellers
at Newcastle University developed a hedgehog population
model in two phases (Shirley, et al., pers. comm. 2007;
Shirley, et al., 2010). These individual-based, simulation
models of the hedgehog population indicated that trapping
and lamping would achieve total eradication of hedgehogs
from the Uists within 30 years, at best, which represents
eradication by 2040. The hedgehog population in the
Uists was estimated from a combination of field data and
the model to be around 3,900 in 2010, whereas this was
estimated to be about 3000 in 2007, (95% confidence limits
±800). This highlighted the shortcomings of our initial
methods and led to a new approach, using sniffer dogs and
all year round removal of hedgehogs.
Hedgehog removal
There are three key methods of removing hedgehogs;
lamping (spot-lighting), live cage trapping and searching
with sniffer dogs. Lamping at night was very effective on
short cropped machair turf. It involves three to five people
transecting areas of land operating in a straight line about
five to 10 metres apart, each using a 10–50 watt, 12 volt
halogen spot-lamp to survey the ground for hedgehogs.
Lamping was not effective in longer vegetation and the
night work caused disturbance to local residents. Lamping
as a method of hedgehog removal was gradually phased
out due to its intrusive nature with regard to light and noise
disturbance at night.
Live cage trapping was tried in 2004 and proved very
effective at removing a large proportion of the population:
80–90% of hedgehogs over an eight week period. Trapping
worked well in all types of habitat and replaced lamping
as the main method of removing hedgehogs. The live cage
traps used for hedgehogs are 180 × 150 × 480 mm, with a
spring-loaded door, activated by a treadle plate. Traps are
installed in large trap grids, designed to intersect the home
range of each potential hedgehog. Two different trapping
densities are used. Low density trapping (30 traps/km2) is
used when initially establishing a trapping route, where the
underlying hedgehog population is expected to be zero (i.e.
monitoring suggests no hedgehogs) or when the underlying
habitat is not particularly suitable for hedgehogs, such
as moorland and bog. Higher density trapping (50 traps/
km2) is used where a known hedgehog population exists
and the underlying habitat is suitable. The traps are baited
with fish, which is placed behind the treadle plate, but
not obstructing it. Once operational, traps are checked
every day. Throughout the project, trap placement was
continuously improved through experience and research
with habitat, location, cover, bait and trap sensitivity
being the most important factors. Trapping proved to be
an effective means of capturing a large proportion of a
population, but not every animal, suggesting that some
were trap shy.
Sniffer dogs are also used to remove hedgehogs. The
dogs are trained to indicate the location of a hedgehog
without harming it and are rewarded with a short period of
play time with a favourite toy when successful. A specialist
trainer was brought in for six days each year to guide the
training process, encourage best practice and work with
each dog handler on a one to one basis. Sniffer dogs can
work effectively for periods of three to four hours and
an experienced sniffer dog and handler can cover up to
two km2 per week in most weather conditions and across
diverse vegetation. Dense vegetation and calm conditions
result in narrower, more condensed search transects, while
wind speeds between eight and 55 kph and short vegetation
allow wider more expansive transects and hence greater
area covered per unit of time. Wind speeds in excess of 55
kph progressively reduce the efficiency of dog searching
due to the scent being dispersed too widely. Sniffer dogs
and trapping complement each other as hedgehog removal
methods, because dogs are more effective in boggy ground
where traps simply can’t be set and traps are more effective
in areas where dens are located deep underground and
hedgehogs only re-emerge at night. There were sometimes
more restrictions on using dogs than traps in fields at
lambing time but sniffer dogs could locate hedgehogs
during the winter, when trapping is ineffective. The use
of dogs was suspended, early on in the project, following
the introduction of legislation banning the hunting of wild
mammals with dogs, The Protection of Wild Mammals
(Scotland) Act 2002. This greatly reduced the efficiency
of removing hedgehogs at lower densities and added
additional time and cost to the Project. Following careful
legal interpretation of how dogs could be used to locate and
‘flush’ hedgehogs, the use of sniffer dogs was reinstated in
2010.
Between 2003 and 2006 all captured hedgehogs were
euthanised, based on the best information available at that
time. Advice from the animal welfare organisation, the
Scottish Society for the Prevention of Cruelty to Animals
(SSPCA), rejected translocation on welfare grounds and
advocated hedgehogs were euthanised. The SNH board
in 2002 stated that there was no scientific evidence or
overriding conservation imperative to justify translocation
of hedgehogs from the Uists to the mainland. During this
time, the Project came under increasing pressure from
animal rights groups and special interest conservation
groups to stop killing hedgehogs and consider moving
them to the Scottish mainland instead. The British public
perceives hedgehogs as an iconic species, which is the
gardener’s friend, and there was strong media and public
pressure against the cull.
New research carried out at Bristol University (Molony,
et al., 2006) showed that translocation of hedgehogs
resulted in low mortality if certain levels of veterinary
care, feeding and general welfare were provided. Based on
this work, the SSPCA advised that the hedgehogs’ welfare
would not be adversely affected by being translocated to
the Scottish mainland. SNH then entered into a partnership
with the animal care sector to translocate hedgehogs.
Fieldworkers pass hedgehogs onto a ‘carer’, based in
South Uist, for onward transport to an animal rescue centre
on the mainland for release under established protocols.
In response to improvements in the ability to identify and
care for pregnant females and to locate dependant young, it
became possible in 2012 to remove hedgehogs throughout
the season, rather than only during the non-breeding season
of three and a half months as done previously.
Monitoring
Monitoring between 2009 and 2010 simply involved
checking traps and lamping, which equates to extending
the removal methods until a period of two years has elapsed
where no capture of hedgehog has occurred.
From 2011 onwards, three monitoring techniques were
deployed: footprint monitoring tunnels, sniffer dogs and
motion-activated cameras.
275
Island invasives: scaling up to meet the challenge. Ch 2A Other taxa: Mammals
The footprint monitoring tunnels were made out of
150 mm plastic drainage pipe, cut to 560 mm lengths.
A rectangular section 100 × 190 mm was cut out of the
middle of the pipe to accommodate a plastic tray, 110 ×
50 × 200 mm. The tray was then filled with one of three
different substrates; clay, sand or carbon plate. The tunnels
were dug into the ground and covered with turf to make
the tunnel as much like a natural burrow as possible. The
inside of the tunnel was fashioned to allow a natural walk
through for an animal over the tray. These tunnels were dug
into the monitoring area at a density of five tunnels/km²
and their positions recorded using GPS.
Trained sniffer dogs were deployed to search at least
25% of the monitoring area following methods similar
to their use for hedgehog removal. Hedgehogs located
in North Uist were removed as re-release was not an
option, whereas hedgehogs located in South Uist were
released, since their removal would have no real impact
on the overall population, which had reached its maximum
carrying capacity.
Motion-activated cameras (model: Bushnell Trophycam
HD max) were deployed at a density of 1.25 cameras/km².
We set them to record 60 second video clips (1280 × 720
px) onto a 32 GB SD card. The camera was focussed on a
120 g ‘tuna tin’ with perforations in its top, filled with fish
and dug into the ground so the surface of the can was level
with the ground. This acts as an attractant to hedgehogs and
a host of other animals, yet prevents them from removing
the fish. The SD card needs to be changed every two weeks
and the rechargeable batteries have a variable lifespan, of
two to three weeks, depending on the rate of triggering.
Six sample areas representing the whole of North Uist
were monitored between 2013 and 2014 using at least two
different monitoring methods. Monitoring highlighted the
areas where hedgehogs were present and allowed a more
strategic and selective approach to checking the total area
of North Uist.
Occupancy model
In the early part of the Project, progress was measured
as ‘number of hedgehogs caught per 1000 trap nights’.
When trapping effort was applied over time, this measure
generally showed a decline. However, we were unsure
if this measure reflected the actual impact of removal
activities on the hedgehog population, or if a significant
number of animals remained undetected due to trap
avoidance. In 2013, a two-year monitoring trial was
established to estimate occupancy and the relative index
of abundance (IOA) of hedgehogs across the Uists, and
evaluate the effectiveness of the removal methods.
Between 2013 and 2014, hedgehog populations were
assessed in 19 locations in the key areas for breeding
waders, using footprint monitoring tunnels. Attempts were
also made to assess populations using motion-activated
cameras and sniffer dogs, but insufficient cameras were
available and the sniffer dog data proved too difficult to
interpret due to a number of factors including experience
of dog and handler, wind speed, and topology of land.
Each plot (route) covered an area of four km² with a
minimum of five monitoring tunnels/km² and was checked
twice per week.
Various occupancy models were tested, and the RoyleNichols single season, abundance-induced heterogeneity
model (Royle & Nichols, 2003) was chosen as the most
appropriate single season occupancy model. This is a
two-parameter model that derives occupancy (ψ) from
estimates of detectability r (the probability of detection per
tunnel) and population density λ (the mean of the Poisson
276
distribution), thus estimating occupancy in a way that
accounts for hedgehogs being easier to detect when there
are more of them. The following formulas represent the
Royle-Nichols model:
Formula (1) represents the likelihood of detections,
where W represents detections, R represents sites, T
represents (route) locations. Formula (2) represents the
site detection probability and (3) represents the probability
density formula for a Poisson distribution, where both
(2) and (3) substitute into (1). Note also how r and λ are
incorporated into this model.
The plots were grouped together by year and modelled
with a constrained detectability and unconstrained
population density. Detectability was estimated, along with
individual population density, for each location and year.
Footprint monitoring results were used in preference to
camera monitoring results due to the limited data sample
from the cameras compared to tunnels (Paul Ross, pers.
comm. 2014).
Hedgehog removal trial
In 2014, we undertook a hedgehog removal trial to
evaluate the effectiveness of the hedgehog removal methods.
A research area of 1.78 km² was selected on Drimore farm
in South Uist, which represented typical machair habitat
with a probable high population of hedgehogs. A perimeter
area of 1.3 km² surrounding this research area was also
created to reduce the effects of dispersion and migration of
hedgehogs following removal from the research area.
The research area was monitored using footprint
monitoring tunnels, motion-activated cameras and sniffer
dogs for a four week period to establish an IOA. The
monitoring tunnels were evenly distributed at a density of
five tunnels/km², motion-activated cameras at a density of
five cameras/km² and sniffer dogs were operated at a rate
of two km² per week. Hedgehogs were then removed from
both areas using 50 traps/km² on the research area only, and
sniffer dogs on both areas, for an eight week period. The
research area was monitored for a further four weeks in the
same design as the pre-removal monitoring, to establish
whether all hedgehogs had been removed.
Scaling up to North Uist
Recent hedgehog removal efforts in North Uist were
guided by the results of the monitoring work; areas that
showed presence of hedgehogs were searched using a
combination of trapping and sniffer dogs. The search effort
was set using the results of the Drimore trial.
The removal phase is expected to be completed by
spring 2018, and will be followed by a further two years
of monitoring to confirm absence of hedgehogs. If a
hedgehog is encountered during the monitoring phase a
rapid-reaction protocol will be initiated.
Rapid-reaction protocol
A one km radius buffer around the sighting of a
hedgehog will be searched for four weeks with sniffer
dogs and 50 traps/km². If further hedgehogs are found, this
process will be repeated.
Thompson & Ferguson: Removing hedgehogs from the Uists
RESULTS
Activites implemented in 2003 to 2008
Sniffer dogs were used only in 2003 during this period.
Lamping and trapping were used as the main methods of
hedgehog removal. Monitoring between 2009 and 2010
confirmed successful eradication.
Initial hedgehog removal: 2003–2008
Hedgehog removal started in 2003 in Locheport and
Carinish in the southern area of North Uist (129 km²) and
was completed by 2008. A further two years of monitoring
were carried out to verify a successful eradication, which
was declared in 2010. Fig. 2 shows the removal of
hedgehogs and effort applied in Carinish and Locheport.
Believing that North Uist was clear, the Project expanded
the removal methods into Benbecula to continue working
southwards. Good progress was made initially, but further
hedgehogs were reported from new areas of North Uist;
from Balranald in 2009 and Lochportain in 2012. Work in
Benbecula was postponed whilst the trapping team was redeployed to eradicate hedgehogs from these new areas.
Monitoring results – occupancy estimates 2013–2014
As expected, the lowest occupancy (ψ) estimates were
in North Uist and the highest ones in South Uist.
The North Uist IOA monitoring results for 2013 and
2014 are shown in Table 1 and Table 2, respectively. Note
that the route names do not correspond to the same areas
between the two years. In 2013, Baleshare (represented
by H1 & H2) showed no occupancy of hedgehogs, and
Balranald (F1) showed a low level of occupancy. In 2014,
Balranald (F1, M1, andG2) showed further dispersal of
hedgehogs.
The South Uist and Benbecula occupancy results
for 2013 and 2014 are shown in Table 3 and Table 4,
respectively. Note that the routes K1 and B2 correspond
between Tables 3 and 4. Fig. 3 shows these results spatially.
The occupancy estimate of hedgehogs is relatively high for
almost all parts sampled in South Uist and Benbecula.
All other parts of North Uist, Benbecula and South
Uist were monitored by sniffer dog but due to a range of
confounding factors it proved impossible to convert these
data into a meaningful occupancy estimate. However, the
sniffer dog monitoring did give a good overview of the
distribution of hedgehogs across the Uists to complement
the formal occupancy estimate results.
Fig. 3 Hedgehogs removed or detected, by method, for the
Drimore trial.
Removal trial results (Drimore, 2014)
Pre-removal monitoring phase
Monitoring was carried out across the research area for
four weeks between 7 April and 4 May. Table 5 shows the
numbers of hedgehogs detected each week by footprint
monitoring, camera monitoring and sniffer dogs. Using
only the footprint monitoring data, it was possible to
derive an occupancy estimate for this area of land during
the four week monitoring phase, which is shown in row
B2 in Table 3. In comparison to other sites monitored in
the Uists, the Drimore site represented a high population
of hedgehogs.
Removal of hedgehogs
This phase of operation involved removing hedgehogs
from the research and perimeter areas using live cage traps
and sniffer dogs over an eight week period between 5 May
and 29 June. Two fieldworkers searched the area using
sniffer dogs and operated 89 live cage traps. Table 6 shows
that the same numbers of hedgehogs were removed from
the research area by sniffer dogs as by trapping. Over the
same period, hedgehogs were removed from the perimeter
area by sniffer dogs alone, as shown in Table 7.
Post-removal monitoring
The final phase of the trial involved repeating the
monitoring over another four week period between 30
June and 28 July to measure the IOA of the hedgehogs
after the removal operation. Table 8 shows that only
two hedgehogs showed up on camera during this period,
both in the perimeter area. No hedgehogs were detected
within the research area, providing an acceptable level of
confidence that all of the hedgehogs had been removed.
Fig. 3 summarises the numbers of hedgehogs detected by
each monitoring method at all three stages.
Recent hedgehog removals in North Uist – scaling up
to North Uist
Lochportain hedgehog removal
Fig. 2 Carinish and Locheport hedgehog removal and
effort.
Lochportain, along with the neighbouring townships,
is located on a peninsula on the east side of North Uist. In
2012 a hedgehog was found by a member of the public on
the road close to Cheesebay (adjacent to Lochportain). Due
to other commitments and limitations on staff resources the
Project was only able to respond to this potential hedgehog
population with a very limited removal effort in 2013,
which yielded no hedgehog captures. The team returned to
this area in 2015, with a concerted removal effort covering
some 75 km2, followed by monitoring in 2016. Fig. 4
shows the hedgehog removal and relative effort appliedin
Lochportain.
277
Island invasives: scaling up to meet the challenge. Ch 2A Other taxa: Mammals
Table 1 2013 – Royle-Nichols parameter estimates for hedgehogs in North Uist. Note: ψ represents the probability
of occupancy, derived from (1 – fk), r represents the probability of detection per hedgehog / tunnel, and l
represents population density as the mean of the Poisson distribution. Route name refers to four km² plot areas.
Route
name
F1
F2
G1
G2
H1
H2
Naïve
occupancy
0.046
0.000
0.000
0.000
0.000
0.000
Occupancy
SE
ψ
0.073
0.322
0.000
0.000
0.000
0.000
0.000
-
Detectability
r
SE
0.142
0.142
0.142
0.142
0.142
0.142
0.027
0.027
0.027
0.027
0.027
0.027
Population density
SE
λ
0.076
0.077
0.000
0.000
0.000
0.000
0.000
-
Table 2 2014 – Royle-Nichols parameter estimates for hedgehogs in North Uist.
Route
name
F1
M1
G2
J1
Naïve
occupancy
0.100
0.025
0.075
0.000
Occupancy
SE
ψ
0.154
0.294
0.041
0.333
0.124
0.304
0.000
-
Detectability
r
SE
0.142
0.142
0.142
0.142
0.027
0.027
0.027
0.027
Population density
SE
λ
0.167
0.086
0.042
0.042
0.133
0.079
0.000
-
Table 3 2013 – Royle-Nichols parameter estimates for hedgehogs in South Uist and Benbecula.
Route
name
A1
A2
B1
B2
C1
C2
D1
D2
E1
E2
K1
B2
Naïve
occupancy
0.286
0.400
0.130
0.400
0.286
0.100
0.000
0.250
0.300
0.100
0.050
0.400
Occupancy
SE
ψ
0.582
0.145
0.531
0.163
0.241
0.264
0.632
0.128
0.405
0.206
0.182
0.284
0.000
0.458
0.188
0.446
0.192
0.171
0.288
0.080
0.319
0.628
0.129
Detectability
r
SE
0.142
0.142
0.142
0.142
0.142
0.142
0.142
0.142
0.142
0.142
0.142
0.142
0.027
0.027
0.027
0.027
0.027
0.027
0.027
0.027
0.027
0.027
0.027
0.027
Population density
SE
λ
0.873
0.347
0.758
0.288
0.275
0.164
1.000
0.366
0.519
0.223
0.201
0.145
0.000
0.613
0.273
0.591
0.255
0.188
0.135
0.084
0.085
0.988
0.332
Table 4 2014 – Royle-Nichols parameter estimates for hedgehogs at Drimore in South Uist and Benbecula.
Route
name
K1
B2
Naïve
occupancy
0.050
0.000
Occupancy
SE
ψ
0.105
0.311
0.000
-
Lochportain was effectively cleared of hedgehogs over
a 10 week period, which matched very closely with the
Drimore removal trial. Migration to and from Lochportain
was minimised by being located on a peninsula with a
narrow isthmus.
Fig. 5 compares the Drimore trial and Lochportain
removal. There is a strong similarity in the pattern of
hedgehog removal even though the starting populations of
hedgehogs and the area of land covered are very different.
Both locations represent declining sequences of weekly
captures ending at one or less over eight to 10 weeks.
Subsequent monitoring on both sites demonstrated that no
further hedgehogs were immediately present.
278
Detectability
r
SE
0.142
0.142
0.027
0.027
Population density
SE
λ
0.110
0.111
0.000
-
Balranald & Paible hedgehog removal
In 2009 hedgehogs were sighted in Balranald and
Paible in the west of North Uist by members of the public.
Trapping began in 2009, and sniffer dogs were introduced
gradually from 2010, so that by 2013 all fieldwork staff
operated a dog.
Fig. 6 shows that the bulk of the hedgehog population
was removed between 2013 and 2015, with just a small
number of hedgehogs removed in 2016. It also shows the
relationship between trapping effort and the number of
hedgehogs removed for Balranald and Paible.
Thompson & Ferguson: Removing hedgehogs from the Uists
Implementation of rapid-reaction protocol
The rapid-reaction protocol has been used only once.
One hedgehog was located by a monitoring camera and
then located and removed by a dog handler and sniffer dog
in the area of east Balranald during April 2017. A search
zone was established using a buffer of a radius of 1 km
from the location of the hedgehog, as shown in Fig. 7.
Four weeks searching using sniffer dogs and trap checks
were carried out, but no further hedgehogs were located.
DISCUSSION
It is essential to have the support of the local community,
not just to report sightings but also to persuade people
not to move hedgehogs to new areas. Hedgehogs were
clearly moved to discrete unconnected areas in North Uist,
including Carinish, Locheport, Balranald and Lochportain.
We had support from most land managers but we failed to
reach all individuals within the wider community. Some
people moved hedgehogs as they thought they would
provide a helpful service such as controlling garden
slugs or snails that host sheep fluke. Once we were able
to discuss these introductions and the potential impacts
with the individuals involved, they usually became more
supportive. Any future removal project should include an
education and promotion resource to assist with community
engagement. There is also a need to secure full support and
commitment right from the start of the project all the way
through until eradication is confirmed.
Fig. 4 Lochportain hedgehog removal and effort.
Fig. 5 Comparison of hedgehogs removed between
Lochportain and Drimore.
Fig. 6 Balranald and Paible removal and effort.
Table 5 Hedgehogs detected during the pre-removal
monitoring phase at Drimore. Monitoring effort: two
footprint monitoring checks per week per tunnel over
10 tunnels, five cameras running continuously and two
sniffer dogs checking two km² per week.
Monitoring
method
Footprint
Camera
Sniffer dog
1
6
10
16
Week
2
3
6
7
19
19
13
23
4
3
18
18
Total
22
66
70
Fig. 7 Rapid-reaction protocol in response to hedgehog
capture at Balranald.
279
Island invasives: scaling up to meet the challenge. Ch 2A Other taxa: Mammals
Table 6 Hedgehogs removed from research area at Drimore during removal phase.
Removal method
Trapping
Sniffer dog
Total hedgehogs
Effort: trapping (hrs)
Effort: dog (hrs)
1
22
17
39
44
24
2
3
8
11
43
17
3
0
0
0
44
1
Week
4
5
3
1
3
0
6
1
44
41
42
47
6
1
3
4
44
40
7
2
0
2
44
37
8
0
1
1
44
39
Total
32
32
64
348
247
Table 7 Hedgehogs removed from perimeter area at Drimore during removal phase.
Removal method
Trapping
Sniffer dog
Total hedgehogs
Effort: trapping (hrs)
Effort: dog (hrs)
1
0
0
0
2
2
2
2
3
12
12
14
The methods used in eradicating hedgehogs from
Carinish and Locheport were limited by the absence of
sniffer dogs and a lack of clarity on the abundance of
hedgehogs in any given area. For animal welfare reasons,
hedgehog removal was restricted to the three and a half
month non-reproductive period. These limitations meant it
took approximately eight years to clear the area and verify
it as clear. Balranald and Paible were also initially limited
to the non-reproductive season and sporadic, exploratory
efforts prior to 2013. However from 2013 onwards Balranald
and Paible had a fully operational team of sniffer dogs and
hedgehog removal progressed relatively quickly, with
captures tailing off by 2016. Lochportain also benefitted
from the use of dogs and from being on a peninsula. The
introduction of monitoring, refined control methods and
strategies meant that removing the Lochportain hedgehog
population took just two years, compared to eight at
Carinish. If there are obstacles or barriers to removal
activities then it will reduce the effectiveness of removal
and it will take longer to reduce the population to zero.
Being able to work all year round made the Project much
more efficient, reducing the predicted minimum time
required for eradication of hedgehogs from the Uists from
30 to five years.
The Drimore trial demonstrated that hedgehog
population density within a discrete area can be effectively
reduced to zero by trapping and sniffer dogs over a relatively
short period of time. The removal phase reduced the IOA
from a high level to zero. The two hedgehogs detected on
camera in the latter weeks of the post-removal monitoring
were located in the perimeter area and it is assumed
these were migrating into the research area. Comparing
the Drimore trial results to the Lochportain eradication
shows that it took roughly the same effort to remove 64
Table 8 Hedgehogs monitored during the post-removal
monitoring phase at Drimore.
Monitoring
method
Footprint
Camera
Sniffer dog
280
Week
1
0
0
0
2
0
0
0
3
0
1
0
4
0
1
0
Total
0
2
0
Week
4
4
4
2
5
1
1
2
6
1
1
6
7
1
1
3
8
0
0
3
Total
21
21
32
hedgehogs as it did 14 hedgehogs from an equivalent area.
This suggests that eradication effort is determined by area
of suitable habitat more than hedgehog density.
The Project needed to estimate the effort required to
reduce the hedgehog population to zero over a given area
of land and prevent re-colonisation from surrounding
areas. The Drimore removal trial enabled us to assess
whether the resource had been sufficient on every bit of
land at Balranald and where to put in additional resource.
The near complete removal of hedgehogs from North
Uist was achieved using an agreed strategy with proven
methods of removal, which were shown to be effective.
Being able to measure the effectiveness of the hedgehog
removal methods used, and the effort required to clear a
given area of land, enables a fairly accurate estimation of
what timescale would be required to clear a specific area of
land. There also needs to be a method of confirming that
the population has been reduced to zero (Russell, et al.,
2016). The IOA has been extremely valuable in that respect,
particularly on areas such as Balranald, with complicated
land tenure and constraints on using dogs whilst livestock
are in fields at certain times of year.
In the early days of the Project we coloured maps
by hand and filled in paper data sheets, whereas now we
use graphic GPS, integrated to GIS systems, connected
to relational databases. This increased data flow has
facilitated a more adaptive approach to managing project
activity. Scientific advice from a wide range of sources
has been extremely helpful but needs to be combined with
practical considerations.
Ideally it would have been desirable to have cleared the
hedgehogs from South Uist to allow the waders to recover
faster, but clearing North Uist first and then moving
south made more strategic sense. Having successfully
removed all of the hedgehogs from North Uist, the next
step is to continue southwards and remove hedgehogs from
Benbecula and South Uist. This will require clearing an
area of almost 400 km². Using the results from the Drimore
trial and the current removal methods, we estimated that
this will take between five and 10 years and will require a
team of 18 staff. It is estimated that this will cost between
£3.5 and £5.0 million and, at the time of writing, SNH is
exploring funding options with partners.
Thompson & Ferguson: Removing hedgehogs from the Uists
ACKNOWLEDGEMENTS
We would like to acknowledge the pioneering wader
survey and research work carried out by RSPB and
BTO which provided the initial evidence for this project
and for the ongoing support and involvement from both
organisations, including financial assistance from RSPB.
SNH took the lead in the project, and we would like to thank
the many staff who were involved, including: the rest of
the UWP project team, fieldworkers, graduate placements,
volunteers, and the Scientific Advisory Committee. Special
thanks are due to the land owners, crofters and residents of
the Uists, who played a major role in supporting this project
by granting access permission, taking part in fieldwork,
and reporting sightings of hedgehogs.
REFERENCES
Angus, S. (1993). ‘Hedgehogs in the Western Isles in 1991’. Hebridean
Naturalist 11: 73–76.
Angus, S. (2006). ‘De tha machair? Towards a machair definition’. Sand
dune machair 4, pp. 7–22. Aberdeen Institute for Coastal Science &
Management, Aberdeen.
Fuller, R.J., Reed, T.M., Buxton, A., Williams, T.D. and Pienkowski,
M.W. (1986). ‘Populations of breeding waders (Charadrii) and
their habitats on the crofting lands of the Outer Hebrides, Scotland’.
Biological Conservation 37: 333–361.
Fuller, R.J. and Jackson, D.B. (1999). ‘Changes in populations of
breeding waders on the machair of North Uist, Scotland, 1983–1998’.
Wader Study Group Bulletin 90: 47–55.
Jackson, D.B. and Green, R.E. (2000). ‘The importance of the introduced
hedgehog (Erinaceus europaeus) as a predator of the eggs of waders
(Charadrii) on the machair in South Uist, Scotland’. Biological
Conservation 93: 333–348.
Jackson, D.B. (2001). ‘Experimental removal of introduced hedgehogs
improves wader nest success’. Journal of Applied Ecology 38: 802–812.
Jackson, D.B., Fuller, R.J. and Campbell, S.T. (2004). ‘Long-term
population changes among breeding shorebirds in the Outer Hebrides,
Scotland, in relation to introduced hedgehogs’. Biological Conservation
117: 151–166.
Jackson, D.B. (2007). ‘Factors affecting the abundance of introduced
hedgehogs (Erinaceus europaeus) on the Hebridean island of South Uist
in the absence of natural predators, and the implications for groundnesting birds’. Journal of Zoology 271: 210–217.
Molony, S.E., Dowding, C.V., Baker, P.J., Cuthill, I.C. and Harris, S.
(2006). ‘The effect of translocation and temporary captivity on wildlife
rehabilitation success: an experimental study using European hedgehogs
(Erinaceus europaeus)’. Biological Conservation 130: 530–537.
Noble, D.G., Aebischer, N.J., Newson, S.E., Ewald, J.A. and Dadam,
D. (2012). A Comparison of Trends and Geographical Variation in
Mammal Abundance in the Breeding Bird Survey and the National
Gamebag Census. JNCC Report No. 468.
Royle, J.A. and Nichols, J.D. (2003). ‘Estimating abundance from
repeated presence-absence data or point counts’. Ecology 84: 777–790.
Russell, J.C., Binnie, H.R., Oh, J., Anderson, D.P., Samaniego-Herrera,
A. (2016) Optimizing confirmation of invasive species eradication with
rapid eradication assessment. Applied Ecology 54(1): 160–169.
Shirley, M., Lurz, P. and Rushton, S. (2010). Development of a Population
Model for the Management of Hedgehogs on the Uists and Benbecula.
Newcastle University, SNH Commissioned Report: 372
281
N. Bunbury, P. Haverson, N. Page, J. Agricole, G. Angell, P. Banville, A. Constance, J. Friedlander, L. Leite, T. Mahoune, E. Melton-Durup, J. Moumou, K. Raines, J. van de
Crommenacker and F. Fleischer-Dogley
Bunbury, N.; P. Haverson, N. Page, J. Agricole, G. Angell, P. Banville, A. Constance, J. Friedlander, L. Leite, T. Mahoune, E. Melton-Durup, J. Moumou, K. Raines, J. van de
Crommenacker and F. Fleischer-Dogley. Five eradications, three species, three islands: overview, insights and recommendations from invasive bird eradications in the Seychelles
Five eradications, three species, three islands: overview, insights and
recommendations from invasive bird eradications in the Seychelles
N. Bunbury, P. Haverson, N. Page, J. Agricole, G. Angell, P. Banville, A. Constance, J. Friedlander, L. Leite,
T. Mahoune, E. Melton-Durup, J. Moumou, K. Raines, J. van de Crommenacker and F. Fleischer-Dogley
Seychelles Islands Foundation, La Ciotat Building, Mont Fleuri, P.O. Box 853, Victoria, Mahé, Seychelles.
<nancy@sif.sc>.
Abstract Management and eradication techniques for invasive alien birds remain in their infancy compared to invasive
mammal control methods, and there are still relatively few examples of successful avian eradications. Since 2011, five
separate eradication programmes for invasive birds have been conducted on three islands by the Seychelles Islands
Foundation (SIF). Target species were prioritised according to their threat level to the native biodiversity of the UNESCO
World Heritage Sites of the Seychelles, Aldabra Atoll and Vallée de Mai, which SIF is responsible for managing and
protecting. Red-whiskered bulbuls (Pycnonotus jocosus) and Madagascar fodies (Foudia madagascariensis) occurred
on Assumption, the closest island to Aldabra, which, at the time, had no known introduced bird species. The growing
population of ring-necked parakeets (Psittacula krameri) on Mahé posed a threat to endemic Seychelles black parrots
(Coracopsis barklyi) on Praslin where the Vallée de Mai forms their core breeding habitat. In 2012, red-whiskered
bulbuls and Madagascar fodies were detected on Aldabra, so an additional eradication was started. All eradications
used a combination of mist-netting and shooting. The intensive part of each eradication lasted three years or less. On
Assumption, 5,279 red-whiskered bulbuls and 3,291 Madagascar fodies were culled; on Mahé, 545 parakeets were culled;
and on Aldabra 262 Madagascar fodies and one red-whiskered bulbul were culled. Each programme underwent 1–2 years
of follow-up monitoring before eradication was confirmed, and four of the five eradications have been successful so far.
None of these species had previously been eradicated in large numbers from other islands so the successes substantially
advance this field of invasive species management. The challenges and insights of these eradications also provide unique
learning opportunities for other invasive avian eradications.
Keywords: Aldabra, Indian Ocean islands, invasive alien bird management and control, mist-netting, parakeets,
passerines, shooting
INTRODUCTION
Birds are currently one of the least represented groups
of terrestrial vertebrates in the field of invasive alien
species research and management, and the development
of successful eradication strategies for introduced birds
remains in its infancy, especially when compared to wellestablished invasive mammal control techniques (see:
Blackburn, et al., 2009; Feare, 2010; Bauer & Woog, 2011;
Strubbe, et al., 2011; Baker, et al., 2014; and Menchetti &
Mori, 2014; for potential reasons for the discrepancy). The
relatively few examples of successful large-scale avian
eradications include rock pigeons (Columba livia) from the
Galápagos Islands (Brand Phillips, et al., 2012), which at
the time was the largest successful eradication of an alien
bird from an island system (with 1,477 birds removed), and
several eradications of the common myna (Acridotheres
tristis) (e.g. Saavedra, 2010; Canning, 2011; Feare, et
al., 2017). There has, however, been little development
of best practices or compilation of lessons learnt so far.
Furthermore, we are not aware of any examples of preemptive invasive bird eradications from islands to protect
native biodiversity on nearby islands.
Since 2011, five separate eradication programmes for
invasive alien birds have been conducted on three islands
in the Seychelles by the Seychelles Islands Foundation
with the aim of protecting endemic biodiversity on Aldabra
Atoll and Praslin from the potential impacts of these
invasive bird species, should they become established.
These eradications targeted: (1) red-whiskered bulbuls
(Pycnonotus jocosus) on the island of Assumption;
(2) Madagascar fodies (Foudia madagascariensis) on
Assumption; (3) red-whiskered bulbuls on Aldabra;
(4) Madagascar fodies on Aldabra; and (5) ring-necked
parakeets (Psittacula krameria) on the main Seychelles
island of Mahé. Red-whiskered bulbuls have a broad
introduced range covering 15 countries (Global Invasive
Species Database (GISD), 2017), and their impacts on
native ecosystems and biodiversity (Clergeau & MandonDalger, 2001; Linnebjerg, et al., 2010; GISD, 2017) have
prompted control efforts and even small-scale eradications,
but these efforts have not been upscaled in most places.
Madagascar fodies are widely introduced across the
Western Indian Ocean islands including many of the
Seychelles islands, where they threaten native avifauna
through hybridisation (Lucking, 1997), and transmission
of pathogens (de Sales Lima, et al., 2015). Ring-necked
parakeets have been introduced to over 35 countries
outside their native range, making them one of the most
successful avian invaders in the world, and are known to
cause detrimental impacts on native wildlife (Strubbe &
Matthysen, 2007; Strubbe & Matthysen, 2009; Strubbe, et
al., 2010; GISD, 2017), but have not yet been eradicated or
substantially reduced in numbers from any of them.
In this paper, we present a general overview of each
eradication including: (i) the main methods applied in each
phase; (ii) the relative success and numbers of birds culled
with each method; and (iii) the difficulties encountered.
Finally, we suggest 10 key insights and recommendations
that can be applied to further eradication attempts and
adopted for best practice, and offer a positive outlook for
the future of introduced bird management.
METHODS
Location and background of project
The Seychelles archipelago consists of 115 islands
across the Western Indian Ocean region (Fig. 1). The
country has two UNESCO World Heritage sites; Aldabra
Atoll, which was inscribed on the World Heritage list in
1982, and the Vallee de Mai, inscribed in 1983. Aldabra
(15,250 ha; 9°24' S, 46°20' E; Fig. 1), one of the largest
raised coral atolls in the world, is famous for its remarkable
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
282
up to meet the challenge, pp. 282–288. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Bunbury, et al.: Invasive bird eradications, Seychelles
biodiversity, including the largest giant tortoise population
in the world, huge seabird colonies, pristine marine
ecology and its relative lack of ecological disturbance.
The Vallée de Mai (4°19′ S 55°44′ E), a 20 ha site on the
island of Praslin (Fig. 1), is a mature palm forest dominated
by the endangered endemic giant palm, the coco de mer
(Lodoicea maldivica). A public trust, the Seychelles
Islands Foundation (SIF), is responsible for managing and
protecting both sites. The sites form crucial strongholds
for many endemic and/or endangered species, and both
sites host endemic bird species that face increasing threats
from the invasive birds present on nearby islands. This
context prompted SIF to consider and initiate pre-emptive
management action of the introduced species in 2010, to
ensure protection of the endemic species.
Assumption (Roberts, 1988). The main threats posed by
the potential spread of these introduced species to Aldabra
were considered to be competition, hybridisation and
transmission of novel pathogens. When the Assumption
eradication of red-whiskered bulbuls and Madagascar
fodies was being planned in 2010/2011, Aldabra was not
known to have any introduced bird species and may have
been the largest tropical island to be free of invasive birds.
Unfortunately however, both of the introduced species
from Assumption were identified on Aldabra in early 2012,
soon after the start of the Assumption eradications. This
was thought to be due to the increasing populations of
both species on Assumption, so an additional eradication
operation for these new populations on Aldabra was
quickly planned.
In the case of Aldabra, Assumption Island (1,171
ha, 9°44' S, 46°30' E; Fig. 1), only 27 km away, had
populations of red-whiskered bulbuls and Madagascar
fodies which were introduced in the 1970s. Aldabra’s native
avifauna, including the endemic Aldabra fody (Foudia
aldabrana) and a native sub-species of Madagascar bulbul
(Hypsipetes madagascariensis rostratus), had long been
considered threatened by the proximity (sensu propagule
pressure, Simberloff, 2009) of these introduced birds on
In the case of the Vallée de Mai, the mature coco de mer
palm forest at this site forms the main breeding area for
the Seychelles black parrot (Coracopsis barklyi), which is
endemic to the island of Praslin and a flagship species for
this island. The main Seychelles island of Mahé (Fig. 1), ca.
37 km away from Praslin, had a rapidly growing population
of introduced ring-necked parakeets since the 1990s. The
increasing probability of their establishment on Praslin,
was accompanied by threats to the black parrot through
competition and pathogen transmission. The presence
of the parakeets on Mahé was thus considered the most
pressing threat to these endemic parrots, which number
only 520–900 birds on one island in the wild (Reuleaux, et
al., 2013). In addition, long-term conservation plans for the
black parrot include possible translocations of the species
to other islands (Rocamora & Laboudallon, 2009) and such
interventions could not be considered while ring-necked
parakeets remained on Mahé.
Eradication time-frames and methods
All of the eradications were initiated in 2011/2012 and
started with a 2–6 month initial phase, which included
surveys to estimate the population size and distribution of
the introduced bird populations, and trials to identify the
most effective eradication methods.
Population estimates were carried out by islandwide distance sampling for Madagascar fodies and redwhiskered bulbuls on Assumption, grid-based surveys on
Aldabra, and standardised roost counts for ring-necked
parakeets on Mahé.
Fig. 1 Location of the Seychelles archipelago in the Indian
Ocean (top, inset), the main islands and island groups of
the Seychelles, including Aldabra and Assumption (top),
and the inner Seychelles islands (bottom), showing
Mahé, Silhouette and Praslin.
The choice of eradication methods trialled in the first
phase of each project (see Table 1) was based on literature
research, staff experience with the species, advice from
experts, and experimentation. The trialled methods
included trapping (using a number of types of trap, bait,
trapping locations, decoys and playback), ground mistnetting, shooting and poisoning, as well as manual
methods such as location and hand-capture of birds at
nests and roosts. For the ring-necked parakeets, high-level
(canopy) mist-netting was also trialled, which involved
mist-nets set up in the canopy at 8–15 m from the ground
using either bamboo poles or tree branches. The outcomes
of these initial trials in terms of capture rates, efficiency,
cost and labour intensiveness were then assessed and
informed the choice of focal method(s) for the main phase
of each eradication (Table 1). Thereby, the methods used
for each eradication varied by island, species and phase of
the project. Nevertheless, amendments needed to be made
throughout the main eradication phase as the situation
changed, so flexibility and adaptability in approach was
essential.
The initial phase was followed by a second phase of
intensive eradication efforts which lasted about three years
for all of the eradications. During this phase the focus was
on reducing the target bird population numbers to zero
283
Island invasives: scaling up to meet the challenge. Ch 2B Other taxa: Birds
Table 1 Trialled methods in phase 1 and focal methods in phase 2 for each of the eradications.
Island
Species
Assumption Red-whiskered
bulbul
Assumption Madagascar
fody
Aldabra
Red-whiskered
bulbul
Aldabra
Madagascar
fody
Methods trialled in phase 1
Trapping, shooting, poisoning, hand-capture at
nests/roosts
Trapping, shooting, poisoning, hand-capture at
nests/roosts
Mist-netting, shooting
Mahé
Trapping, canopy mist-netting, ground mistnetting, nest cavity targeting, shooting along
flight lines and feeding areas
Ring-necked
parakeet
Mist-netting, shooting, hand capture
as quickly and efficiently as possible using the methods
identified in the trial phase. The second phase started in
2012 for Madagascar fodies and red-whiskered bulbuls
on both islands, and in 2013 for the ring-necked parakeets
after approval to use firearms was granted. It ended when
no more birds could be detected.
Outreach was an important part of the ring-necked
parakeet eradication in particular and efforts were made
at the start of the intensive phase of this project to reach
as many people as possible to encourage them to call
the team with any information on sightings. We initially
used all means available (including radio, TV, talks and
presentations, newspaper and magazine articles, social
media posts, website, newsletters, posters, stickers) to
spread the message, and fine-tuned this according to
responses over time.
The third and final phase consisted of monitoring
(direct observations at all sites, island-wide point counts
on Assumption; grid-based surveys in and surrounding the
invaded area of Aldabra; roost and feeding tree checks at all
known sites on Mahé) to confirm that no individuals of the
target species remained. The monitoring was implemented
in four 2–3 week periods with a team of 2–4 local scientific
staff who had experience in one or more bird eradications,
every 3–6 months.
Phase 2 focal method(s)
Mist-netting, then shooting
Mist-netting, then shooting
Mist-netting
Mist-netting, supplemented by
shooting with air rifle & hand
capture of fledglings
Shooting along flight lines and
feeding areas with shotgun
RESULTS
Bird removal
Table 2 summarises pre-eradication population
estimates and the total number of birds culled in each
eradication, with estimates of the size of the introduced
bird populations ranging from two to 4,300.
To date, four of the five eradications have been
successful, with only the ring-necked parakeet eradication
still in the monitoring phase. On Assumption and Aldabra,
there were no sightings of either introduced bird species
in two years of monitoring so both islands are again
considered free of invasive birds.
Efficiency of control methods
The proportion of birds culled using different methods
varied in each eradication (Fig. 2) and only a summary
is provided here. The predominant and most effective
methods for all campaigns were shooting and mist-netting
(Table 1; Fig. 2).
For the ring-necked parakeets on Mahé, mist-netting
caught 25 birds in the trials and first two months of the
intensive phase of the campaign, but quickly became
unfeasible as the birds learnt to avoid the nets even when
set up in different places. Trapping caught no birds. The
Table 2 Summary of the pre-eradication population estimate and the number of birds culled for each of the target
populations.
Island
Assumption
Assumption
Aldabra
Aldabra
Mahé
Species
Red-whiskered
bulbul
Madagascar fody
Red-whiskered
bulbul
Madagascar fody
Ring-necked
parakeet
Pre-eradication
Number of birds Population estimation method and
population estimate
culled
reference
4,300
5,279
Distance sampling; Feare & FriesLinnebjerg, 2012
1,600
3,291
Distance sampling; Feare & FriesLinnebjerg, 2012
2–3
1
Direct observations; van de
Crommenacker, 2012
150–200
262 (incl. hybrids) Point counts; van de Crommenacker,
2012
288
545*
Roost counts; Birch, et al., 2012
* The 545 ring-necked parakeets included 543 from Mahé, one bird from Silhouette and one bird from Praslin. The single ring-necked
parakeets culled on Praslin and Silhouette were assumed to have flown there from the Mahé population as there were no records of
captive birds on either island.
284
Bunbury, et al.: Invasive bird eradications, Seychelles
was by far the most effective capture method in the early
part of the intensive eradication phase. This was labourintensive and most successful when targeted at flight lines
to and from nocturnal communal roosts. The propensity
for communal roosting varied seasonally and the location
of flight lines required constant monitoring to maximise
mist-net captures. As numbers of target birds fell and
mist-netting became less effective, shooting became the
dominant method in the last year of the eradication. Both
bird species on Assumption also appeared to be extremely
wary of humans, even before the start of the project, and
this became more marked as the eradication progressed.
The originally planned methods therefore had to be reassessed early in the project and underwent continual
assessment as the eradication progressed.
Fig. 2 The proportion (%) of birds culled on each
introduced bird eradication (ring-necked parakeets
[Psittacula krameri] from Mahé, red-whiskered bulbuls
[Pycnonotus jocosus] and Madagascar fodies [Foudia
madagascariensis] from Assumption; and Madagascar
fodies from Aldabra) using the main eradication methods
of shooting and mist-netting. ‘Other methods’ include
trapping and manual capture. The red-whiskered bulbul
eradication on Aldabra is not included in this figure
because there was only one bird (mist-netted).
parakeet on Silhouette was culled by a member of the
public using a catapult – a method not trialled on Mahé.
The bird on Praslin was hand-caught.
Ring-necked parakeet eradication outreach
Television adverts were found to have prompted ca.
70% of all callers with information on sightings of the
parakeets, with less than 10% of responders prompted by
newspaper and magazine articles and the remainder from
presentations, social media and having seen the posters.
DISCUSSION
Difficulties encountered
For each island and species, there was a particular
set of challenges to overcome. On Aldabra, Madagascar
fodies very closely resembled the endemic Aldabra fodies.
This caused problems with capture of non-target species,
and identification of introduced vs endemic species at a
distance. The two species also quickly hybridised (van
de Crommenacker, et al., 2015), making the eradication
decisions more complex. Most birds targeted therefore
needed to be identified at very close range to ensure that no
Aldabra fodies were culled. Aldabra’s physical challenges
also included impenetrable vegetation, treacherous terrain
and extremely demanding logistics. The invaded area was
in the most remote part of Aldabra, initially had no field
station, freshwater or facilities, and is only accessible
via boat on a high tide, followed by a one-hour hike.
Establishing basic infrastructure was therefore an essential
pre-requisite for this eradication to proceed.
On Assumption the main challenges were higher than
estimated population sizes of both target species, and
the fact that neither species behaved as predicted from
previous observations elsewhere. For example, trapping
was initially anticipated to be an important and relatively
simple capture method throughout the eradication, but
this method appeared almost completely ineffective in
extensive trials of the first phase. The failure of birds to
accept bait (without traps) or to enter traps, combined with
the high densities of both species, meant that mist-netting
On Mahé, mist-netting and trapping of ring-necked
parakeets proved ineffective or inefficient – the birds
were found to fly and roost usually too high for mistnetting, and several specialised trap designs (including
the use of decoys) were unsuccessful. Poisoning could
not be considered on Mahé because of possible effects on
humans and non-target species. This left shooting as the
only viable alternative, which was a politically and socially
difficult method to adopt. Mahé is an inhabited island,
with a population of ca. 80,000 people, and eradication
activities had to occur in inhabited areas as the birds were
predominantly observed in agricultural and cultivated
areas with crops and fruit trees. The Seychelles is, for
historical and security reasons, highly sensitive about the
use of firearms and this resulted in a delay of two years
before firearms were approved for use on the project.
Shooting was then permitted to external hunters, provided
they were accompanied by a military escort at all times and
used only shot-gun and air rifle. Ring-necked parakeets are
also highly intelligent birds and became ‘educated’ and
wary very quickly. For example, we think they learnt to
recognise and avoid the project car and staff uniforms.
Shooting therefore had to be done with extreme caution
(e.g. from cryptic locations, wearing camouflage gear,
only shooting at groups of one or two birds, and shooting
only when the hunter was very confident of a strike). A
final critical issue with working on an inhabited island was
public perceptions concerning the project, especially with
such a charismatic target species and because the success
of any eradication partly depends on public support and
contribution (Mack, et al., 2000). SIF tackled this potential
problem from the outset by conducting intensive outreach
campaigns to try to ensure that as many people on Mahé
as possible were aware of the eradication and the reasons
for it. Lack of support did cause occasional problems with
access to private land and misinformation. Fortunately, the
parakeets were a known pest and commonly viewed as a
threat to farming and endemic wildlife, so the majority of
people encountered were in favour of the project and very
supportive.
Ten key insights and recommendations
Here is a list of 10 key insights from these eradications,
which can serve as a basis for recommendations for
practitioners who are considering invasive alien bird
eradications. The eradications presented here cover islands
from both ends of the ecological disturbance spectrum,
from the most ecologically depauperate (Assumption), to
the least disturbed and most biodiversity-rich (Aldabra)
making the lessons relevant to a broad suite of islands.
1. Large-scale invasive bird eradications are feasible
Red-whiskered bulbuls in the Seychelles occurred
only on Assumption, plus the single bird on Aldabra,
so the outcome of these eradications has been national
285
Island invasives: scaling up to meet the challenge. Ch 2B Other taxa: Birds
elimination of the species. With the parakeet’s range on the
Seychelles encompassing only Mahé, if this eradication is
successful, it will mark a second national eradication of an
invasive bird species of high concern. Madagascar fodies
remain established in high numbers on many islands of the
Seychelles, but their eradication from two very different
islands confirms the feasibility of this approach, should
there be a need to consider their eradication elsewhere.
Therefore, our first key message is that eradications of
invasive alien birds from islands are feasible, even if the
population of the target species exceeds 5,000 birds.
2. Pre-emptive action should be considered as a means to
remove perceived threats
The three initial eradications of red-whiskered bulbuls
and Madagascar fodies on Assumption, and ring-necked
parakeets on Mahé, were based on the precautionary
principle, i.e. the aim was to protect threatened endemic
biodiversity pre-emptively based on perceived threats. This
was justified in the case of the red-whiskered bulbul and
the ring-necked parakeet, which have known detrimental
impacts in their introduced ranges. However, even in the
case of the Madagascar fody, the impacts of which on
endemic birds have been questioned (Garrett, et al., 2007),
the perceived threats were verified during the course of the
eradications: (i) all three target species reached the islands
of concern and at least one of these species established a
breeding population (Madagascar fodies on Aldabra); (ii)
hybridisation was confirmed to occur between introduced
and endemic fodies on Aldabra (van de Crommenacker,
et al., 2015); and (iii) several potentially novel pathogens
were identified in the invasive species (SIF, unpubl. data).
3. Don’t assume what you know of a species from other
locations will apply in a new area – plan to conduct
initial trials
Based on experience of the same species elsewhere, we
expected a main method for catching Madagascar fodies
and red-whiskered bulbuls to be trapping, and planned
accordingly with respect to equipment and logistics.
Trapping can be an effective capture method elsewhere
for these species (N.B. & P.H., pers. obs.; C.G. Jones,
pers. comm., all in Mauritius), but was found to be almost
completely ineffective on Assumption for reasons that are
unclear, and the birds never became accustomed to baited
areas. This was despite several members of staff working
on the project who had extensive experience successfully
trapping these species in other locations. This caused
delays at the beginning of the eradication while methods
were re-assessed and other equipment sourced. A similar
problem was encountered with the ring-necked parakeets,
which have been successfully trapped in other countries
(e.g. Bashir, 1979; Hussain, et al., 1992), but could not be
trapped using the same or similar trap designs on Mahé,
although these problems were less significant, as trapping
parakeets had not been assumed as a main method of
capture. It is important to note that we are not ruling out any
particular method for targeting these species elsewhere.
Trapping may still be a highly effective capture technique
in other places for these species, so our advice here is
simply that initial small-scale trials should be conducted to
determine the feasibility of several different methods and
save time and funding.
4. One size doesn’t fit all birds
Shooting was by far the most effective method for ringnecked parakeets in the Seychelles, while mist-netting
proved to be generally more effective for passerines
(although this depended on the phase of the eradication).
However, a flexible approach and willingness to modify the
286
strategy was critical for the success of these eradications.
Even within the same species and island, our techniques
needed to be assessed and ‘tweaked’ frequently (and
often substantially) to maintain efficient capture rates. For
example, on Assumption, there was a switch in the final
year of the intensive phase of the eradication, from using
mist-nets as the main method of capture to firearms (this
switch also applies to mynas; Feare, et al., 2017). This
was decided when catch rates in mist-nets (i.e. the density
of target population) had dropped too low for continued
progress with the eradication (i.e. population recruitment
rates were thought likely to be equal to or higher than
capture rates).
5. Don’t count your eggs before they hatch
For all three species, there were more birds present
than had been estimated by survey methods. This was the
case regardless of which estimation method was used.
The higher numbers are likely to have been primarily due
to recruitment of young birds into the populations since
distance sampling is based on classical closed population
sampling (Cassey & McArdle, 1999) but the survey
methods (roost counts, distance sampling) could also
have produced underestimates. The higher figures had
implications for the planning and especially the costs of
completing the eradications.
6. Identify the weak points of your target species
Each target species was found to have at least one
trait or habit which either increased their vulnerability at
certain times or to certain methods, or could be used to
improve eradication effectiveness. The communal roosting
sites of ring-necked parakeets enabled regular standardised
counts to be conducted, which initially provided a valuable
way to monitor the population numbers, flight lines and
the impacts of the eradication efforts and later formed an
essential location for targeting the remaining birds. These
sites proved so useful that parakeets were not targeted at
roosts until close to the end of the project to ensure that
the roost sites were not disturbed or compromised. Redwhiskered bulbuls also roosted communally in the early
stages of the project and could be targeted with mist-nets
along their flight lines towards roosts, which maximised
the mist-net catch. Later in the eradication, their habit of
vocalising from prominent perches meant that they could
be reliably located from several hundred metres away,
which greatly helped in the search for and targeting of the
last few birds. Madagascar fodies were found to have a
tendency to form large foraging groups, especially in the
non-breeding season, which, when spotted, provided key
areas for mist-netting.
7. Use research to aid management decisions on the
ground
A scientific and research-based approach was an
important aspect of the eradications and greatly facilitated
management decisions on the ground. This included
collecting comprehensive data and samples from all
birds caught, regularly analysing the effectiveness of
methods and approaches, and setting up external research
partnerships for analysis which could not be done on
site. The strongest example of this was the case of the
Madagascar fody introduction to Aldabra, for which
SIF was able to quickly establish a collaboration with
university researchers, ensuring that the samples and
data collected could be rapidly and effectively analysed.
The resulting research outputs included analysis of origin
(Assumption) and timing (recent, but probably pre-dating
the start of the eradication) of the invasion, as well as
confirmation of hybridisation between the endemic and
Bunbury, et al.: Invasive bird eradications, Seychelles
introduced species and more insights into this process (van
de Crommenacker, et al., 2015). A collaboration was also
established for disease-screening of ring-necked parakeet
and black parrot blood samples to provide information on
the pathogen status of each species.
Seychelles are the first of their kind suggests that a change
in approach and mindset to invasive bird eradications
is timely. We believe that insights gained from these
programmes can be used as a basis to significantly advance
the field of invasive bird management and to initiate the
development of best practices for eradication attempts.
8. Training of local staff is essential for project success
Few people with the necessary technical skills needed
for the eradications existed in the Seychelles when the
project started, so more than 30 local staff were intensively
trained on the job throughout the eradications. Five of these
staff members subsequently led parts of the eradications and
were crucial to their success. Several of the staff members
have subsequently been recruited in other invasive species
management positions within SIF and elsewhere, so the
eradications have increased in-country capacity in this
field. Indeed, local staff training is seen as one of the
biggest achievements of the eradications and has had the
additional benefit of providing a strong sense of national
ownership to the eradications.
9. Assess effectiveness of publicity and focus on the most
appropriate means
Outreach activities are important in any eradication but
in some, they are an essential means of achieving success.
For the ring-necked parakeet eradication, on assessing
where callers had heard about the project we found that
the vast majority were prompted by the TV advert so
we were able to focus on this for the rest of the project,
which reduced costs and time without compromising the
information received. In addition to public outreach, we
found it was essential to liaise with other stakeholders
in the environmental sector about the importance of the
eradications. We noticed that the eradications tended to
bring out strong feelings either for or against the project,
and most people appreciated an opportunity to ask questions
and understand the reasons for it. Our impression was that
the outreach and education carried out for these projects
went a long way to increase public support although we
have no way of quantifying this
10. The early bird catches the worm
In the case of these eradications, we are certain that preemptive action has been a more effective and cost-efficient
strategy to protect endemic species than would have been
the case had we waited for the introduced species to spread
to Aldabra and the Vallée de Mai (or other islands in the
Seychelles) and establish populations. Indeed, this had
already started to happen with all three species and, had
we waited much longer, eradication may have proved an
impossible task. Finally, at least one and potentially two of
these invasive bird species are now nationally eradicated
from the Seychelles and there is minimal risk of them
being reintroduced to the sites in the future. We therefore
consider the biodiversity and ecological integrity of the
Seychelles World Heritage sites to have been safeguarded
from these particular threats by these eradications.
CONCLUSION
Although all three species targeted here are known
invasive species, and control efforts have been made or are
underway in several places, there were no previous records
of them being removed in such large numbers, or their
complete eradication from any other islands or countries.
The challenges and successes of these eradications provide
a unique learning opportunity and offer a positive outlook
for the future of introduced bird eradications. The fact
that these eradication successes (or near successes) in the
ACKNOWLEDGMENTS
We thank the Seychelles Ministry for Environment,
Energy and Climate Change for their support and
facilitation of all of these eradications. We are also very
grateful to our key partners – the Islands Development
Company, the Police Special Services Wing, the Seychelles
People’s Defence Force, and the Seychelles National Parks
Authority. The red-whiskered bulbul and Madagascar fody
eradications on Assumption, and the ring-necked parakeet
eradication on Mahé were carried out under an EU-funded
project (project DCIENV/2010/220-252: “Mainstreaming
the management of invasive alien species to preserve
the ecological integrity and enhance the resilience of
Seychelles World Heritage Sites”) implemented by SIF
between 2011 and 2015. Later financial support for the
ring-necked parakeet eradication was kindly provided by
the Environment Trust Fund Seychelles and the Global
Environmental Faculty. The red-whiskered bulbul and
Madagascar fody eradications on Aldabra were initiated
under an Emergency Assistance project funded by
UNESCO. We thank all of our funders for their essential
support of these eradication programmes. This paper
includes as co-authors all staff who have worked directly
on the eradication programmes on two or more of the
three eradication locations (Assumption, Aldabra, Mahé)
for at least a year. For the Assumption eradications, Chris
Feare and Jannie Linnebjerg developed and led the initial
research into the eradication and monitoring methodology
and conducted the pre-eradication population estimates.
Phase 1 of the Mahé eradication of ring-necked parakeets
was led by Darryl Birch and Pete Haverson. In addition,
all of the many consultants and staff who have worked
hard on one or more of the eradications to make them a
success are thanked, including Wilna Accouche, Nyara
Anacoura, Lina Barbe, Samuel Bassett, Paul Benoit, Jude
Brice, Unels Bristol, April Burt, Sheril de Commarmond,
Stan Denis, Steve Denis, Martin van Dinther, Marcus
Dubel, Roland Duval, Rebecca Fillipin, Ronny Gabriel,
Frankie Gamble, Rosanna Gordon, Murvin Green, Arjan
de Groene, Oskar Guy, Helga Hoareau, Jakawan Hoareau,
Lauren Koehler, Israel Labrosse, Ronny Marie, Stephanie
Marie, Jamie McAuley, Alex McDougall, Pete McIntosh,
Glenn McKinlay, Adam Mitchell, Julio Moustache, Reza
Moustache, Catherina Onezia, Hendrick Quatre, Terance
Payet, Jeremy Raguain, Lotte Reiter, Heather Richards,
Marvin Roseline, Jovani Simeon, Abel Sorry, Joel
Souyave, Chris Tagg, Rowana Walton, Jeremy Waters and
Jack West. Thanks to Carl Jones for his expertise on the
trapping of Madagascar fodies and red-whiskered bulbuls.
Finally, we would like to thank the II17 organisers and
Dick Veitch for the opportunity to present and submit this
manuscript.
REFERENCES
Baker, J., Harvey, K.J. and French, K. (2014). ‘Threats from introduced
birds to native birds’. Emu 114: 1–12.
Bashir, W.A. (1979). ‘A new “PAROTRAP” adapted from the Mac trap
for capturing live parakeets in the field.’ Wildlife Damage Management,
Internet Center for Bird Control Seminars Proceedings 23, pp. 167–
171. Lincoln, US: University of Nebraska.
Bauer, H-G. and Woog, F. (2011). ‘On the ‘invasiveness’ of non-native
bird species’. Ibis 153: 204–206.
287
Island invasives: scaling up to meet the challenge. Ch 2B Other taxa: Birds
Birch, D., Haverson, P., Simeon, J. and Constance, A. (2012). Phase
1 Report: Inception Report and Plan to Eradicate the Ring-necked
Parakeet (Psittacula krameri) from Mahé, Seychelles (Project Objective
2b). Unpublished report. Seychelles. Seychelles Islands Foundation.
Strubbe, D. and Matthysen, E. (2009). ‘Experimental evidence for nestsite competition between invasive ring-necked parakeets (Psittacula
krameri) and native nuthatches (Sitta europaea)’. Biological
Conservation 142: 1588–1594.
Blackburn, T.M., Lockwood, J.L. and Cassey, P. (2009). Avian Invasions:
The Ecology and Evolution of Exotic Birds. Oxford, UK: Oxford
University Press.
Strubbe, D., Matthysen, E. and Graham, C.H. (2010). ‘Assessing the
potential impact of invasive ring-necked parakeets Psittacula krameri
on native nuthatches Sitta europaea in Belgium’. Journal of Applied
Ecology 47: 549–557.
Brand Phillips, R., Cooke, B.D., Carrión, V. and Snell, H.L. (2012).
‘Eradication of rock pigeons, Columba livia, from the Galápagos
Islands’. Biological Conservation 147: 264–269.
Canning, G. (2011). ‘Eradication of the invasive common myna,
Acridotheres tristis, from Fregate Island, Seychelles’. Phelsuma 19:
43–53.
Cassey, P. and McArdle, B. H. (1999). ‘An assessment of distance
sampling techniques for estimating animal abundance’. Environmetrics
10: 261–278.
Clergeau, P. and Mandon-Dalger, I. (2001). ‘Fast colonization of an
introduced bird: the case of Pycnonotus jocosus on the Mascarene
Islands’. Biotropica 33: 542–546.
Feare, C.J. (2010). ‘Invasive bird eradication from tropical oceanic
islands’. Aliens: The Invasive Species Bulletin 30: 12–19.
Feare, C.J. and Fries-Linnebjerg, J. (2012). Eradication of Avian Invasive
Alien Species on Assumption Island Including Feasibility Studies,
and Development and Implementation of Eradication Plan (Project
Objective 1a). Phase 1 Report: Inception Report and Eradication Plan.
Unpublished report. Seychelles. Seychelles Islands Foundation,
Feare, C.J., van der Woude, J., Greenwell, P., Edwards, H.A., Taylor, J.A.,
Larose, C.S., Ahlen, P-A., West, J., Chadwick, W., Pandey, S., Raines,
K., Garcia, F., Komdeur, J. and de Groene, A. (2017). ‘Eradication of
common mynas Acridotheres tristis from Denis Island, Seychelles’.
Pest Management Science 73: 295–304.
Garrett, L.J., Jones, C.G., Cristinacce, A., and Bell, D.J. (2007).
‘Competition or co-existence of reintroduced, critically endangered
Mauritius fodies and invasive Madagascar fodies in lowland Mauritius?’
Biological Conservation 140: 19–28.
GISD (2017). ‘Global Invasive Species Database’ <http://www.iucngisd.
org/gisd/speciesname/Psittacula+krameri> and <http://www.iucngisd.
org/gisd/speciesname/Pycnonotus+jocosus>. Accessed 18 August 2017.
Hussain, I., Khan, A.A. and Munir, S. (1992). ‘Trapping success of
PAROTRAP.’ Pakistan Journal of Zoology 24: 170–172.
Linnebjerg, J., Hansen, D.M., Bunbury, N. and Olesen, J.M. (2010). ‘Diet
composition of the invasive red-whiskered bulbul Pycnonotus jocosus
in Mauritius’. Journal of Tropical Ecology 26: 347–350.
Lucking, R.S. (1997). ‘Hybridization between Madagascan red fody
Foudia madagascariensis and Seychelles fody Foudia sechellarum on
Aride Island, Seychelles’. Bird Conservation International 7: 1–6.
Mack, R.N., Simberloff, D., Lonsdale, W.M., Evans, H., Clout, M.N. and
Bazzaz, F.A. (2000). ‘Biotic invasions: causes, epidemiology, global
consequences, and control.’ Ecological Applications 10: 689–710.
Menchetti, M. and Mori, E. (2014). ‘Worldwide impact of alien parrots
(Aves Psittaciformes) on native biodiversity and environment: A
review’. Ethology Ecology and Evolution 26: 172–194.
Reuleaux, A., Bunbury, N., Villard, P. and Waltert, M. (2013). ‘Status,
distribution and recommendations for monitoring of the Seychelles
black parrot Coracopsis (nigra) barklyi’. Oryx 47: 561–568.
Roberts, P. (1988). ‘Introduced birds on Assumption Island—a threat to
Aldabra’. Oryx 22: 15–17.
Rocamora, G. and Laboudallon, V. (2009). Seychelles Black Parrot
Coracopsis (nigra) barklyi Conservation Assessment and Action Plan.
2009–2013. FFEM Project ‘Réhabilitation des Ecosystèmes Insulaires’.
Seychelles: Island Conservation Society and MENRT.
Saavedra, S. (2010). ‘Eradication of invasive mynas from islands: Is it
possible?’ Aliens: The Invasive Species Bulletin: 29: 40–47.
de Sales Lima, F.E., Gil, P., Pedrono, M., Minet, C., Kwiatek, O., Campos,
F.S., Spilki, F.R., Roehe, P.M., Franco, A.C., Maminiaina, O.F., Albina,
E. and de Almeida, R.S. (2015). ‘Diverse gammacoronaviruses
detected in wild birds from Madagascar’. European Journal of Wildlife
Research 61: 635–639.
Simberloff, D. (2009). ‘The role of propagule pressure in biological
invasions’. Annual Review of Ecology, Evolution, and Systematics 40:
81–102.
Strubbe, D. and Matthysen, E. (2007). ‘Invasive ring-necked
parakeets Psittacula krameri in Belgium: Habitat selection and impact
on native birds’. Ecography 30: 578–588.
288
Strubbe, D., Shwartz, A. and Chiron, F. (2011). ‘Concerns regarding the
scientific evidence informing impact risk assessment and management
recommendations for invasive birds’. Biological Conservation 144:
2112–2118.
van de Crommenacker, J. (2012). First Annual Report: Takamaka
Invasive Bird Project 2012. Unpublished report. Seychelles. Seychelles
Islands Foundation,
van de Crommenacker, J., Bourgeois, Y.X.C., Warren, B.H., Jackson,
H., Fleischer-Dogley, F., Groombridge, J. and Bunbury, N. (2015).
’Using molecular tools to guide management of invasive alien species:
Assessing the genetic impact of a recently introduced island bird
population’. Diversity and Distributions 21: 1414–1427.
E. Hagen, J. Bonham and K. Campbell
Hagen, E.; J. Bonham and K. Campbell. House sparrow eradication attempt on Robinson Crusoe Island, Juan Fernández Archipelago, Chile
House sparrow eradication attempt on Robinson Crusoe Island,
Juan Fernández Archipelago, Chile
E. Hagen1, J. Bonham1 and K. Campbell1,2
1
Island Conservation, Las Urbinas 53 Santiago, Chile. <erin.hagen@islandconservation.org>.2School of Geography,
Planning & Environmental Management, The University of Queensland, St Lucia 4072, Australia.
Abstract House sparrows (Passer domesticus) compete with native bird species, consume crops, and are vectors for
diseases in areas where they have been introduced. Sparrow eradication attempts aimed at eliminating these negative
effects highlight the importance of deploying multiple alternative methods to remove individuals while maintaining
the remaining population naïve to techniques. House sparrow eradication was attempted from Robinson Crusoe Island,
Chile, in the austral winter of 2012 using an experimental approach sequencing passive multi-catch traps, passive singlecatch traps, and then active multi-catch methods, and finally active single-catch methods. In parallel, multiple detection
methods were employed and local stakeholders were engaged. The majority of removals were via passive trapping, and
individuals were successfully targeted with active methods (mist nets and shooting). Automated acoustic recording, point
counts and camera traps declined in power to detect individual sparrows as the population size decreased; however, we
continued to detect sparrows at all population densities using visual observations, underscoring the importance of local
residents’ participation in monitoring. Four surviving sparrows were known to persist at the conclusion of efforts in 2012.
Given the lack of formal biosecurity measures within the Juan Fernández archipelago, reinvasion is possible. A local
network of citizen observers is the best tool available to detect house sparrows at low density, however ongoing, dedicated
eradication funding does not exist. Opportunistic removals via shooting have been possible from 2013–2016, but elusive
individual sparrows were seen during a small number of days each year suggesting remnant group(s) exist in yet unknown
forest locations.
Keywords: invasive bird, management, Passer domesticus, trapping
INTRODUCTION
House sparrows (Passer domesticus) have a wide
range of negative impacts in areas where they have been
introduced. They affect native bird species, pushing eggs
and nestlings from nests and chasing adults (McGillivray,
1980; Gowaty, 1984); they consume crops and ornamental
plants; and are vectors of at least 29 diseases affecting
people, livestock and wildlife (Clergeau, et al., 2004;
Fagerstone, 2007). This species is an effective invader
owing to its generalist diet; rapid rate of increase,
facilitated by colonial-communal nesting, large clutch
sizes and extended breeding seasons; effective range
expansion in human-altered landscapes; and aggression
against similar and smaller sized birds (MacGregorFors, et al., 2010). The risks of house sparrows are often
underestimated and delays in rapid responses to incipient
or small localised populations can result in much more
complex and costly future actions for their management
once population growth and negative impacts on native
species are documented (Clergeau, et al., 2004). Complete
removal of invasive house sparrow populations should be
considered to eliminate risk for negative impacts arising
from the species’ invasion.
House sparrow eradication attempts on other islands
have demonstrated that the effectiveness of some methods
may decline over time, if sparrows learn to avoid them
(Bednarczuk, et al., 2010) emphasising the importance of
using a variety of techniques in an adaptive management
approach. Campaigns for house sparrow eradication should
employ multiple methods and aim to remove the entire
population within as short a time as possible. Otherwise,
given the species’ reproductive potential, there is a risk that
house sparrows will breed faster than they are removed. To
maintain naïveté of the population to methods for as long
as possible and reduce the likelihood of house sparrows
dispersing in response, methods should be implemented
strategically. The detection and removal of the last
individuals must be considered in planning the deployment
of the multiple alternative methods available (Morrison, et
al., 2007). To increase likelihood of successful eradication,
some methods should be deployed consecutively and
others sequentially with attention to maintaining sparrows
naïve to methods.
The Juan Fernández Archipelago in Chile is comprised
of three islands (Robinson Crusoe (4,790 ha), Alexander
Selkirk (4,950 ha),) and Santa Clara (220 ha)) with globally
significant biodiversity and endemism due to its isolation
and topographic variation. However, invasive species
continue to drive catastrophic changes to these unique
natural values including species extinctions and massive
erosion, as well as precipitous declines in plant and animal
species and loss of native vegetation cover (Sanders, et al.,
1982; Bourne, et al., 1992; Arroyo, 1999; Hahn & Römer,
2002). Feasibility of the complete removal of invasive
species has been explored and participatory planning with
the islands’ inhabitants and varied stakeholders continues
to advance as benefits of invasive species removals and
restoration are prioritised (Saunders, et al., 2011; Glen, et
al., 2013; Ministerio del Medioambiente, 2017).
House sparrows have been present on Robinson Crusoe
Island (RC) since 1943 as a wild population (Hahn, et al.,
2006) and none are kept as pets. The population appeared
stable at around 80 individuals and to be restricted to the
island’s only human settlement of San Juan Bautista (Hahn,
et al., 2006; Hagen pers. obs.); however, observations in
2011–2012 indicated population expansion within San
Juan Bautista into new home construction areas following
a tsunami in February 2010. The potential increased
risk from this species to single-island endemic birds and
local food production prompted a review of control and
eradication options within a local multi-stakeholder group
focused on animal issues related to conservation and local
development.
The study reports on an attempt to eradicate the
local house sparrow population within an experimental
framework to examine the efficacy of methods for house
sparrow eradications and protect local biodiversity. The
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 289–294. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
289
Island invasives: scaling up to meet the challenge. Ch 2B Other taxa: Birds
objectives of the study were to keep house sparrows naïve
and eliminate the potential for survivors to learn to avoid
methods (e.g. escape from traps).
METHODS
A range of potential methods for use in house sparrow
eradication from RC were considered (see Table 22 of
Saunders, et al., 2011). Removal techniques were evaluated
and prioritised based on previous success in bird removals,
permissibility in this urban setting, and likelihood to
contribute to sparrow learning. Toxicants were assessed,
but none were considered suitable for house sparrow
eradication (Fisher, et al., 2012). Trapping was identified
as having the greatest potential to provide a large reduction
in the house sparrow population on RC while minimising
risks to native birds and poultry. Pre-baiting was initiated
one month before removals began (15 June 2012) at 10
sites to allow house sparrows to become accustomed to
feeding at a given location on provided crushed maize
(1.6–3.2 mm diameter) and to confirm minimal attraction
of non-target species to these sites.
Passive removal techniques were employed in the
first phase of this trial, to minimise education of house
sparrows to future methods (10 July 2012–14 September
2012). Active removal techniques were added to the trial
beginning 27 July 2012.
Passive removal techniques
To minimise education of house sparrows in the
population, passive traps were employed in the initial
phase of removals.
Elevator multi-catch traps have demonstrated good
capture and low escape rates (Fitzwater, 1981). House
sparrows enter a compartment alone to feed on bait, their
body weight causes an “elevator” to lower the individual
to its “escape” into a closed cage. Without the bird’s
weight, the counterbalanced “elevator” springs back into
the original position ready for another passenger. Birds
trapped in the closed cage act as live decoys. We purchased
traps without the central mesh body for ease of transport,
and then assembled the mesh over a plywood base forming
the holding cage once on the island. Trap dimensions were
60 × 40 × 20 cm (<http://www.sparrowtraps.net/index.
htm>). Elevator traps were placed on an elevated platform,
approximately 2 m in height, to reduce the potential for trap
interference by domestic animals and private citizens. We
added a covered plywood compartment with a perch within
each elevator trap’s holding cage to provide protection
from the elements for live decoys. Decoys had primary
flight feathers on one wing clipped so that they couldn’t
fly in the event of escape. Food and water were provided.
Trio multi-catch traps are comprised of two
compartments which each function as a single-catch trap,
whose sprung doors must be manually reset after each
catch (Nature-House ST1 Trio house sparrow trap http://
www.amazon.com/Nature-House-ST1-Trio-Sparrow/dp/
B001GIP2MG). The bird drops into the compartment, onto
a perch over the feed tray which triggers the compartment
door to close. Captured individuals can freely move into
the third compartment, where they act as live decoys. Three
trio traps were deployed, mounted at least 1.5 m above the
ground to reduce potential for trap interference by domestic
animals and private citizens. We provided flooring in each
compartment to increase bait retention and partial roofs to
decrease interference from natural elements.
Modified Australian crow (MAC) traps function when
birds drop into the MAC trap to access bait and are unable
290
to fly through the trap entrance to escape as their wingspan
exceeds the diameter of the entrance. Captured individuals
alight on perches in the higher parts of this trap (Clark &
Hygnstrom, 1994). Exclusive use of a ‘mini’ MAC trap has
enabled local populations of house sparrows to be entirely
removed (McGregor & McGregor, 2008). We constructed
two mini MAC traps, retaining traditional width of slats
and height of centre board to avoid birds jumping to escape,
reducing overall length (82 × 137 × 71 cm). MAC traps
were placed on the ground given their robust size.
Nest box traps were made from nest boxes which were
converted into single-catch traps (http://www.vanerttraps.
com/urban.htm) to capture house sparrows investigating
nest cavities. In areas where house sparrows were seen
entering and exiting cavities, known cavities were covered
to exclude sparrows and nest box traps were deployed
with small feathers and fine nesting material added to the
entrances to encourage investigation.
Traps were placed within open areas where birds could
easily see them, and near frequented flyways, perches and
feeding areas. For 2–3 days before arming traps, wired-open
traps were placed at pre-baiting sites, with crushed maize
on and around the open traps, to permit birds to explore
them without risk of capture. When birds were trapped, the
trap would be covered with a bed sheet to assist calming
the birds during transport and reducing visibility to the
general public. Covered traps were then transported to a
room where any escapees could be recaptured, prohibiting
escape. Within this facility birds were removed from traps
and either selected for use as live decoys, or euthanised.
Euthanasia was via cervical dislocation; possibly the
easiest means for this species and a practical means for
mass euthanasia (Sharp & Saunders, 2005; AVMA, 2007).
Active removal techniques
As capture rates declined with passive traps, active
removal techniques were added to the trial. We continued
using passive traps simultaneously with active removal
techniques.
Walk in cage traps were used to target individual
sparrows unable to be trapped in other trap types. A
wooden box with mesh sides was set up as a walk-in cage
trap by propping open a door that opens from the bottom.
When the prop is pulled out by a nearby observer (Sharp &
Saunders, 2005), bungee cords add to downward force to
close the door quickly.
Clap traps utilise a spring-loaded throw net triggered
remotely by the trapper, which is placed on the ground
and pre-baited with crushed maize (<http://pestbarrier.
com/store/itemdesc.asp?xCc=8u4u3>). The trap was not
triggered unless all birds in a flock were able to be captured.
Mist nets are a common ornithological capture
technique for small birds and were deployed on flyways to
capture house sparrows that had avoided traps. Continual
monitoring was required to quickly remove any house
sparrows or non-target species.
Nest destruction can be used during the breeding season
to slow or halt recruitment, and may make adult birds more
susceptible to other techniques such as clap traps baited
with nest material (Fitzwater, 1994). Eggs are crushed
and nestlings euthanised (Sharp & Saunders, 2005). Nest
destruction, although planned, was not needed in our trial.
Shooting was employed in specific scenarios where
traps were proving ineffective. A 0.177 caliber air rifle with
4–12 times magnification scope (Beeman R9, Weirauch,
Germany) was utilised, targeting only individuals alone
Hagen, et al.: House sparrow, Robinson Crusoe Island, Chile
or in pairs, to avoid wariness. Adult females were targeted
first, to limit potential growth of the local population. After
2012, shooting was employed opportunistically.
Detection techniques
Eradication campaigns rely on effective detection of
the target species to indicate when individuals of the target
species no longer exist and the campaign can conclude. We
assessed potential detection techniques for house sparrows
throughout the trial, to examine their efficacy at varying
house sparrow population densities. We anticipated that
some detection methods may become innefective at low
population densities as changes in flocking, calling and
movements may result from individuals. Therefore we
deployed multiple detection techniques simultaneously in
order to ensure at least one technique was effective at even
low population densities.
Autonomous recording units (ARUs) were deployed
at 15 sites within San Juan Bautista. Ten ARUs were colocated with pre-baiting locations while the remaining units
were in locations without pre-baiting. We programmed
ARUs to record every other day for a 4-hour period
around dawn (starting 30 minutes before sunrise) when
house sparrows are known to be acoustically active. In
addition, each sensor was programmed to record one of
every 10 minutes throughout the rest of the day until 30
minutes after sundown. Data from these recordings was
available only after post-processing in a sound laboratory.
Automated analysis of all field recordings was carried out
with the eXtensible BioAcoustic Tool (XBAT, <http://
www.xbat.org>) using an image processing technique
known as spectrogram cross correlation to detect and
classify sounds on our field recordings that were correlated
with the spectral qualities of typical house sparrow calls.
Sensitivity in the detection analysis was increased to
improve the probability of detecting house sparrow calls
when few individuals remained, which led to manual
review of all events to confirm accuracy of detecting true
house sparrow calls (McKown, 2013).
Visual observations were conducted over the same
period to provide alternative detection methods in the case
that a given method failed to detect individuals even though
a population remains present. Fixed radius point counts
(Bibby, 2000; Buckland, et al., 2001) were conducted
weekly beginning 15 June 2012. Project personnel
conducted point counts 14 times throughout the trial period
at 21 locations throughout San Juan Bautista, 15 of these
locations were co-located with ARU deployment sites
and six of which were not located with acoustic sensors
or pre-baiting locations. Point counts were analysed
using the fixed-radius point count equation as detailed
by Buckland, et al. (2001), generating density estimates
by habitat type, based on the estimated total surface area
of coverage class occupied by sparrows (settlement and
cultivated Eucalyptus, Cupressus and Pinus per Greimler,
et al., 2002). Point count density estimates were compared
to recorded call rates and sparrow removals each week.
In addition to point counts, citizens were encouraged
to report opportunistic sightings of house sparrows, which
were all investigated by project staff. Multiple reports of
the same individuals, as well as uncorroborated reports
prevented clear calculations of the number of individuals
remaining.
Camera traps (Reconyx, Holmen WI) were deployed
opportunistically at pre-baiting and passive trapping
locations. Camera traps were used as an additional
technique for visual confirmation of surviving individuals.
After the intensive 2012 campaign, an early observer’s
network attempting to harness the interest and participation
of island residents was developed. This network has grown,
and has become a formalised early detection network for
invasive species, with individuals’ observations of invasive
species combined with a common smartphone application
(WhatsApp) which allows researchers to capture reports
within a database.
Stakeholder communications
Throughout the project, a communications campaign
was undertaken to highlight the threats that house sparrows
pose to local endemic species. Announcements via radio,
signs, fliers, and a booth at a children’s day event, were
complemented with active participation in the local
conservation committee, opportunistic presentations for
local institutions and a nest box design contest for local
endemic bird species. We also promoted the needs for
biosecurity and a municipal ordinance to be established to
regulate entry and possession of invasive species.
RESULTS
Methods to maximise personnel efficiency were
deployed while reducing the risk of educating animals.
Passive multi-catch traps (elevator, trio and mini-MAC
traps) were deployed first. As nest-building behaviour
was observed, passive single-catch nest box traps were
deployed. As the number of individual sparrows was
reduced, specific individuals were targeted with more timeintensive active multi-catch traps (mist net, clap trap and
walk-in trap). Shooting (active, single-catch technique)
was reserved for specific scenarios once other methods
appeared ineffective.
Personnel contributed a total of 2,600 person hours
across two months of sustained effort. A total of 814 trap
days were conducted during the trial, resulting in 89 house
sparrows removed. The majority of removals resulted from
elevator traps (46 individuals, 275 trap days), followed
by mist nets (22 individuals, 22 trap days) and trio traps
(15 individuals, 70 trap days). Additional methods did not
capture birds (modified MAC, walk-in cage, and clap traps)
or were used in specific situations, after the population
was reduced, and thus removed fewer birds (nest box trap,
1; shooting, 5). At the conclusion of the trial, four house
sparrows were known to remain on the island (two males
and two females).
Mist nets and shooting were the most effective removal
techniques when effectiveness is assessed as the number
of individual sparrows removed as a function of the days
the technique was deployed. However, both of these active
methods can educate individuals in the target population
and require much higher personnel effort as compared
to passive trap deployment (for example elevator traps
and trio traps), demonstrating that this calculation of
effectiveness is incomplete. Also, house sparrows captured
in traps appeared to be useful as decoys; however, data
specific to differential capture rates is not available.
Detection techniques
Both point counts and automated surveys detected a
decline of house sparrows after house sparrow removals
occurred. Point count density estimates showed abrupt
declines after 60 individuals had been removed from the
population, while call rates estimated from ARUs varied
more gradually over the trial period (McKown, 2013; Fig.
1). Point count observers did not detect house sparrows
291
Island invasives: scaling up to meet the challenge. Ch 2B Other taxa: Birds
Camera traps effectively captured images of house
sparrows visiting known food sources. Given the trial
setting in San Juan Bautista, some sites were inefficient for
house sparrow detections via camera given that domestic
animals, people, and objects moving in the wind would
trigger the camera traps resulting in a significant number
of images without the target species present. Camera traps
did not capture images of individuals when population
density was lowered by removals (after 15 August 2012),
demonstrating ineffectiveness as a detection method for
sparrows at low population densities.
The remnant house sparrows were infrequently
detected within the town area between 2012 and 2016 and
were reported by residents. Observations were limited
to isolated localities and dates (20–23 June 2013, one
individual detected and removed; 14 and 23 November
2015, one individual detected; 1 November 2016, five
individuals detected; 19–30 October 2016, six individuals
detected, three removed). Remaining house sparrows
successfully avoided removal techniques and, based on
inability to detect them, are thought to spend most of the
year outside of the town area. It is uncertain whether or not
house sparrows have continued to arrive via cargo ships
from mainland Chile.
In addition to house sparrow detections, shiny cowbirds
(Molothrus bonariensis) have been detected through the
citizen observers network (15 March 2016, two individuals
detected and removed; 20–24 April 2017, two individuals
detected, one removed; Hagen, unpublished data).
Stakeholder communications
Fig. 1 Results of house sparrow removals over time
(month/day/year), as well as detections from point count
estimates and acoustic recordings. The cumulative total
of house sparrows removed is presented (A) over the
same time period that weekly density estimates were
calculated from point count observations (B) and mean
call rates by house sparrows (C), reported as averages
over the previous survey week (McKown, 2013).
after 15 August 2012, while ARUs continued to detect
house sparrow activity for 10 additional days. Both point
counts and automated surveys failed to detect individual
house sparrows known to be present by opportunistic
observations on the island in early September 2012; neither
point counts nor automated acoustic surveys were effective
detection methods at low sparrow densities. Reports and
observations made by community members were initiated
in July 2012 and continue to date. These observations are a
critical component of visual observations as they increase
the effective coverage of the dedicated eradication team in
area as well as time. In 2014 observations were also being
made through the smartphone network, as well as through
personal communications.
A total of 1,179 hours of acoustic recordings were
collected and analysed from July to September 2012.
All 79,822 events detected as potential house sparrow
vocalisations were manually reviewed to confirm accuracy.
Mean house sparrow acoustic activity, at all surveyed sites
with data, declined from an average of 0.3 calls per minute
in July 2012 to no calls by the end of August when a low
number of individual house sparrows remained on the
island (McKown, 2013).
292
Dedicated efforts for regular, personalised and
transparent communications about the trial and its goal
to benefit native biodiversity were invested before,
during and after the trial. Emphasis was given towards
communications with homeowners at or near removal
sites, as well as broad community-wide communications
to minimise misinformation. Project personnel questioned
while working always provided community members
their attention, answering questions and continuing
conversations as needed. A dedicated outreach coordinator
led interactions with site owners and local institutions,
served as primary point of contact for stakeholder concerns
and provided regular updates to stakeholders regarding
trial status and advances.
DISCUSSION
The house sparrow has aggressive foraging and nesting
behaviour towards native bird species and is one of the
most widespread invasive bird species throughout the
world (Anderson, 2006). The house sparrow population
expansion on Robinson Crusoe Island caused concerns
for impacting vulnerable island endemic birds such as the
Juan Fernández firecrown (Sephanoides fernandensis) and
the Juan Fernández tit-tyrant (Anairetes fernandezianus),
species which already co-occur with house sparrows
(Hahn, et al., 2005). Given that house sparrows were proactively eliminated from neighbouring Alejandro Selkirk
Island in 1994 (Hahn, et al., 2009), there was local interest
in their removal from Robinson Crusoe Island while they
were still restricted to one area of the island.
Worldwide, invasive bird eradications have received
criticism for perhaps not being the highest need or having
substantial evidence related to their impacts (Strubbe, et
al., 2011). The precautionary principle may be invoked in
decisions of eradicating potential threats before ecological
Hagen, et al.: House sparrow, Robinson Crusoe Island, Chile
damage is documented and the invasive bird establishes a
population; in fact, this early action may be the only option
for removing highly mobile bird species in some places
and can definitely be the most economical option (IUCN,
2000; Baker, et al., 2014; Martin-Albarracin, et al., 2015).
On Robinson Crusoe Island, house sparrow eradication
and related activities as a community engagement and
invasive species awareness-building technique for a
broader invasive species programme (Glen, et al., 2013)
were used. By working within the island’s only town and
with dedicated transparent communications focused on
native species conservation, a coalition of homeowners was
built that not only actively asked questions about invasive
species management, but also contributed observations
regularly to an early observer’s network. This network
has grown and today is formalised as an early detection
network, continuing to rely on individuals’ observations of
invasive species by word of mouth, phone and smartphone
application as a critical part of invasive species management
(Ministerio del Medioambiente, 2017).
At the conclusion of the trial in late 2012, local
decision-makers were interested in completing the house
sparrow eradication, however the only detection techniques
effective at low population densities are opportunistic
visual observations. A wide network of citizen observers
has successfully indicated presence and locations of house
sparrows on Robinson Crusoe in following years; however
detailed observations that lead to successful removals
require effort-intensive follow-up by specialized personnel.
Follow-up trapping has not been successful, however
removals by shooting have occurred. It is unclear how
many individuals remain, however they tend to be reported
in the period from October to January. Multiple methods
were ineffective at detecting the presence of remaining
house sparrows at low population density, complicating
the ability to assess eradication success probability
without considerable observer effort across the island.
Statstical frameworks developed to assess the probability
of eradication confirmation success for other species
may lend themselves to adjustments for invasive bird
eradications and should continue to be explored (Ramsey,
et al., 2011; Samaniego-Herrera, et al., 2013). There is
no local institution able to dedicate staff to responding
to observations, and so reported sparrow sightings and
opportunistic removals are recorded in an exotic species
database, including detections and removals from Alejandro
Selkirk Island in 2016. The complete removal of house
sparrows from the Juan Fernández Archipelago is possible
with continued observations and removals; however, the
arrival of additional individuals from continental sources
via cargo boats is likely as no formal biosecurity measures
exist and established municipal ordinances cannot restrict
these movements. Persistent threats to native avifauna
from introduced species continue to exist in the absence
of formal biosecurity and environmental protection
legislation.
Worldwide we are aware of at least 23 documented bird
eradication attempts (DIISE, 2016, using data classified as
good or satisfactory quality, and whole island eradications
only). Bird eradication projects are more challenging
compared to mammal eradications because volant birds
fly more readily between adjacent islands, leading to
higher rates of reinvasion (thus necessitating definition of
eradication units for eradication planning, e.g. Robertson
& Gemmell, 2004; Abdelkrim, et al., 2005) and it is often
harder to define if treatment of the whole island or only
part of the island is required. Recently, six successful bird
eradications in the Seychelles were implemented (Bunbury,
et al., 2019) adding to the global knowledge pool for
planning and implementing invasive bird eradications. We
are aware of only two other attempts to eradicate invasive
house sparrows from island habitats, an unsuccessful
attempt on Round Island in Mauritius (Bednarczuk, et
al., 2010), and a successful attempt of a restricted range
population on Mahe in the Seychelles, where repeated
invasions (due to international ship traffic) are treated on
an ongoing basis (Beaver & Mougal, 2009).
ACKNOWLEDGEMENTS
Thanks to J.L. Herrera, W. Jolley, A. de Rodt, D.
Arredondo, R. Recabarren, M. Pérez, E. Oberg, U.
Partarrieu, C. López, P. González and S. de Rodt for
assisting in planning, logistics, and the field work
conducted to implement this project. I. Leiva and the park
guards at Parque Nacional Archipiélago Juan Fernández,
especially B. López, M. Recabarren and D. Arredondo,
provided support throughout the work described here.
Island Conservation’s eradication advisory team provided
peer review of plans. Matthew McKown and Conservation
Metrics provided excellent support with detection analysis.
Permits were provided by Servicio Agrícola y Ganadero
as well as the Corporación Nacional Forestal. Thanks also
to residents of the Juan Fernández Islands who provided
information on sparrow sightings, and landowners within
San Juan Bautista who provided access to their land. We are
grateful for the support and interest from the Municipality
of Juan Fernández. This project received financial support
from Island Conservation, The David and Lucile Packard
Foundation, and National Fish & Wildlife Foundation.
Thanks to two anonymous reviewers for constructive
feedback on an earlier draft of this manuscript.
REFERENCES
Abdelkrim, J., Pascal, M. and Samadi, S. (2005). ‘Island colonization and
founder effects: The invasion of the Guadeloupe Islands by ship rats
(Rattus rattus)’. Molecular Ecology 14: 2923–2931.
Anderson, T.R. (2006). Biology of the Ubiquitous House Sparrow: From
Genes to Populations, Oxford: Oxford University Press.
Arroyo, M.T.K. (1999). ‘The Juan Fernández Islands’. In: R.A.
Mittermeier, N. Myers, G.P. Robles and C.G. Mittermeier (eds.)
Hotspots: Earth’s Biologically Richest and Most Endangered Terrestrial
Ecoregions, p. 173. Mexico City: CEMEX.
AVMA (2007). ‘AVMA Guidelines on Euthanasia’. Journal of the
American Veterinary Medical Association, p. 14. American Veterinary
Medical Association.
Baker, J., Harvey, K.J. and French, K. (2014). ‘Threats from introduced
birds to native birds’. Emu 114: 1–12.
Beaver, K. and Mougal, J. (2009). Review of IAS Control and Eradication
Programmes in Seychelles. Seychelles: Plant Conservation Action
Group. Unpublished report. <http://www.pcusey.sc/index.php/
component/remository/func-startdown/107/?Itemid=>
Bednarczuk, E., Feare, C., Lovibond, S., Tatayah, V. and Jones, C. (2010).
‘Attempted eradication of house sparrows Passer domesticus from
Round Island (Mauritius), Indian Ocean’. Conservation Evidence 7:
75–86.
Bibby, C.J., Burgess, N.D., Hill, D.A., and Mustoe, S.H. (2000). Bird
Census Techniques. London: Academic Press of Elsevier.
Bourne, W.R.P., Brooke, M.d.L., Clark, G.S. and Stone, T. (1992).
‘Wildlife conservation problems in the Juan Fernández Archipelago,
Chile’. Oryx 26: 43–51.
Buckland, S.T., Anderson, D.R., Burnham, K.P., Laake, J.L., Borchers,
D.L. and Thomas, L. (2001). Introduction to Distance Sampling
Estimating Abundance of Biological Populations. Oxford, UK: Oxford
University Press.
293
Island invasives: scaling up to meet the challenge. Ch 2B Other taxa: Birds
Bunbury, N., Haverson, P., Page, N., Agricole, J., Angell, G., Banville,
P., Constance, A., Friedlander, J., Leite, L., Mahoune, T., MeltonDurup, E., Moumou, J., Raines, K., van de Crommenacker, J. and
Fleischer-Dogley, F. (2019) ‘Five eradications, three new species,
three islands: overview, insights and recommendations from invasive
bird eradications in the Seychelles’. In: C.R. Veitch, M.N. Clout, A.R.
Martin, J.C. Russell and C.J. West (eds.) Island invasives: scaling up to
meet the challenge, pp. 282–288. Occasional Paper SSC no. 62. Gland,
Switzerland: IUCN.
Clark, J.P. and Hygnstrom, S.E. (1994). ‘Crowned Sparrows’. In: S.
Hyngstrom, R.M. Timm and G. Larson (eds.) Prevention and Control of
Wildlife Damage, pp. E97-E100. Lincoln, NE: University of Nebraska
Cooperative Extension Service.
Ministerio del Medioambiente (2017). Plan de Acción del Archipiélago
Juan Fernández para la Prevención, Control y/o Erradicación de
Especies Exóticas Invasoras. Santiago, Chile: Imprenta Grafic Suisse.
Morrison, S.A., Macdonald, N., Walker, K., Lozier, L. and Shaw, M.R.
(2007). ‘Facing the dilemma at eradication’s end: Uncertainty of absence
and the Lazarus effect’. Frontiers in Ecology and the Environment 5:
271–276.
Ramsey, D.S.L., Parks, J.P., Will, D., Hanson, C.C., and Campbell, K.J.
(2011). ‘Quantifying the success of feral cat eradication, San Nicolas
Island, California’. New Zealand Journal of Ecology 35: 163–173.
Robertson, B.C. and Gemmell, N.J. (2004). ‘Defining eradication units to
control invasive pests’. Ecology 41: 1042–1048.
Clergeau, P., Levesque, A. and Lorvelec, O. (2004). ‘The precautionary
principle and biological invasion: The case of the house sparrow on the
Lesser Antilles’. International Journal of Pest Management 50: 83–89.
Samaniego-Herrera, A., Anderson, D.P., Parkes, J.P. and Aguirre-Muñoz,
A. (2013). ‘Rapid assessment of rat eradication after aerial baiting’.
Journal of Applied Ecology 50: 1415–1421.
DIISE. (2016). The Database of Island Invasive Species
Eradications.<http://diise.islandconservation.org>.
Accessed
16
February, 2016.
Sanders, R.W., Stuessy, T.F. and Marticorena, C. (1982). ‘Recent changes
in the flora of the Juan Fernandez Islands, Chile’. Taxon 31: 284–289.
Fagerstone, K.A. (2007). ‘Mitigating Impacts of Terrestrial Invasive
Species’. In: D. Pimentel (ed.) Encyclopedia of Pest Management,
Volume II, pp. 347–349. Boca Raton: CRC Press.
Saunders, A., Glen, A., Campbell, K., Atkinson, R., Sawyer, J., Hagen, E.
and Torres, H. (2011). Feasibility of Managing Invasive Species in the
Juan Fernández Archipelago, Chile. Unpublished report. Lincoln, New
Zealand: Invasive Species International.
Fisher, P., Booth, L. and Campion, M. (2012). Review of Avicides for
Use in Eradication of Introduced Sparrows and Pigeons from the Juan
Fernández Islands, Chile. Unpublished report. Lincoln: Landcare
Research.
Sharp, T. and Saunders, G. (2005). ‘BIR002: Trapping of Pest Birds’. In:
Sharp, T. and Saunders, G. (eds.) Humane Pest Animal Control: Codes
of Practice and Standard Operating Procedures, p. 11. New South
Wales, Australia: NSW Department of Primary Industries.
Fitzwater, W. (1981). ‘Outwitting the House Sparrow [Passer domesticus
(Linnaeus)]’. Great Plains Wildlife Damage Control Workshop
Proceedings, pp. 244–251. Lincoln, NE: University of Nebraska.
Strubbe, D., Shwartz, A. and Chiron, F. (2011). ‘Concerns regarding the
scientific evidence informing impact risk assessment and management
recommendations for invasive birds’. Biological Conservation 144:
2112–2118.
Fitzwater, W. (1994). ‘House Sparrows’. In: S. Hyngstrom, R.M. Timm
and G. Larson (eds.) Prevention and Control of Wildlife Damage,
pp. E101-E108. Lincoln, NE: University of Nebraska Cooperative
Extension Service.
Glen, A.S., Atkinson, R., Campbell, K.J., Hagen, E., Holmes, N.D.,
Keitt, B.S., Parkes, J.P., Saunders, A., Sawyer, J. and Torres, H. (2013).
‘Eradicating multiple invasive species on inhabited islands: The next
big step in island restoration?’ Biological Invasions 15: 2589–2603.
Gowaty, P.A. (1984). ‘House sparrows kill eastern bluebirds’. Journal of
Field Ornithology 55: 378–380.
Greimler, J., Lopez S., P., Stuessy, T.F. and Dirnböck, T. (2002). ‘The
vegetation of Robinson Crusoe Island (Isla Masatierra), Juan Fernández
Archipelago, Chile’. Pacific Science 55: 263–284.
Hahn, I. and Römer, U. (2002). ‘Threatened avifauna of the Juan
Fernández Archipelago, Chile: The impact of introduced mammals and
conservation priorities’. Cotinga 17: 66–72.
Hahn, I., Römer, U. and Schlatter, R. (2005). ‘Distribution, habitat use,
and abundance patterns of landbird communities on the Juan Fernández
Islands, Chile’. Ornitología Neotropical 16: 371–385.
Hahn, I., Römer, U. and Schlatter, R. (2006). ‘Population numbers and
status of land birds of the Juan Fernandez Archipelago, Chile: (Aves:
Falconiformes, Columbiformes, Strigiformes, Caprimulgiformes,
Passeriformes)’. Senckenbergiana Biologica 86: 109–125.
Hahn, I., Römer, U., Vergara, P. and Walter, H. (2009). ‘Biogeography,
diversity, and conservation of the birds of the Juan Fernández Islands,
Chile’. Vertebrate Zoology 59: 103–114.
IUCN. (2000). ‘IUCN Guidelines for the Prevention of Biodiversity
Loss Caused by Alien Invasive Species’. < http://www.issg.org/pdf/
guidelines_iucn.pdf>. Accessed 10 May, 2017.
MacGregor-Fors, I., Morales-Perez, L., Quesada, J. and Schondube, J.E.
(2010). ‘Relationship between the presence of house sparrows (Passer
domesticus) and neotropical bird community structure and diversity’.
Biological Invasions 12: 87–96.
Martin-Albarracin, V.L., Amico, G.C., Simberloff, D. and Nuñez, M.A.
(2015). ‘Impact of non-native birds on native ecosystems: A global
analysis’. PLOS ONE 10: e0143070.
McGillivray, W.B. (1980). ‘Communal nesting in the house sparrow’.
Journal of Field Ornithology 51: 371–372.
McGregor, D.B. and McGregor, B.A. (2008). ‘Eliminating an avian pest
(house sparrow Passer domesticus) population: The role of trapping at a
homestead scale’. Victorian Naturalist 125: 4–10.
McKown, M. (2013). Acoustic Monitoring of House Sparrows Eradication
Trials on Isla Robinson Crusoe, Chile Final Report. Unpublished report.
Santa Cruz, California: Conservation Metrics.
294
B.J. Hughes, R.C. Dickey and S.J. Reynolds
Hughes, B.J.; R.C. Dickey and S.J. Reynolds. Predation pressures on sooty terns by cats, rats and common mynas on Ascension Island in the South Atlantic
Predation pressures on sooty terns by cats, rats and common mynas
on Ascension Island in the South Atlantic
B.J. Hughes1,2, R.C. Dickey1 and S.J. Reynolds1,2
Army Ornithological Society, c/o Prince Consort Library, Knollys Road, South Camp, Aldershot, Hampshire GU11
1PS, UK. <rasuk@btconnect.com>. 2Centre for Ornithology, School of Biosciences, College of Life & Environmental
Sciences, University of Birmingham, Edgbaston, Birmingham B15 2TT, UK.
1
Abstract Despite the presence of invasive black rats (Rattus rattus), common mynas (Acridotheres tristis), and feral
domestic cats (Felis catus), sooty terns (Onychoprion fuscatus) breed in large numbers on Ascension Island in the tropical
South Atlantic Ocean. These introduced predators impact the terns by destroying eggs or interrupting incubation (mynas),
eating eggs (mynas and rats), eating chicks (rats and cats), or eating adults (cats). Between 1990 and 2015, 26 censuses of
sooty terns and five of mynas were completed and myna predation was monitored on 10 occasions. Rat relative abundance
indices were determined through trapping around the tern colonies and rat predation was monitored by counting chick
carcasses. Cat predation was quantified by recording freshly killed terns. Prior to their eradication in 2003, cats had the
greatest impact on sooty terns and were depredating 5,800 adults and 3,600 near-fledging chicks (equivalent to the loss
of 71,000 eggs) each breeding season. We estimated that 26,000 sooty tern eggs (13% of all those laid) were depredated
by approximately 1,000 mynas. Rats were not known to depredate sooty terns prior to cat eradication but in 2005, 131 of
596 ringed (monitored) chicks (22%) were depredated by rats. In 2009 chick carcass density was 0.16 per m2. Predation
by rats hugely increased in the absence of cats and was the equivalent of 69,000 eggs. Care is needed when applying our
findings to seabirds globally. The scarcity of alternative food sources and seasonally high density of easily available prey
in the sooty tern colony may have magnified predation by cats, rats and mynas.
Keywords: Non-native species, population size, predation rate, United Kingdom Overseas Territory (UKOT)
INTRODUCTION
Comparative studies of global declines in faunal
biodiversity have concluded that harvesting, habitat
loss and introduced invasive species are leading causes
(see refs in Young, et al., 2016). Of extinction events for
which causes have been investigated, 54% have been
attributed in part to invasive species (Clavero & GarciaBerthou, 2005). Globally, terrestrial invertebrate invaders
have reduced faunal diversity by 29% (Cameron, et al.,
2016). Lowe, et al. (2000) compared the severity of alien
species on animal and plant diversity by compiling a list
entitled “100 of the world’s worst invasive alien species”.
The list includes invasive predators that are commensal
with man; they pose major threats to seabirds and they
persist following anthropogenic introduction to 90% of all
island archipelagos (Towns, et al., 2006). Global seabird
population size has declined by 70% between 1950 and
2010 (Paleczny, et al., 2015) with introduced commensal
predators being one of the major proposed causes of such
declines (Moors & Atkinson, 1984).
Of introduced predators, feral domestic cats (Felis
catus) (Medina, et al., 2011) and black rats (Rattus rattus)
(Jones, et al., 2008) inflict the most severe impacts on
native avifauna. Common mynas (Acridotheres tristis)
(hereafter referred to as ‘mynas’) also have significant
negative impacts on native avifauna through competition
for food and nest sites (Grarock, et al., 2012). When
cats, rats and mynas invade islands on which seabirds are
breeding, cats have a direct effect on the size of the seabird
population through predation of adults (van Aarde, 1983)
while rats and mynas have a less immediate, but a more
indirect, effect through predation of eggs or chicks (Jones,
et al., 2008). Therefore, rats and mynas reduce breeding
success and inflict downstream impacts on seabird
demography through reduced recruitment to the breeding
population (Harper & Bunbury, 2015). The direct impacts
of cats on breeding seabirds are more readily observed
than the indirect effects of rats and mynas that are more
difficult to quantify because rat and myna predation is less
obvious and is confounded by rats scavenging on chicks
that have died from causes other than direct predation (e.g.
starvation).
Alien invasive predators are the potential cause of
precipitous declines in the population size of breeding
sooty terns (Onychoprion fuscatus) on Ascension Island
during the 20th century (Hughes, et al., 2017a). Sooty terns
are the most numerous avian species in tropical waters
and Ascension Island accommodates the largest breeding
population in the Atlantic (Schreiber, et al., 2002). Three
of the world’s ‘worst’ invasive predators are found in the
seabird colonies on Ascension Island. Black rats probably
arrived when HMS Roebuck was abandoned close to the
island in 1701 (Ashmole & Ashmole, 2000), and by 1725
rats were so numerous that a castaway on the island lived in
fear of being eaten alive (Ritsema, 2006). In 1815 domestic
cats were introduced to control the rat population. Mynas
were introduced in the 1880s to reduce damage to crops by
black cutworms (Agrotis ipsilon) (Duffey, 1964). Common
mynas in their home range (i.e. India) are regarded as a
beneficial species (BirdLife International, 2015) because
typically more than 80% of their food mass comprises insects
regarded as pests (e.g. cutworms – larvae of Noctuidae).
Since the arrival of these invasive species on the island, the
once vast colonies of seabirds, estimated to contain > 10
million birds (Ashmole & Ashmole, 2000), have dwindled
to less than half a million birds (Bell & Ashmole, 1995).
Of the 11 seabird species that breed on Ascension Island,
only sooty terns now breed in large numbers on the main
island. Numerically, 97% of all seabirds breeding on the
main island are sooty terns (Hughes, 2014). Remnant
populations of other seabird species nest on cat- and ratfree offshore stacks and Boatswainbird Islet (Ratcliffe, et
al., 2009).
In 1958 and 1959, cats were the only non-native
predatory species known to depredate seabirds and an
aspiration for their eradication was conceived (Ashmole,
1963). During a feasibility study for cat eradication in
1992, rats were also considered a major threat to seabirds
(Ashmole, et al., 1992) but the threat that mynas posed
was not recognised at that time. More recently, in the
Seychelles, Feare, et al. (2015) recorded mynas inflicting
intense predation on seabird eggs. On Ascension Island
cats were eradicated in 2003 (Bell & Boyle, 2004) and
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 295–301. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
295
Island invasives: scaling up to meet the challenge. Ch 2B Other taxa: Birds
rat control measures (Pickup, 1999) were implemented.
The eradication of apex predators is generally associated
with an increase in the abundance of smaller predators
with this trophic interaction referred to as ‘mesopredator
release’ (Prugh, et al., 2009). However, Russell, et al.
(2009) modelled the effects of mesopredator release and
concluded that the negative impact of more mesopredators
is outweighed by the benefit of apex predator removal,
allowing recovery of prey populations. If we apply their
conclusions to Ascension Island then cat eradication
should have resulted in an increase in the population size
of sooty terns but, to date, no such effect has been detected
(Hughes, et al., 2017a).
Prey population size
Here, we have collated data from published outputs
and from a 25-year Army Ornithological Society (AOS)
dataset on introduced species to calculate the relative
impacts of cat, rat and myna predation on the sooty tern
breeding population.
Myna population size and their predation pressure on
sooty terns
METHODS
Study area and period
Ascension (07º57′S, 14º24′W, 97 km²) is one of the
volcanic islands that make up the UK Overseas Territory
(UKOT) of St Helena, Ascension and Tristan da Cunha,
and is isolated in the tropical South Atlantic Ocean midway
between South America and Africa (Fig. 1; Hughes, et al.,
2010). Its nearest neighbour is the island of St Helena some
1,300 km to the south-east. The territory is an Important
Bird Area (IBA reference number SH009; BirdLife
International, 2017). More than half of its surface consists of
cinder plains, ash cones and basaltic lava flows. The average
annual rainfall is 144.0 mm (Anon., 1998) and plant species
richness on the plain is < 11 species (Duffey, 1964). The dry
coastal plain is the traditional nesting site for seabirds and
sooty terns nest at Mars Bay and Waterside in the southwest corner of the island (Fig. 1).
Fieldwork lasted two weeks per breeding season and
was timed to coincide with the peak of the sooty tern
breeding season (see further details in Reynolds, et al.,
2014). Time in the field amounted to 1,691 person-days.
Sooty terns, the primary avian prey species of rats, cats
and mynas, are a migratory species and are absent from
Ascension Island for approximately three in every 9.6
months that constitute the sub-annual cycle of the species
(Reynolds, et al., 2014). Modal clutch size of sooty terns is
one (Schreiber, et al., 2002). The population was censused
on 26 occasions between 1990 and 2015. We calculated
the number of eggs laid by measuring the area of the
two breeding colonies using conventional land survey
techniques and determined egg density by counting eggs in
quadrats (see full details in Hughes, et al., 2008).
Censuses of the myna population were obtained from
a consolidation of counts in 1994, 2004, 2005, 2006 and
2015 and included counts of birds feeding on the two
rubbish tips, in 116 1-km grid squares covering the whole
island, and at night-roosts.
Rates of predation were estimated by marking focal
sooty tern eggs and following their fates. In each sooty
tern breeding season, egg predation by mynas was
measured for approximately seven days (i.e. for 25% of the
incubation period of 28.8 days; Ashmole, 1963) and mean
egg failure rates for the core and periphery of the colony
(Hughes, et al., 2008) were calculated using the Mayfield
method (Johnson & Shaffer, 1990). Causes of egg failure
were categorised according to egg damage likely caused
by mynas: ‘consumption’ was defined as the opening of
a viable egg and feeding on some (usually < 10%) of the
contents, and ‘puncturing’ was defined as the creation of
a single small hole that destroyed egg integrity. The ratio
of consumed:punctured sooty tern eggs was obtained from
quadrat counts of depredated eggs. To establish causation
of egg desertion, sets of focal eggs that contained deserted
eggs were separated into two categories: those containing
eggs consumed or punctured by mynas, and those that did
not. We had previously found that the apparent association
between these egg fates was significant (see full details in
Hughes, et al., 2017b).
Cat population size and their predation pressure on
sooty terns
On Ascension Island the cat population size in 1958
was estimated to be in the hundreds (Ashmole, 1963). Of
the 1,100 feral cats that were removed from the island in
the eradication programme of 2002, approximately 50
were removed from the tern colonies (Bell & Boyle, 2004).
Predation of adult sooty terns was monitored by
removing all corpses of terns from the breeding area and
then re-visiting the colonies to record the number of freshly
killed birds. The mortality data gathered over two weeks
may sometimes under-estimate the level of predation.
Ashmole (1963) found that towards the end of each
season cats began to take large chicks as well as adults. To
compensate for this unknown level of chick predation, cats
were assumed to take equal numbers of adults and chicks
for 110 days (i.e. the period when some adults incubate
while others feed chicks close to fledging; see full details
in Hughes, et al., 2008). Because cats have been observed
consuming seabird eggs elsewhere, albeit on rare occasions
(Plantinga, et al., 2011), we also assessed this source of egg
loss by inspecting cat middens for cat-predated eggs.
Fig. 1 Map of Ascension Island in the South Atlantic
showing sites of human habitation and ground above
300 m (shaded). Sooty terns nest in the south-west
corner of the island in the areas marked as ‘Mars Bay’
and ‘Waterside’.
296
Hughes, et al.: Predation of sooty terns on Ascension Island
Rat population size and their predation pressure on
sooty terns
The size of the rat population on Ascension Island has
not been estimated but anecdotal data indicate that it has
been (and remains) large. For example, 70,148 rats were
killed between 1878 and 1887 (Hart-Davis, 1972). Relative
abundance of rats in the tern colonies was estimated using
a simple index calculated as the number of rat captures
per 100 trap-nights (C/100TN) corrected for traps tripped
(after Cunningham & Moors, 1983). During field seasons
50 ‘Victor’ break-back rat traps baited with peanut butter
and cornflakes were set out in pairs along the edge of both
tern breeding colonies. Nest density was too high to allow
traps to be set within the colonies without significantly
disturbing breeding birds. Trapping occurred over two
consecutive nights.
We studied the rate of chick predation by rats by
counting chick carcasses (Towns, et al., 2006). We
eliminated the possibility that starvation was the ultimate
cause of chick mortality by recording the muscle score and
body mass of live chicks in the same parts of the colonies
as carcass surveys. The shape of the pectoral muscles was
scored between 0 and 2 according to the prominence of the
keel as described in Gosler (1991). A muscle score of 0 on
this scale is indicative of low body condition most likely
caused by malnourishment. Body mass of live chicks was
recorded to the nearest 1 g with a Pesola spring balance.
Chicks aged 28–30 days that were underweight weighed
approximately 80 g and those that were in higher condition
were > 150 g (Ashmole, 1963).
Prior to cat eradication in 2002 we found two cavities
in rocks on the perimeter of the tern colonies that contained
many broken sooty tern eggs but only later did we attribute
the find to rat predation. Rats will roll eggs away from
avian nests to a place of safety where they can open them
(Zarzoso-Lacoste, et al., 2011). After cat eradication, we
studied the rate of egg predation by rats by marking focal
eggs and recording their losses. The rate of egg losses to
rats was calculated as for that to mynas. We calculated the
level of egg predation by rats prior to cat eradication by
scaling up our findings from the above focal study. We used
rat indices to generate relative rat abundance estimates
before and after cat eradication.
Comparison of the three predation pressures
To evaluate the impact of chick and egg losses on the
size of the breeding population of sooty terns, ratios of
adults to chicks, and adults to eggs were required. In other
words, on average, how many eggs need to hatch, and thus
how many chicks need to survive until recruitment, to
replace one adult in the breeding population? Furthermore,
cats depredate near-fledging chicks while rats take halfgrown chicks and thus we also required a ratio of eggs to
both cat- and rat-depredated chicks.
The ratio of near-fledging chicks to adults was obtained
from demographic data and estimates of adult and juvenile
survival rates were calculated from ringing-re-capture data
of adults and near-fledging chicks that were ringed during
the same breeding seasons and re-captured in subsequent
seasons (see further details in Reynolds, et al., 2014). Adult
and juvenile survival rates, age at first breeding and mean
age of birds in the breeding population were determined
each breeding season by the re-capture from each cohort of
adults and new recruits, and a mean with a 95% confidence
limit (CL) calculated using the program MARK (White &
Burnham, 1999).
The ratio of eggs to near-fledging chicks (i.e. those
depredated by cats) was calculated from density counts
of eggs and near-fledging chicks in quadrats (Bibby, et
al., 2000; Schreiber & Burger, 2002). The ratio of eggs
to half-grown chicks (i.e. those depredated by rats) was
calculated by taking the average of near-fledging chick
survival (see above) and nestling survival rates. The age at
which nestlings leave the nest was approximately five days
(Schreiber, et al., 2002). We calculated nestling survival
rate for the five days by applying a hatchability rate of
0.91 (i.e. the number of eggs that hatched at the end of the
incubation period; after Koenig, 1982) and a predation rate
from Ascension frigatebirds (Fregata aquila) of 0.98 (i.e.
the number of nestlings that escape frigatebird predation;
BJH, unpubl. data) to the incubation success rate.
RESULTS
Sooty tern population size
Each season between 1990 and 2015, sooty terns laid
on average 180,000 ± 8,000 (1 standard error [SE]) eggs
(range: 70,000–270,000 eggs, n = 26 censuses). The mean
number of nestlings in the tern colony each season was
94,000 ± 14,000 (n = 12 breeding seasons). The mean size
of the breeding population was 360,000 ± 14,000 (95%
CL) birds (Fig. 2).
Myna population size and their predation pressure on
sooty terns
Between 1992 and 2015 the mean size of the myna
population was 935 ± 265 (95% CL) birds (Fig. 3a). We
found no evidence to suggest that mynas killed tern chicks.
Mynas were recorded every field season in the tern colonies.
Between 2000 and 2008 we monitored 1,238 eggs (935 on
the periphery and 303 in the core). Of the 331 nest failures
at the periphery of the colonies, 87 (26.3%) failed as a
direct result of mynas. We calculated the mean egg failure
rate at the periphery of the colonies as being 0.35 ± 0.07 (±
1 SE) eggs per season (n = 10 breeding seasons). The core
of each colony appeared largely free from egg predation
by mynas. The mean rate of egg loss to mynas in the two
colonies was 0.19 ± 0.04 eggs per pair of terns (range:
0.02–0.37 eggs per pair, n = 1,238 breeding pairs over 10
breeding seasons). The ratio of consumed:punctured sooty
tern eggs was 1:1.83 (n > 500 eggs in five sample quadrats
across three breeding seasons). In summary, of all sooty
tern eggs lost to mynas, 21% were consumed, 39% were
punctured and 40% were deserted. We calculated that
sooty tern mean egg losses to myna predation per season
amounted to 26,000 ± 12,000 eggs (range: 4,000–50,000
eggs) that represented an average of 13% of all eggs laid (n
= 10 breeding seasons).
Fig. 2 Estimated size of the sooty tern breeding population
(mean + 95% confidence limits) on Ascension Island
between 1990 and 2015. Filled columns are censuses
carried out during the cat-rat-myna (three) predator
regime while open columns are those conducted during
the rat-myna (dual) predator regime. Note that the subannual breeding cycle results in birds breeding twice in
1996, 2004, 2008, 2012 and 2015.
297
Island invasives: scaling up to meet the challenge. Ch 2B Other taxa: Birds
three breeding seasons). Towards the end of the sooty tern
breeding season cats were killing near-fledging chicks
as well as adults. If cat predation continued at the same
intensity in the second half of the season as in the first, the
overall percentage of the adult population depredated by
cats would have been 1.8% (or 5,800 birds on average, n =
3 breeding seasons). Predation of chicks was not monitored
but we estimated that the overall percentage of chicks that
were depredated or died of starvation because a parent was
killed by cats, was 3,600 chicks (i.e. 29 cat kills per night
for the four months that chicks were in the colony, yielding
a total of 3,600 chicks, 3.8% of the chick population of
94,000). We found no evidence that feral cats were taking
any sooty tern eggs.
Rat population size and their predation pressure on
sooty terns
During 473 days of fieldwork prior to the eradication of
cats, no rat predation of tern chicks was observed. Between
1992 and 2002 the mean relative abundance of rats pre-cat
eradication on the dry coastal plain close to the two tern
colonies was 1.3 ± 1.0 (± 1 SE) C/100TN (range: 0–6.0
C/100TN, n = 6 trap-lines over three breeding seasons).
Between 2005 and 2015 after the cat eradication the mean
relative abundance of rats was 15.2 ± 3.8 C/100TN (range:
0–74.5 C/100TN, n = 25 trap-lines over 12 breeding
seasons) (Fig. 3c).
Carcasses of chicks depredated by rats were first
observed in 2005 when 131 of 596 ringed chicks (22.0%)
were depredated. In 2009 mean carcass density in quadrats
was 0.16 (range: 0–0.9 per m2, n = 68 quadrats). The area of
the colony was 12.21 ha and thus it contained an estimated
19,500 ± 27,000 carcasses (20%) of the chick population.
We found no evidence of mass starvation as live chicks had
a mean muscle score of 1.05 ± 0.31 (range: 0–2, n = 998
chicks) and a mean body mass of 157.5 ± 29.2 g (range:
54.8–220.0 g, n = 946 chicks). In 2005 and 2009 the mean
number of chicks depredated by rats was 20,000 (21% of
the chick population).
Fig. 3 Population sizes of the three introduced species
that prey on various life stages of sooty terns breeding
on Ascension Island in the South Atlantic: (a) common
mynas, (b) domestic cats, and (c) black rats. Note that in
(a) and (b) population estimates are based upon wholeisland counts while (c) provides a relative abundance
index calculated as the mean number of rats captured
(C) per 100 trap-nights (TN) at both colonies (see further
details in the Methods).
Cat population size and their predation pressure on
sooty terns
We were unable to quantify cat numbers in the sooty tern
colonies. Estimates of the size of the feral cat population
across the whole island are shown in Fig. 3b. The number
of adult sooty terns killed by cats from collection of corpses
in 1990, 1992 and 1994 amounted to 2,996, 340 and 310,
respectively. The mean number of adults killed by cats
in the two colonies was 29 per night (n = 32 nights over
298
During nine sooty tern breeding seasons between 2003
and 2012, we monitored 1,067 single egg clutches (792
on the periphery and 275 in the core) for rat predation. Of
the 327 nest failures, 314 were on the periphery and 13
were in the core of the colonies. Of the 327 that failed, 51
(15.6%) were missing eggs and these were attributed to rat
predation. The mean rate of egg loss to rats was 0.17 ± 0.06
(± 1 SE) eggs per pair of terns (range: 0.00–0.49 eggs per
pair, n = 1,067 breeding pairs over nine breeding seasons).
Assuming rats only depredated eggs at the periphery of
the colonies as so few eggs failed in the core, the overall
percentage of the eggs depredated in the tern colony by
rats was 4.8% representing an egg total of 9,000 (n = 9
breeding seasons).
From a comparison of mean relative abundances of rats
pre- (1.3 C/100TN, n = 3 breeding seasons) and post-cat
eradication (15.2 C/100TN, n = 12 breeding seasons), we
estimated that the rat population was only 8.6% as large
prior to, compared with after, the cat eradication. We also
estimated that rats depredated 800 eggs per season prior to
cat eradication.
Comparison of the three predation pressures
A summary of the comparative predation pressures in
terms of egg losses (i.e. the lowest currency to represent
all tern life stages) is provided in Fig. 4. We generated
ratios to eggs laid of survival estimates at various life
stages using life-history data in Hughes (2014). The ratios
generated were eggs preyed upon by mynas (1:1); chicks
succumbing to rat predation (2.99:1); chicks preyed upon
by cats (4.13:1); and adults preyed upon by cats (9.67:1).
Hughes, et al.: Predation of sooty terns on Ascension Island
DISCUSSION
We compared carcass counts of adults killed by cats with
those of chicks killed by rats to assess their relative impacts
on the tern population. Carcass density of chicks killed by
rats was likely to provide an under-estimate of chick losses
as decomposition of corpses was rapid and rough ground
made it easy to overlook them. Carcass counts of adults
and near-fledging chicks killed by cats (and adjusted by
the adult:chick ratio), and those of chicks solely killed by
rats, were similar to each other, varying by 18%. If we take
into account that rats also depredated 9,000 eggs then the
variation between rat and cat predation is just 3%. Of the
three sources of predation on the tern population, cats had
the greatest impact on the tern population but following
their eradication, rats replaced them as the primary source
of predation pressure (Fig. 4).
The third source of predation on the island was mynas
that depredated 26,000 eggs every sooty tern breeding
season but their overall impact on the population size of
sooty terns was less than half that of cats or rats. Mynas
depredated more tern eggs than did rats (i.e. 26,000 versus
9,000) and very many more than did cats. Bell and Boyle
(2004) found egg remains in stomachs of one of five cats
culled close to the tern colony. We found no evidence that
mynas depredated chicks or adults. Mynas depredated more
tern eggs than rats or cats depredated chicks and before cat
eradication (i.e. pre-2002) mynas had a greater detrimental
impact on the size of the breeding tern population than
did rats. There were large variations between sooty tern
breeding seasons in the relative abundance of rats in the
tern colonies (Fig. 3c) and in the extent of egg losses to
mynas (i.e. 0.02–0.37 eggs per pair) suggesting that sooty
terns were not the main driver of the population dynamics
of these two omnivorous predators (Towns, et al., 2006).
Other comparative studies
Fig. 4 (a) Pre- and (b) post-cat eradication losses of sooty
tern egg equivalents (i.e. eggs and post-hatching chick
and adult life stages) to three introduced predator
species on Ascension Island in the South Atlantic. The
average number of eggs laid was 180,000. Note that in
(b) there are only two predators contributing to losses of
egg equivalents while the third category represents the
benefit to the tern population of cat eradication, equating
to 2,000 eggs (see further details in the Results).
The mean number of eggs lost to myna predation per
sooty tern breeding season was 26,000. We calculated
that losses of chicks to rats translated into the ‘loss’ of
60,000 eggs (i.e. 20,000 chicks × 2.99). Rats also directly
removed 9,000 eggs from the breeding colony. The sum of
rat predation translated into the ‘loss’ of 69,000 eggs. Cats
depredated 5,800 adults and 3,600 near-fledging chicks
translating into egg losses of 56,000 (i.e. 5,800 adults ×
9.67) and 15,000 (i.e. 3,600 near-fledging chicks × 4.13),
respectively. Therefore, the sum of cat predation translated
into the ‘loss’ of 71,000 eggs.
A meta-analysis by Baker, et al. (2013) of threats to
native avian species posed by introduced ones concluded
that introduced invasive avian species are not a major
threat. However, we found that mynas posed a major threat
to native sooty terns on the island (Hughes, et al., 2017b).
For every egg that mynas consumed, they punctured or
caused desertion of four others. The only quantitative
comparative study of seabird egg predation by mynas
was of 350 wedge-tailed shearwaters (Puffinis pacificus
cuneatus) on Hawaii where mynas punctured 74 (21%) of
all eggs laid during one season (Byrd, 1979).
Ashmole (1963) estimated that on Ascension Island
cats were killing approximately 10,000–20,000 sooty tern
adults (i.e. 0.5 to 1.0% of the adult population) and up to
40% of chicks in 1958 and 1959. On Juan de Nova Island
in the Mozambique Channel in the western Indian Ocean
where predator/prey constituent members were similar to
those on Ascension Island, Peck, et al. (2008) found that
cats were killing 2,205 sooty terns per week (0.1% of the
breeding population).
Prior to cat eradication, we saw no live rats in the tern
colony and we did not suspect any rat predation of tern life
stages. Similarly, Ashmole (1963) saw no such incidents of
predation by rats during numerous day and night visits to
the tern colonies in 1958 and 1959. On Juan de Nova Island
where black rats co-exist with cats, and both depredate
sooty terns, Ringler, et al. (2015) reported that losses of
sooty terns to rats were relatively low. On Ascension Island
losses of sooty terns to rats increased dramatically when
cats were eradicated (Fig. 4). The severity of the predation
post-eradication was similar to that found by Jones, et al.
(2008) in their meta-analysis of the severity of rat predation.
299
Island invasives: scaling up to meet the challenge. Ch 2B Other taxa: Birds
Mesopredation
ACKNOWLEDGEMENTS
We found that the intensity of predation by rats varied
depending on whether rats were the apex predator. Under
the cat-rat-myna predator regime, rats exerted the least
predation pressure on the tern population (Fig. 4a) but
following apex predator (cat) removal, they exerted the
greater predation pressure in the dual predator regime (Fig.
4b). Cats were eradicated from the tern colony in 2002 (Bell
& Boyle, 2004) and the rat population increased sevenfold in size following their eradication as determined from
the relative abundance index (Fig. 3c). The eradication of
cats is seen as particularly beneficial to seabirds (Nogales,
et al., 2013) but, to the best of our knowledge, this only
applies to islands without rats in the first place (e.g. Natividad
Island, Marion Island in the sub-Antarctic and Baker Island
in the Pacific). We found clear evidence that when black
rats are ‘released’ by apex predator removal the size of
the rat population increased and rats started to depredate
tern chicks. Rats as apex predators exerted a predation
pressure on terns that was 97% of that in the regime of cats
and rats. Our findings are at odds with those of Ringler,
et al. (2015) who predicted that cat eradication would be
beneficial to sooty terns. They also oppose McCreless, et
al. (2016) who found that the potential for extirpation of
seabird populations was greater in the twin predator regime
of cats and rats and they also disagree with Ratcliffe, et al.
(2009) who reported that on Ascension Island five seabird
species had re-colonized the mainland following the
eradication of cats. There are three possible explanations
for this disparity: 1) despite major changes in predator
population sizes (Fig. 3), there has been little fluctuation
in that of breeding sooty terns on the island (Fig. 2) which
suggests that predation may not be the primary driver of
the tern’s population size; 2) a change in the habitat on the
tern colonies on Ascension Island occurred concurrently
with cat eradication which rats, with their catholic diets,
took advantage of by switching to alternative food sources
such as seeds of the invasive plant mesquite (Prosopis
juliflora) (Pickup, 1999); and 3) the sub-annual breeding
cycle of sooty terns on Ascension Island may provide rats
with more opportunities to breed than if sooty terns were
breeding annually as they do elsewhere in their range
(Reynolds, et al., 2014).
This study could not have taken place without the
enthusiasm, energy and industry of more than 50 members
of the Army Ornithological Society (AOS). We owe
Colin Wearn a large debt of gratitude for ringing and recapturing sooty terns and Ali Johnson at the British Trust for
Ornithology (BTO) for calculation of the sooty tern juvenile
and adult survival rates. We are also grateful to Chris Feare
for his help and advice during the 2006 breeding season. BJH
is particularly grateful to his five children who forfeited their
inheritance so that he could follow his passion for seabirds
on Ascension Island. We are grateful to the Royal Society
for the Protection of Birds (RSPB) for their encouragement
and for providing flights for one of our team and to the
staff at the Ascension Island Government Conservation and
Fisheries Department for their ongoing assistance. We thank
two anonymous referees and the editor for their constructive
comments on previous drafts of the manuscript.
CONCLUSIONS
Care is needed when applying our findings related to
predation pressures on Ascension Island sooty terns to other
seabird species on the island and to other places in the world.
When sooty terns are present, the super-abundance of prey
as represented by eggs, chicks and adults may magnify
predation pressures. As far as we are aware, our study is
the first to provide a comparison of predation pressures
by cats, rats and mynas on seabirds. Such empirical
evidence of invasive species’ impacts on native avifauna
is critical for the prioritization of management options
directed towards introduced species (Jeschke, et al., 2014;
McCreless, et al., 2016). Here, we present strong evidence
that mynas can be major egg predators of seabirds. We
have quantified changes in predation pressures resulting
from the eradication of cats and we have highlighted that
rats in the absence of cats have impacted upon breeding
success of sooty terns sufficiently to bring into serious
question the benefits of cat eradication to the recovery of
the sooty tern breeding population on Ascension Island.
Conversely, pressures on sooty terns from predators have
declined by 3% following the removal of cats. How rats
have largely replaced the predation pressure posed by cats
following their removal and why the population of sooty
terns on Ascension Island has not recovered in response
to seabird conservation efforts to date are questions that
require considerable future investigation.
300
REFERENCES
Anon. (1998). Metchat – The Ascension Island Climate. Ascension Island:
Meteorological Office.
Ashmole, N.P. (1963). ‘The biology of the wideawake or sooty tern
Sterna fuscata on Ascension Island’. Ibis 103b: 297–364.
Ashmole, P., Ashmole, M. and Simmons, K.E.L. (1992). Seabird
Conservation and Feral Cats on Ascension Island, South Atlantic. ICBP
Technical Publication No.1, pp. 94–121. Cambridge, UK: ICBP.
Ashmole, N.P. and Ashmole, M. (2000). St Helena and Ascension Island:
A Natural History. Oswestry, UK: Antony Nelson.
Baker, J., Harvey, K.J. and French, K. (2013). ‘Threats from introduced
birds to native birds’. Emu 114: 1–12.
Bell, B.D. and Ashmole, N.P. (1995). The Feasibility of Eradication of
Feral Cats and Rats from Ascension Island. Wellington, New Zealand:
Wildlife Management International Ltd.
Bell, M. and Boyle, D. (2004). The Eradication of Feral Cats from Ascension
Island. Wellington, New Zealand: Wildlife Management International Ltd.
Bibby, C.J., Burgess, N.D., Hill, D.A. and Mustoe, S.H. (2000). Bird
Census Techniques. London, UK: Academic Press.
BirdLife International (2015). ‘Species Factsheet: Acridotheres tristis’.
BirdLife International Data Zone [website]. <www.birdlife.org>.
Accessed 9 May 2015.
Birdlife International (2017). ‘Important bird areas factsheet: Wideawake
fairs’. BirdLife International Data Zone [website]. <www.birdlife.org>.
Accessed 5 July 2017.
Byrd, V.G. (1979). ‘Common myna predation on wedge-tailed shearwater
eggs’. Elepaio 39: 69–70.
Cameron, E.K., Vilà, M. and Cabeza, M. (2016). ‘Global meta-analysis of
the impacts of terrestrial invertebrate invaders on species, communities
and ecosystems’. Global Ecology and Biogeography 25: 596–606.
Clavero, M. and Garcia-Berthou, E. (2005). ‘Invasive species are a leading
cause of animal extinctions’. Trends in Ecology and Evolution 20: 110.
Cunningham, D.M. and Moors, P.J. (1983). A Guide to the Identification
and Collection of New Zealand Rodents. Wellington, New Zealand:
Department of Internal Affairs.
Duffey, E. (1964). ‘The terrestrial ecology of Ascension Island’. Journal
of Applied Ecology 1: 219–247.
Feare, C.J., Lebarbenchon, C., Dietrich, M. and Larose, S. (2015).
‘Predation of seabird eggs by common mynas Acridotheres tristis on
Bird Island, Seychelles, and its broader implications’. Bulletin of the
African Bird Club 22: 162–170.
Gosler, A.G. (1991). ‘On the use of greater covert moult and pectoral
muscle as measures of condition in passerines with data for the great tit
Parus major’. Bird Study 38: 1–9.
Grarock, K., Tidemann, C.R., Wood, J. and Lindenmayer, D.B. (2012).
‘Is it benign or is it a pariah? Empirical evidence for the impact of the
common myna on Australian birds’. PLOS ONE 7(7): e40622.
Harper, G.A. and Bunbury, N. (2015). ‘Invasive rats on tropical islands:
Their population biology and impacts on native species’. Global
Ecology and Conservation 3: 607–627.
Hughes, et al.: Predation of sooty terns on Ascension Island
Hart-Davis, D. (1972). Ascension: The Story of a South Atlantic Island.
London, UK: Constable.
Hughes, B.J., Martin, G.R. and Reynolds, S.J. (2008). ‘Cats and seabirds:
Effects of feral domestic cat Felis silvestris catus eradication on the
population of sooty terns Onychoprion fuscata on Ascension Island,
South Atlantic’. Ibis 150(Suppl. 1): 121–129.
Hughes, B.J., Martin, G.R. and Reynolds, S.J. (2010). ‘Sooty terns
Onychoprion fuscata on Ascension Island in the South Atlantic are a
reproductively isolated population’. Revista Brasileira Ornitologia 18:
194–198.
Hughes, B.J. (2014). ‘Breeding and Population Ecology of Sooty Terns
on Ascension Island’. PhD thesis. Birmingham, UK: University of
Birmingham.
Hughes, B.J., Martin, G.R., Giles, A.G. and Reynolds, S.J. (2017a).
‘Long-term population trends of sooty terns Onychoprion fuscatus:
Implications for conservation’. Population Ecology 59: 213–224.
Hughes, B.J., Martin, G.R. and Reynolds, S.J. (2017b). ‘Estimating the
extent of seabird egg depredation by introduced common mynas on
Ascension Island in the South Atlantic.’ Biological Invasions 19: 843–
857.
Jeschke, J.M., Bacher, S., Blackburn, T.M., Dick, J.T.A., Essl, F., Evans,
T., Gaertner, M., Hulme, P.E., Kühn, I., Mrugała, A., Pergl, J., Pyšek,
P., Rabitsch, W., Ricciardi, A., Richardson, D.M., Sendek, A., Vilŕ, M.,
Winter, M. and Kumschick, S. (2014). ‘Defining the impact of nonnative species’. Conservation Biology 28: 1188–1194.
Johnson, D.H. and Shaffer, T.L. (1990). ‘Estimating nest success: When
Mayfield wins’. The Auk 107: 595–600.
Jones, H.P., Tershy, B.R., Zavaleta, E.S., Croll, D.A., Keitt, B.S.,
Finkelstein, M.E. and Howald, G.R. (2008). ‘Severity of the effects of
invasive rats on seabirds: A global review’. Conservation Biology 22:
16–26.
Ringler, D., Russell, J.C. and Le Corre, M. (2015). ‘Trophic roles of
black rats and seabird impacts on tropical islands: Mesopredator
release or hyperpredation?’ Biological Conservation 185: 75–84.
Ritsema, A. (2006). A Dutch Castaway on Ascension Island in 1725.
Deventer, The Netherlands: Alex Ritsema.
Russell, J.C., Lecomte, V., Dumont, Y. and Le Corre, M. (2009).
‘Intraguild predation and mesopredator release effect on long-lived
prey’. Ecological Modelling 220: 1098–1104.
Schreiber, E.A. and Burger, J. (2002). Biology of Marine Birds. Boca
Raton, FL, USA: CRC Press.
Schreiber, E.A., Feare, C.F., Harrington, B.J., Murray, B.G. Jr, Robertson,
W.B. Jr, Robertson, M.J. and Woolfenden, G.E. (2002). ‘Sooty tern
(Sterna fuscata)’. In: A. Poole and F. Gill (eds.) Birds of North America
No. 665, pp. 1–32. Philadelphia, PA, USA: American Ornithologists’
Union.
Towns, D.R., Atkinson, I.A.E. and Daugherty, C.H. (2006). ‘Have the
harmful effects of introduced rats on islands been exaggerated?’ Biological
Invasions 8: 863–891.
Van Aarde, R.J. (1983). ‘Demographic parameters of the feral cat Felis
catus population at Marion Island’. South African Journal of Wildlife
Research 13: 12–16.
White, G.C. and Burnham, K.P. (1999). ‘Program MARK: Survival
estimation from populations of marked animals’. Bird Study 46(suppl.):
S120–S139.
Young, H.S., McCauley, D.J., Galetti, M. and Dirzo, R. (2016). ‘Patterns,
causes, and consequences of Anthropocene defaunation’. Annual
Review of Ecology, Evolution, and Systematics 47: 333–358.
Zarzoso-Lacoste, D., Ruffino, L. and Vidal, E. (2011). ‘Limited predatory
capacity of introduced black rats on bird eggs: An experimental
approach’. Journal of Zoology 285: 188–193.
Koenig, W.D. (1982). ‘Ecological and social factors affecting hatchability
of eggs’. The Auk 99: 526–536.
Lowe, S., Browne, M., Boudjelas, S. and De Poorter, M. (2000). 100 of
the World’s Worst Invasive Alien Species: A Selection from the Global
Invasive Species Database. Auckland, New Zealand: The Invasive
Species Specialist Group (ISSG) of the World Conservation Union
(IUCN).
McCreless, E.E., Huff, D.D., Croll, D.A., Tershy, B.R., Spatz, D.R.,
Holmes, N.D., Butchart, S.H.M. and Wilcox, C. (2016). ‘Past and
estimated future impact of invasive alien mammals on insular
threatened vertebrate populations’. Nature Communications 7: 12488.
Medina, F.M., Bonnaud, E., Vidal, E., Tershy, B.R., Zavaleta, E.S.,
Donlan, C.J., Keitt, B.S., Le Corre, M., Horwath, S.V. and Nogales,
M. (2011). ‘A global review of the impacts of invasive cats on island
endangered vertebrates’. Global Change Biology 17: 3503–3510.
Moors, P.J. and Atkinson, I.A.E. (1984). ‘Predation on seabirds by introduced
animals, and factors affecting its severity’. In: J.P. Croxall, P.G.H. Evans
and R.W. Schreiber (eds.) Status and Conservation of the World’s Seabirds
ICBP Technical Publication No. 2, pp. 667–690. Cambridge, UK: ICBP.
Nogales, M., Vidal, E., Medina, F.M., Bonnaud, E., Tershy, B.R.,
Campbell, K.J. and Zavaleta, E.S. (2013). ‘Feral cats and biodiversity
conservation: The urgent prioritization of island management’.
Bioscience 63: 804–810.
Paleczny, M., Hammill, E., Karpouzi, V. and Pauly, D. (2015). ‘Population
trend of the world’s monitored seabirds, 1950-2010’. PLOS ONE 10(6):
e0129342.
Peck, D.R., Faulquier, L., Pinet, P., Jaquemet, S. and Le Corre, M.
(2008). ‘Feral cat diet and impact on sooty terns at Juan de Nova Island,
Mozambique Channel’. Animal Conservation 11: 65–74.
Pickup, A.R. (1999). Ascension Island Management Plan. Cambridge,
UK: BirdLife International.
Plantinga, E., Bosch, G. and Hendriks, W. (2011). ‘Estimation of the
dietary nutrient profile of free-roaming feral cats: Possible implications
for nutrition of domestic cats’. British Journal of Nutrition 106: 35–48.
Prugh, L.R., Stoner, C.J., Epps, C.W., Bean, W.T., Ripple, W.J., Laliberte,
A.S. and Brashares, J.S. (2009). ‘The rise of the mesopredator’.
Bioscience 59: 779–791.
Ratcliffe, N., Bell, M., Pelembe, T., Boyle, D., Benjamin, R., White, R.,
Godley, B., Stevenson, J. and Sanders, S.M. (2009). ‘The eradication of
feral cats from Ascension Island and its subsequent recolonization by
seabirds’. Oryx 44: 20–29.
Reynolds, S.J., Martin, G.R., Dawson, A., Wearn, C.P. and Hughes, B.J.
(2014). ‘The sub-annual breeding cycle of a tropical seabird’. PLOS
ONE 9(4): e93582.
301
S. Saavedra Cruz and S.J. Reynolds
Saavedra Cruz, S. and S.J. Reynolds. Eradication and control programmes for invasive mynas (Acridotheres spp.)
and bulbuls (Pycnonotus spp.): defining best practice in managing invasive bird populations on oceanic islands
Eradication and control programmes for invasive mynas (Acridotheres
spp.) and bulbuls (Pycnonotus spp.): defining best practice in
managing invasive bird populations on oceanic islands
S. Saavedra Cruz1 and S.J. Reynolds2,3
Invasive Bird Management (INBIMA), P.O. Box 6009, Tenerife 38007, Canary Islands, Spain. <odisea64@hotmail.
com>. 2 Centre for Ornithology, School of Biosciences, College of Life & Environmental Sciences, University of
Birmingham, Edgbaston, Birmingham B15 2TT, UK. 3 Army Ornithological Society, c/o Prince Consort Library, Knollys
Road, South Camp, Aldershot, Hampshire GU11 1PS, UK.
1
Abstract Invasive plants and animals inflict much damage on native species and this is particularly the case on isolated
oceanic islands with high degrees of endemism. Such islands commonly are important refugia for species of high
conservation value. Some of the most pervasive and potent of invasive animal species are birds of the myna (Acridotheres)
and bulbul (Pycnonotus) genera that historically were introduced to isolated islands as biological control agents for
the management of insect pest species that can cause considerable economic damage to agricultural crops and wider
ecosystems. In this paper we consider a number of ‘successful’ eradication and control programmes targeting mynas
and bulbuls. We review the locations and taxa where 17 such programmes took place and report that the common myna
(Acridotheres tristis) has been the most heavily targeted species in eradication efforts followed by the red-whiskered
bulbul (Pycnonotus jocosus). Common mynas were also at the focus of control programmes as were jungle mynas
(Acridotheres fuscus) and red-vented bulbuls (Pycnonotus cafer). By far the most favoured method of eradication and
control was trapping whereas mist-netting was employed rarely. We discuss ‘best practice’ in planning and executing such
eradication and control programmes on oceanic islands so as to maximise their benefits to local human communities. We
outline measures that must be adopted pre-, during and post-intervention in both programme types. They include adequate
resourcing, local engagement and the integration of both traditional ecological knowledge and established conservation
theory.
Keywords: engagement, ethno-ornithology, invasive bird management, local community, multispecies approach
INTRODUCTION
The modern world is experiencing unprecedented
anthropogenic inputs that are resulting in the sixth global
wildlife extinction (Foley, et al., 2013) with concomitant
phenomena such as accelerating climate change (Crowley,
2000) and increased frequencies of invasions of alien
species (Vitousek, et al., 1997; Dukes & Mooney, 1999),
resulting in losses of biodiversity (Lowe, et al., 2000).
The establishment of early human societies resulted
in the trade of goods and services (Zeder, 2008), and
the accompanying development of transport modes and
infrastructure, such as roads and other trading routes,
resulted in commodities traded over greater distances
(Earle, 1994). Inevitably, this resulted in the movement
of species out of their native ranges into areas where
they were alien (exotic). Today, we continue to trade
goods and services internationally and in so doing we
move thousands of exotic species, approximately 10% of
which will become established as invasive (Williamson
& Fitter, 1996; Westphal, et al., 2008). By definition, an
alien species occurs outside of its natural (past or present)
range and it has dispersal potential, including any part of
it (e.g. propagules, gametes) surviving and subsequently
reproducing (Lowe, et al., 2000). Invasive species are
targeted for conservation actions because they are alien
species that become established in natural or semi-natural
ecosystems or habitats where they present problems to
native species (Colautti & MacIsaac, 2004). The impacts
of avian invasive species on ecosystems are pervasive and
enduring; they include, for example, competitive exclusion
and predation of native species, disease transmission
and dilution of native gene pools through hybridisation
(reviewed in Blackburn, et al., 2009).
Oceanic islands are known to be more susceptible
to negative impacts of exotic species compared with
continental land masses because of their increased endemism
as a result of their geographical isolation (Coblentz, 1990;
Reaser, et al., 2007; Feare, 2017). Furthermore, their
ecological fragility is magnified on smaller islands that
accommodate more simple native ecological communities
than larger ones (Donlan & Wilcox, 2008). Therefore,
conservation priorities for insular environments are often
defined by the need for effective eradication and/or control
programmes of invasive species (e.g. Dulloo, et al., 2002;
Donlan & Wilcox, 2008).
In this study we focus on two invasive avian genera (i.e.
mynas Acridotheres and bulbuls Pycnonotus), consisting of
six different species. These two genera are both represented
on the list of ‘100 of the World’s Worst Invasive Alien
Species’ (Lowe, et al., 2000), a subset of the Global Invasive
Species Database, by common mynas (Acridotheres tristis)
and red-whiskered bulbuls (Pycnonotus jocosus). The 18th
and 19th centuries saw a series of introductions of mynas to
oceanic islands as biocontrol agents to counter insect pests
that threatened agricultural production. They were also
transported to oceanic islands as cage birds. On Tutuila in
American Samoa, for example, common mynas arrived in
1980 and jungle mynas (Acridotheres fuscus) in 1985 (SSC,
unpubl. data) while on Ascension Island (Hughes, et al.,
2017) and St Helena (Burns, 2011) they were introduced in
the 19th century. Bulbuls were kept widely as caged birds
but escaped captivity on islands such as Tahiti in 1925,
Assumption in the 1970s and Tenerife and Fuerteventura
at the turn of the 21st century.
The aim of our study is to provide an account of the
characteristics of successful eradication and control
programmes and then to discuss how they can be used to
define ‘best practice’. Through material presented in the
discussion, we indicate how the conservation community
can take effective measures to combat avian invasive
species on remote oceanic islands.
METHODS
We used the following keywords – ‘common myna’,
‘bulbul’, ‘Acridotheres’, ‘Pycnonotus’, ‘trapping’, ‘control’,
‘shooting’, ‘island*’, and ‘eradication’ – in searches of
several bibliographic databases including Webspire, Web
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
302
up to meet the challenge, pp. 302–308. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Saavedra Cruz & Reynolds: Mynas and bulbuls on oceanic islands
of Knowledge, Ovid SP, Inist, Blackwell Publishing and
Science Direct to identify primary scientific literature about
management programmes for avian invasive species on
islands. References in literature cited/bibliography sections
of the resulting sources were also considered for inclusion
in our study. Whether programmes were considered further
as successful eradications or controls, or rejected followed
correspondence with programme managers to obtain
further details about the interventions (Table 1). At the time
of writing all programme managers have been contacted
and we have received responses from all but one of them.
Following inclusion in the study, the information obtained
from these programme managers combined with that in
publications was examined to assess a number of factors
determining the effectiveness of programmes: the target
species; and the numbers of birds of each species and the
methods used as part of the intervention.
red-whiskered bulbuls from two islands (Tenerife and
Assumption; Table 4). Control efforts targeting common
mynas are ongoing on North Island in Seychelles (Table
4). Short-term isolated control programmes targeting
common mynas were carried out twice on Ascension
Island and once on St Helena, each being conducted in late
2009 (Table 4). An ongoing project on Tahiti is carrying
out long-term control of common mynas and red-vented
bulbuls (Pycnonotus cafer). In the Tarawa eradication
two Acridotheres species were targeted and they were
the common myna and the jungle myna. The only multispecies long-term control programme running today is in
American Samoa where the two aforementioned myna
species and the red-vented bulbul are being successfully
targeted through trapping, with approximately 9,600 birds
being captured in two consecutive trapping campaigns.
RESULTS
Table 4 provides details of the numbers of birds of each
species that have been targets of population eradication and
control programmes in the 17 projects (see also Table 3). In
total, over 57,000 invasive birds have been captured. With
ongoing projects such as the work on, for example, Tahiti
and Tutuila (long-term control programmes) and North
Island (an eradication programme), numbers are predicted
to climb steeply in the near future. The vast majority of
birds were common mynas and most were captured in
Atiu in the Cook Islands (officially ‘eradicated’ but one
remaining bird currently being tracked; SSC, unpubl. data),
and on Tahiti in the control programme (Table 4). All but
one of the 4,606 jungle mynas were caught in the ongoing
control programme on Tutuila. The majority of redwhiskered bulbuls were caught as part of the eradication
programme on Assumption, while most red-vented bulbuls
were caught in the control programme on Tahiti (Table 4).
Literature searches
We did not consider every programme where eradication
or control of avian invasive species had been attempted
because after apparently successful removal of invasive
birds, they reappeared on some islands (Table 1). The
publications that were not considered further are detailed
in Table 2. Following exclusions of these published studies
we were left with 17 programmes (Table 3); their locations
are shown in Fig. 1.
Invasive species targeted
Common mynas have been successfully eradicated
from eight islands (Tenerife, Gran Canaria, Mallorca,
Fuerteventura, Fregate, Denis, Tarawa and Atiu), and
Total numbers of birds of each species by island
Table 1 Inclusive sets of categorisation criteria that allowed us to exclude (reject) or include published studies as successful
eradication and control programmes targeting avian invasive species on oceanic islands as a result of questioning
programme managers (see Methods for further details).
Reject
Birds were present as of April 2017
No post-intervention monitoring
No defined milestones during
intervention
Pathways of invasion remain open
No defined period of quarantine
Eradication
Birds were absent as of April 2017
Post-intervention monitoring
found no birds
Defined milestones during
intervention
Pathways of invasion closed
Defined period of quarantine
Control
Birds were present as of April 2017
Post-intervention monitoring found
reproductive birds
Defined milestones during intervention
Pathways of invasion remain open
No defined period of quarantine
Fig. 1 Locations of islands where eradication (square symbols), and control (round symbols) programmes have
been carried out to address problems of invasive myna and bulbul genera (see Tables 3 & 4 for further details).
303
Island invasives: scaling up to meet the challenge. Ch 2B Other taxa: Birds
Table 2 Details of eradication and control programmes on islands involving myna and bulbul genera, including the
location, focal species, the number of birds eradicated/controlled (where known), and notes (including references) to
explain why programmes were excluded from further consideration in our study.
Location
Ascension
Island, Atlantic
Ocean
Species
Common myna
Acridotheres
tristis
No of birds Notes (including references where available)
Seychelles,
Indian Ocean
Common myna
-
Fakaofo
(Tokelau),
Pacific Ocean
Common myna
40
Western Samoa,
Pacific Ocean
Bulbul
Pycnonotus
spp. and myna
Acridotheres spp.
Mainland
Australia
Common myna
>69,000
Moturoa Island,
Bay of Islands,
New Zealand
Common myna
45
40
6,000
Trapped birds were non-target species during feral domestic cat
(Felis silvestris catus) removal in 2004 (Hughes, et al., 2008)
Some birds remained after the eradication project ended
(Canning, 2011). Eradication was abandoned when rats were
discovered on site on Denis (Millett, et al., 2005)
Birds targeted in 2006 with their egg and nest destruction
resulting in no further sightings in 2011, but in early 2012 birds
were seen on Nukunonu Atoll, 64 km north of Fakaofo (Parkes,
2012)
Feeding of ©DCR-1339 (3-chloro-p-toluidine hydrochloride)
has been effective but to date no strategy to control avian
invasive species has been formalised and consistently
implemented island-wide
The Canberra Indian Myna Action Group (CIMAG) work,
removing birds over 11 years using volunteer trappers, has taken
place on a continental land mass and is of limited applicability
to oceanic islands
No detailed results were reported from trapping which has been
criticised as an inappropriate method to control this species
(Parkes, 2012)
Table 3 Island groups and islands where eradication and control programmes targeting myna and bulbul
genera were carried out and the year when they ended and started, respectively.
Eradication
Island group
Island
Balearic Islands
Mallorca
Canary Islands
Fuerteventura
Canary Islands
Gran Canaria
Canary Islands
Tenerife
Canary Islands
Tenerife
Cook Islands
Atiu
Kiribati
Tarawa
Seychelles
Assumptiona
Seychelles
Denis
Seychelles
Fregate
Year
2007
2008
2006
2000
2007
2016
2015
2014
2015
2011
Control
Island group
Island
American Samoa
Tutuila
Canary Islands
Fuerteventura
French Polynesia
Tahiti
Seychelles
North
UK Overseas Territories
Ascensionb
UK Overseas Territories
Ascensionc
Year
2016
2010
2012
2016
Sept. 2009
Nov. 2009
a
Eradication of one target genus (i.e. red-whiskered bulbul) and one non-target genus (i.e. red fody Foudia
madagascariensis) was achieved
b
A control programme carried out by SSC by trapping
c
A separate one carried out by C.J. Feare by poisoning, in the same year
Methods employed on projects
Methods used in eradication and control programmes
included trapping, shooting, poisoning and mist-netting
(Table 5). The method of choice for both programme
types was live-trapping of invasive birds using live decoys
and edible baits such as bread, fruit, pet food and tinned
fish. Few programmes used shooting, with four out of
five programmes employing firearms being conducted
for population eradication purposes. Three out of the four
programmes using poisoning were controlling (as opposed
to eradicating) populations of invasive species. Only two
eradication (but no control) programmes employed mistnetting to capture birds.
DISCUSSION
It was clear when we reviewed published studies and
contacted programme managers that some programmes
304
described as eradications should have been categorised
as ongoing control programmes, according to our
classification criteria outlined in Table 1. For those that did
not carry out post-intervention monitoring, had not defined
milestones during the intervention, had not identified
invasion pathways or had not stipulated a period of
quarantine post-intervention, we suggest that they should
not be considered successful control programmes. We
also request that programme managers consider carefully
the contents of Table 1 as they plan and execute their
intervention. Many studies were published before 2000
and they were unsuccessful in the case of eradication
programmes because pathways of invasion were not closed
and/or programme managers failed to remove all targeted
birds (Tables 1 and 2). Remaining populations therefore
recovered in numbers and, as a result, they expanded
their ranges once again on islands. Such an example was
provided by Millett, et al., (2005).
Saavedra Cruz & Reynolds: Mynas and bulbuls on oceanic islands
Table 4 Total number of mynas and bulbuls of six different invasive species caught on islands (see Table 3 for further
details) as part of eradication and control programmes. Note that red-vented bulbul has been split into two subspecies
– cafer and bengalensis – for historical reasons.
Island (year)
Eradication
Assumption (2014)
Atiu (2016)
Denis (2015)
Fregate (2011)
Fuerteventura (2008)
Gran Canaria (2006)
Mallorca (2006)
Tarawa (2015)
Tenerife (2000)
Tenerife (2007)
Total birds
Control
Ascension (Sept. 2009)
Ascension (Nov. 2009)
Fuerteventura (2010)
North (2016)
St Helena (2009)
Tahiti (2012)
Tutuila (2016)
Total birds
Total birds for both
programme types
Common
myna
A. tristis
Invasive species (A.= Acridotheres; P.= Pycnonotus)
Jungle
A. hybrid
Red-whiskered Red-vented Red-vented
myna
bulbul
bulbul
bulbul P. c.
A. fuscus
P. jocosus
P. cafer cafer bengalensis
5,279
24,375
1,186
758
21
3
22
3
11
26,376
1
1
3
7
5,286
623
114
7
1,600
342
6,170
2,915
11,764
4,605
4,605
38,140
4,606
Our empirical results document the species, the
numbers of birds of each taxon and the methods employed
during the targeting of birds in eradication and control
programmes. It is clear that traps should be favoured to
‘capture’ invasive birds as we understand more about the
biology of the target species and because trap design has
markedly improved over recent years. From a practical
perspective, the construction and establishment of traps on
the ground are more preferable to applying continuously for
permits from authorities on isolated islands to import and
use firearms and poison. This said, national governmental
agencies would be well advised to facilitate the use of
complementary and effective management methods that
can be combined with trapping to allow programme staff
to progress invasive bird management on these islands and
others in the future. As an example, the experience from
Assumption suggests that combining mist-netting with
shooting can result in removal of large numbers of redwhiskered bulbuls (now eradicated) and red fodies (Foudia
madagascariensis).
As a result of considerations of both excluded (Table
2) and included studies (Tables 3 and 4) documenting
eradication and control programmes, we briefly discuss
below some of the fundamental considerations that
should be undertaken in their future planning, execution
and reporting. The outcome should be the adoption of
processes that lead to best practice in managing invasive
bird populations on oceanic islands.
3
5,286
7
9,123
2,401
11,524
7
11,524
Community engagement
Many programme managers historically argued that it
was impossible to rely on local people to instigate actions
on the ground, to remain committed to the programme
and thus to constitute the main task force addressing the
problems posed by the invasive species, as the programme
will be destined to fail because of local apathy (SSC, pers.
obs.). Nowadays, programme managers often assume
that the programme’s aims will thrive mediated by the
locals’ sense of community and shared aspirations for the
programme. Success comes through the development of
simple ‘tools’ that can be employed by the local community
to manage invasive species for the benefit of the whole
community. While people who want to become volunteers
(whether trapping or otherwise) in any invasive species
management programme have their own motivations for
doing so, the success of any such intervention lies in the
effective coordination of human power directed towards an
achievable and beneficial community goal. This sustains
commitment to the programme, especially from the
community itself.
A successful programme will not only engage with
the local community but also with wider audiences,
requiring widespread availability of well-designed and
well-delivered education campaigns, and comprehensive
media coverage. The Canberra Indian Myna Action
Group (CIMAG) provides an excellent example (albeit a
mainland one) of a society-driven movement of volunteer
305
Island invasives: scaling up to meet the challenge. Ch 2B Other taxa: Birds
Table 5 Methods employed on eradication and control programmes targeting mynas and bulbuls of six
different invasive species caught on islands (see Table 3 for further details).
Island (year)
Ocean
Trapping
Method
Shooting
Poisoning
Mist-netting
Eradication
Assumption (2014)
Atlantic
Atiu (2016)
Pacific
Denis (2015)
Indian
Fregate (2011)
Indian
Fuerteventura (2008)
Atlantic
Gran Canaria (2006)
Atlantic
Mallorca (2006)
Mediterranean
Tarawa (2015)
Pacific
Tenerife (2000)
Atlantic
Tenerife (2007)
Atlantic
8
Totals
Control
6
1
Ascension (Sept. 2009)
Atlantic
Ascension (Nov. 2009)
Atlantic
Fuerteventura (2010)
Atlantic
North (2016)
Indian
St Helena (2009)
Atlantic
Tahiti (2012)
Pacific
Tutuila (2016)
Pacific
6
1
3
Totals
community trappers that has removed >69,000 common
mynas and 8,900 common starlings (Sturnus vulgaris)
through trapping over the last 11 years (CIMAG, pers.
comm.). Their programme started in 2006 and it has
achieved unprecedented successes in controlling birds on a
continental scale, thereby demonstrating the effectiveness
of well-coordinated volunteer efforts. Invasive birds have
been managed effectively on Tahiti for the last seven years
and on Tutuila for nearly the last three years.
Programme resourcing
We make a few general points about resourcing,
based upon experiences of SSC gained from the control
programme carried out on Tahiti in 2012 (Tables 3 and
4). This programme was driven by the need for urgent
conservation action to promote the survival of the critically
endangered Tahiti monarch (Pomarea nigra) (Blanvillain,
et al., 2003; Ghestemme, 2011). It was a success because
the programme engaged fully with the local community,
and maintained high levels of motivation among local
community members by sustaining frequent and dynamic
communication between the local community and the
programme’s management team. It provided many insights
that could be transferred to other such programmes.
Contractors should provide an upfront realistic budget to
meet the costs incurred in mobilising materials and having
personnel in post at the start of actions on the ground.
Mobilisation requires transport logistics, appropriate
personnel to be available and fuel costs to be met at the
start of the programme. Materials can include components
for trap construction, mist-nets and their associated poles,
firearms and ammunition, bait stations, bait and poisons,
and storage facilities. Often equipment like traps has been
306
2
used for centuries but knowledge about the appropriate
deployment of them has been lost trans-generationally.
Money spent on re-education and re-training to address the
deployment of single traps and of coordinated networks of
traps is particularly well received, especially in locations
such as the Pacific islands (SSC, pers. obs.) where remote
communities rely upon subsistence agriculture for food
security and invasive bird species in part threaten their
very existence.
There are costs associated with employing appropriate
(i.e. informed) staff on such programmes (e.g. advertising,
interviewing) and submitting applications for permits to
relevant on- or off-island authorities for activities such as
the use of mist-nets and traps, the handling of hazardous
chemicals and the safe disposal of managed birds. Funding
is also needed to maintain surveillance efforts to ensure that
invasive birds have not returned (eradication programmes)
or exist in low numbers as a result of sustained trapping
efforts (control programmes).
Perceptions of invasive (and native) species
In many locations outside of their native ranges invasive
species may be the first birds that locals observe and
become familiar with (CIMAG, pers. comm.; SSC, pers.
obs.). Their overwhelming presence can result in native
species becoming ‘invisible’ in local communities both in
terms of reduced numbers of birds on the ground and a
loss of natural history knowledge through education and
personal experiences. This erosion of so-called ‘traditional
ecological knowledge’ (TEK) is a widespread phenomenon
(Sinclair, et al., 2010) and is not just restricted to remote
oceanic islands. Children often tend to consider invasive
species as ‘normal’ because they observe them constantly
Saavedra Cruz & Reynolds: Mynas and bulbuls on oceanic islands
throughout their formative years. In local communitybased management projects, public awareness of native
species for aesthetic, as well as ecosystem service, benefits
is crucial in gaining public support, resulting in potent
public engagement with invasive eradication and control
programmes. Local people become highly motivated
rapidly, especially if provided with effective management
‘tools’ to control invasive bird species. The challenge to
the conservation manager is to promote native species’
survival as a positive outcome of effective invasive species
management in addition to other benefits to the local
community. Whether this generates a conservation ethic
in local peoples beyond that of their livelihoods remains
aspirational but realistic, given experiences of SSC in
the last seven years of control in Tahiti and three years in
Tutuila.
Expertise networks
If invasive species are to be targeted successfully we
must develop networks of expertise that are constituted
not just by species experts (e.g. invasive species managers,
professional ornithologists, avian pest controllers), but also
by local experts who have developed detailed knowledge
of the target species on the ground after training. Networks
can thereby provide a detailed knowledge of the species’
biological traits such as flocking patterns (Sinu, 2011),
responses to novel foods (Martin & Fitzgerald, 2005),
changes in food preference in relation to their location in
their distributional range (Liebl & Martin, 2014), and trap
shyness (Camacho, et al., 2017). For example, common
mynas can be trapped for long periods of time without
developing ‘trap shyness’ (SSC, pers. obs.), but only if
trappers follow the recommended protocols.
Usually, local people have some biological knowledge
of targeted invasive species, but on rare occasions some
have detailed local knowledge about birds. All such
knowledge can be obtained from full engagement with the
local community who may have attempted eradication and
control methods albeit in an uncoordinated manner that
invariably results in unsuccessful outcomes. Knowledge
can relate to where birds roost, favoured routes between
roost and foraging sites, where they drink, their preferred
foods and even how they behave in response to presentation
of novel foods (e.g. Lermite, et al., 2017, SSC, pers. obs.).
In some cases, ethno-ornithological knowledge (Tidemann
& Gosler, 2010) could prove to be fundamental in the
successful deployment of methods on the ground but to the
best of our knowledge it has failed to inform eradication
and control programmes to date.
Creating and sustaining networks of trappers on
single islands and on chains of islands are fundamental
in targeting high numbers of birds to be removed. Full
engagement in terms of commitment and motivation
by programme managers is key to retaining network
integrity. Communication is the principal way to enlist
assistance from trap builders and volunteer trappers,
to inform the island population, to recruit local people
to the programme, to educate the community about its
benefits, and to update local people about the results of the
programme to date. It is not just the general public that
needs to be updated but, just as importantly, members of
the trapper network itself. Sharing positive results from the
ongoing programme motivates everyone and if a problem
in the network is described in sufficient detail, a solution
can be found rapidly because of shared experience and
capacity in problem solving. Of course, a sustained line of
communication also engages with stakeholders beyond the
programme’s location such as international agencies who
might be partially funding the work.
The reality of most programmes is that training of
staff takes the form of native biodiversity conservation but
that of volunteers is focussed on local habitat protection,
whether cash crop, farmland or otherwise. The practical
training to build and deploy traps should be similar for
both of the above groups, but is often viewed as being less
exigent for volunteers. However, if local people are trained
in partnership with programme staff through an established
expertise network, often trap design and deployment can
Table 6 Attributes defining best practice in planning and executing effective eradication and control
programmes of avian invasive species on oceanic islands.
Attribute
Pre- and during intervention
Eradication
Control
Local government support
Stakeholders identified and engaged with
Internal and external communication channels identified and open
Training of local and contract staff
Milestones identified
Full financial resources
Full non-financial resources
Full financial resources (including contingencies)
Full non-financial resources (including contingencies)
Refresher training of local and contract staff
All communication channels remain open
Birds absent
Monitoring for birds
Pathways of invasion closed
Ongoing management of pathways of invasion
Defined period of quarantine
Post-intervention
307
Island invasives: scaling up to meet the challenge. Ch 2B Other taxa: Birds
be improved through inputs of local knowledge (Tidemann
& Gosler, 2010). Part of such training should include
emphasising the importance of record keeping. Recording
data is key to a programme’s success but can sometimes
be problematic when carried out by local trappers without
an appreciation for its importance. The transmission of
data between trappers, programme managers and their
staff can result in the loss of data when resources such
as standardised datasheets, time, computer hardware
and software etc. are lacking. Data collection should run
smoothly with full commitment of participants on such
programmes if training has been effective and expertise
networks are maintained.
What constitutes best practice in control and eradication
programmes targeted at invasive bird populations on
oceanic islands?
To conclude, we refer the reader to Table 6 where
we summarise the main attributes of effective control
and eradication programmes. These attributes should be
considered alongside others that we have discussed in this
study. In conclusion, we have provided an account of the
most common invasive avian species that have been targets
for conservation action on oceanic islands where they
threaten native species and the livelihoods of local human
communities. Mynas and bulbuls still pose major threats
to local economies and to native biodiversity, and we must
find ways to plan and execute their eradication and control
that engage with local communities while guaranteeing
that programme outcomes are attained. Above, we have
discussed effective planning through full engagement with
and between the local community, programme managers
and team members (whether volunteers or otherwise) to
capacity build through education and training. This results
in the construction and maintenance of expertise networks
that are built on the ideas of local people, harnessing
their local knowledge about the target species and on an
appreciation of the benefits of the proposed actions to the
local community. Executing plans involves coordinated
action on the ground between programme managers,
their staff and local volunteers that arises from sustained
communication and motivation in meeting all of the
programme’s goals. Success involves far more than simply
providing financial resources to cover various elements
of a programme. If we were to propose one overarching
recommendation it would be that programmes share
information using standardised reporting protocols as
everyone strives to adopt best practice.
ACKNOWLEDGEMENTS
We thank all of the programme managers who have so
kindly answered our queries relating to finer details of their
work. Their names are withheld because of the sensitivities
surrounding such invasive species management. Félix M.
Medina and two anonymous referees provided helpful
comments that significantly improved the manuscript. We
extend a special thanks to the Editor of the conference
proceedings for his patience and encouragement as the
manuscript was prepared for submission.
REFERENCES
Blackburn, T.M., Lockwood, J.L. and Cassey, P. (2009). Avian Invasions:
The Ecology and Evolution of Exotic Birds. Oxford, UK: Oxford
University Press.
Blanvillain, C., Salducci, J.M., Tutururai, G. and Maeura,
M. (2003). ‘Impact of introduced birds on the recovery of the Tahiti
flycatcher (Pomarea nigra), a critically endangered forest bird of
Tahiti’. Biological Conservation 109: 197–205.
Burns, F.E. (2011). ‘Conservation Biology of the Endangered St Helena
Plover Charadrius sanctaehelenae’. PhD thesis. Bath, UK: University
of Bath.
Camacho, C., Canal, D. and Potti, J. (2017). ‘Lifelong effects of trapping
experience lead to age-biased sampling: Lessons from a wild bird
population’. Animal Behaviour 130: 133–139.
308
Canning, G. (2011). ‘Eradication of the invasive common myna,
Acridotheres tristis, from Fregate Island, Seychelles’. Phelsuma 19:
43–53.
Coblentz, B.E. (1990). ‘Exotic organisms: A dilemma for conservation
biology’. Conservation Biology 4: 261–265.
Colautti, R.I. and MacIsaac, H.J. (2004). ‘A neutral terminology to define
‘invasive’ species’. Diversity and Distributions10: 135–141.
Crowley, T.J. (2000). ‘Causes of climate change over the past 1000 years’.
Science 289: 270–277.
Donlan, C.J. and Wilcox, C. (2008). ‘Diversity, invasive species and
extinctions in insular ecosystems’. Journal of Applied Ecology 45:
1114–1123.
Dukes, J.S. and Mooney, H.A. (1999). ‘Does global change increase the
success of biological invaders?’ Trends in Ecology and Evolution 14:
135–139.
Dulloo, M.E., Kell, S.P. and Jones, C.G. (2002). ‘Conservation of endemic
forest species and the threat of invasive species’. International Forestry
Review 4: 277–335.
Earle, T. (1994). ‘Positioning Exchange in the Evolution of Human
Society’. In: T.G. Baugh and J.E. Ericson (eds.) Prehistoric Exchange
Systems in North America. Interdisciplinary Contributions to
Archaeology, pp. 419–437. Boston, MA, USA: Springer.
Feare, C.J. (2017). Orange Omelettes & Dusky Wanderers: Studies and
Travels in Seychelles Over Four Decades. Mahé, Seychelles: Calusa
Bay Publications.
Foley, S.F., Gronenborn, D., Andreae, M.O., Kadereit, J.W., Esper, J.,
Scholz, D., Pöschl, U., Jacob, D.E., Schöne, B.R., Schreg, R., Vött,
A., Jordan, D., Lelieveld, J., Weller, C.G., Alt, K.W., GaudzinskiWindheuser, S., Bruhn, K.-C., Tost, H., Sirocko, F. and Crutzen, P.J.
(2013). ‘The Palaeoanthropocene – the beginnings of anthropogenic
environmental change’. Anthropocene 3: 83–88.
Ghestemme, T. (2011). ‘Impact of introduced birds on Tahiti monarch’.
Pacific Invasives Initiative Newsletter 12: 4.
Hughes, B.J., Martin, G.R. and Reynolds, S.J. (2008). ‘Cats and seabirds:
Effects of feral domestic cat Felis silvestris catus eradication on the
population of sooty terns Onychoprion fuscata on Ascension Island,
South Atlantic’. Ibis 150 (Suppl. 1): 121–129.
Hughes, B.J., Martin, G.R. and Reynolds, S.J. (2017). ‘Estimating the
extent of seabird egg depredation by introduced common mynas on
Ascension Island in the South Atlantic.’ Biological Invasions 19: 843–
857.
Lermite, F., Perneaux, C. and Griffin, A.S. (2017). ‘Personality and
problem-solving in common mynas (Acridotheres tristis)’. Behavioural
Processes 134: 87–94.
Liebl, A.L. and Martin, L.B. (2014). ‘Living on the edge: Range edge
birds consume novel foods sooner than established ones’. Behavioral
Ecology 25: 1089–1096.
Lowe, S., Browne, M., Boudjelas, S. and De Poorter, M. (2000). 100 of
the World's Worst Invasive Alien Species: A Selection from the Global
Invasive Species Database. Auckland, New Zealand: The Invasive
Species Specialist Group (ISSG) of the World Conservation Union
(IUCN).
Martin II, L.B. and Fitzgerald, L. (2005). ‘A taste for novelty in invading
house sparrows, Passer domesticus’. Behavioral Ecology 16: 702–707.
Millett, J., Climo, G. and Shah, N.J. (2005). ‘Eradication of common
myna Acridotheres tristis population in the granitic Seychelles:
Successes, failures and lessons learned’. Advances in Vertebrate Pest
Management 3: 169–183.
Parkes, J. (2012). Review of Best Practice Management of Common
Mynas (Acridotheres tristis) with Case Studies of Previous Attempts at
Eradication and Control: A Working Document. Lincoln, New Zealand:
Landcare Research.
Reaser, J.K., Meyerson, L.A., Cronk, Q., de Poorter, M., Eldredge,
L.G., Green, E., Kairo, M., Latasi, P., Mack, R.N., Mauremootoo, J.,
O’Dowd, D., Orapu, W., Sastroutomo, S., Saunders, A., Shine, C.,
Thrainsson, S. and Vaiutu, L. (2007). ‘Ecological and socioeconomic
impacts of invasive alien species in island ecosystems’. Environmental
Conservation 34: 98–111.
Sinclair, R.J., Tuke, L. and Opiang, M. (2010). ‘What the Locals Know:
Comparing Traditional and Scientific Knowledge of Megapodes in
Melanesia’. In: S. Tidemann and A. Gosler (eds.) Ethno-ornithology:
Birds, Indigenous Peoples, Culture and Society, pp. 115–137. London,
UK: Earthscan.
Sinu, P.A. (2011). ‘Avian pest control in tea plantations of sub-Himalayan
plains of Northeast India: Mixed-species foraging flock matters’.
Biological Control 58: 362–366.
Tidemann, S. and Gosler, A. (2010). Ethno-ornithology: Birds, Indigenous
Peoples, Culture and Society. London, UK: Earthscan.
Vitousek, P.M., D’Antonio, C.M., Loope, L.L., Rejmánek, M. and
Westbrooks, R. (1997). ‘Introduced species: A significant component
of human-caused global change’. New Zealand Journal of Ecology 21:
1–16.
Westphal, M.I., Browne, M., MacKinnon, K. and Noble, I. (2008). ‘The
link between international trade and the global distribution of invasive
alien species’. Biological Invasions 10: 391–398.
Williamson, M. and Fitter, A. (1996). ‘The varying success of invaders’.
Ecology 77: 1661–1666.
Zeder, M.A. (2008). ‘Domestication and early agriculture in the
Mediterranean Basin: Origins, diffusion, and impact’. Proceedings of
the National Academy of Sciences USA 105: 11597–11604.
M. Thibault, E. Vidal, M.A. Potter, F. Masse, A. Pujapujane, B. Fogliani, G. Lannuzel, H. Jourdan, N. Robert, L. Demaret, N. Barré and F. Brescia
Thibault, M.; E. Vidal, M.A. Potter, F. Masse, A. Pujapujane, B. Fogliani, G. Lannuzel, H. Jourdan, N. Robert, L.
Demaret, N. Barré and F. Brescia. Invasion by the red-vented bulbul: an overview of recent studies in New Caledonia
Invasion by the red-vented bulbul: an overview of recent
studies in New Caledonia
M. Thibault1,3, E. Vidal2, M.A. Potter3, F. Masse1,4, A. Pujapujane1, B. Fogliani1, G. Lannuzel1,
H. Jourdan2, N. Robert5, L. Demaret5, N. Barré1 and F. Brescia1
Institut Agronomique Néo-Calédonien (IAC), Equipe ARBOREAL (AgricultuRE BiOdiversité Et vAlorisation) BP 73,
98890 Païta, Nouvelle-Calédonie. <thibault@iac.nc>. 2Institut Méditerranéen de Biodiversité et d’Ecologie marine et
continentale (IMBE), Aix Marseille Université, CNRS, IRD, Avignon Université Centre IRD Nouméa - BP A5, 98848
Nouméa Cedex, Nouvelle-Calédonie. 3Wildlife & Ecology Group, School of Agriculture and Environment, Massey
University, Palmerston North 4442, New Zealand. 4 Faculté des arts et des sciences, Université de Montréal, P.O. Box
6128, Succursale Centre-ville, Montréal, H3C 3J7, Québec, Canada. 5Institut Agronomique Néo-Calédonien (IAC),
Equipe SOLVEG (SOL et VEGétation), BP 711, 98810, Mont Dore, Nouvelle-Calédonie.
1
Abstract New Caledonia is a tropical archipelago of the South Pacific Ocean, and is one of the 36 world biodiversity
hotspots. However, its unique biodiversity is increasingly threatened by habitat fragmentation and introductions of invasive
alien species. Among these invaders, the red-vented bulbul (Pycnonotus cafer) is currently expanding towards the north
of the main island. This passerine features in the IUCN-ISSG list of the 100 worst invasive species of the world because
of impacts caused by its diet. Thirty-five years after its introduction, we present an overview of data from recent studies
conducted in New Caledonia that describe the local status of the red-vented bulbul, its range expansion, and potential
impacts on both the local biodiversity and agriculture. Biannual monitoring of the distribution coupled with surveillance
at the edges of native forests highlighted a tight association of the bulbul with man-modified habitats. Using a distance
sampling method, we estimated that bulbul densities within the distribution core varied from a peak of 200 individuals/
km2 in the main city of Nouméa, where the species has been introduced, to 30 individuals/km2 in rural habitats located 50
km away from Nouméa. We conducted a diet analysis on 40 bulbul corpses and found that 82% and 55% of individuals
had consumed plant and animal items, respectively. We identified plant and insect species that may be of concern in the
contexts of seed dispersal and predation by the red-vented bulbul. Finally, a food colour selection experiment and an
open field test showed that the red-vented bulbul had a significant preference for red and sweet fruits. We estimated the
economic loss caused by bulbuls to a tomato grower and discuss the result with respect to the development of an adapted
management strategy, to prevent further impacts of the red-vented bulbul on the biodiversity and agriculture in the tropical
island hotspot of New Caledonia.
Keywords: density, diet, distribution, impacts, invasive bird, Pycnonotus cafer
INTRODUCTION
New Caledonia is a tropical archipelago located to
the east of Australia, in the SouthPacific Ocean. The
archipelago has been classified as one of the world’s 36
biodiversity hotspots because of its high levels of endemism
in such a small territory (Williams, et al., 2011). Among
notable features of local biodiversity in New Caledonia,
Myers, et al. (2000) highlighted five endemic families and
112 endemic genera of plants, and one endemic family
and three endemic genera of birds. However, a significant
proportion of this biological richness is increasingly
threatened by human activities and global changes, as is the
case for most of the world’s biodiversity hotspots (Bellard,
et al., 2014). Among factors that foster these changes,
habitat fragmentation and climate change are widely
recognized (Garcia, et al., 2014; Haddad, et al., 2015),
although the best response from scientists and managers to
species’ introductions is still a matter of debate (Russell &
Blackburn, 2017; Davis & Chew, 2017).
The effects of invasive species have been widely
documented (Early, et al., 2016). Impacts are accentuated
in island ecosystems (Russell, et al., 2017), often because
of the naivety of insular species (Gerard, et al., 2016)
and environmental, ecological and evolutionary factors
associated with geographic isolation (Cabral, et al., 2017).
Humans play a key role in the transportation of plant and
animal species worldwide (Ricciardi, et al., 2017). Trade
in animals (Cardador, et al., 2017; Su, et al., 2016) and the
release or escape of cage birds are frequently identified as
the main mechanisms for alien bird introductions and the
dispersal of wild birds outside of their native ranges (Dyer,
et al., 2017).
Tropical bird species, particularly those from Southeast Asia, occupy an important place in global bird trade
(Nijman, 2010), with bulbuls, starlings, mynas and robins
figuring amongst the most traded species from this region
(Harris, et al., 2015). As a result, two out of three species
considered in the IUCN-ISSG list of 100 worst invasive
species are native to Southern Asia: the red-vented bulbul
(Pycnonotus cafer) and the common myna (Acridotheres
tristis) (Lowe, et al., 2000). These two species historically
were widely transported from India to Pacific Islands
(Watling, 1978) and both are now established in New
Caledonia (Brochier, et al., 2010). Our global review on
the impact and management of alien red-vented bulbuls
identified 37 islands in the alien distribution of this species
(Thibault, et al., 2018a). This study also highlighted the
lack of quantitative data and evidence-based assessments
of the impacts associated with this invasive species. The
red-vented bulbul was introduced into New Caledonia in
1983 (Gill, et al., 1995) and its local distribution range
is currently expanding from Nouméa toward the north
and south of the main island. For 25 years following its
introduction into Nouméa, no studies were conducted
to investigate the ecology, distribution or impacts of
the species at a local scale. This lack of information has
precluded any detailed assessment of the threats posed
by the establishment of the red-vented bulbul in New
Caledonia. Consequently, it has not thus far been possible
to implement an evidence-based management strategy.
The goal of this paper is to present an overview of
data from recent studies conducted in New Caledonia
to describe the local status of the red-vented bulbul, its
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 309–316. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
309
Island invasives: scaling up to meet the challenge. Ch 2B Other taxa: Birds
range expansion, and potential impacts on both the local
biodiversity and agriculture. We firstly report the local
distribution range of the species, the rate and nature of its
range expansion, and its habitat selection and densities
in different habitats. We then use diet analysis to explore
potential negative effects of the red-vented bulbul on
natural and agricultural systems. We present original
data on an ongoing invasion process in a tropical island
biodiversity hotspot and highlight priority areas for local
red-vented bulbul research and risk management.
METHODS
Red-vented bulbul range expansion
Red-vented bulbul dispersal was monitored over time
using static 10-min point counts combined with 2-min
playback of recorded calls to increase detection probability
(Ralph, et al., 1995). Points were sampled within the
four hours following sunrise, between November and
December in 2008, 2012, 2014 and 2016. Each point was
geo-referenced, and the observers accounted for seen
and heard individuals. In 2008, 136 points were sampled
that covered Nouméa and suburbs as well as borders of
the two main roads going to the north and south. Random
points were also located in major urban areas along these
roads to search for potential pioneering individuals. The
method was replicated in 2012, 2014 and 2016, covering
203, 96 and 99 points respectively. Data were compiled
and plotted in Qgis software version 2.18.1 (Quantum GIS
Development Team, 2016).
In April 2016, we selected six additional sites across
native and man-modified habitats to explore the future
establishment of the red-vented bulbul in forests. We chose
three sites across urban and dry forest habitats, and three
across urban and wet forest habitats. These sites were
located close to the core of the distribution range, where
red-vented bulbul densities were highest. We placed 10
points spaced at least 250 m apart at each site (five points
per habitat) and counted red-vented bulbul individuals seen
and heard. The method used was the same as for distribution
monitoring. Data were compiled in Qgis software version
2.18.1 (Quantum GIS Development Team, 2016) and
plotted in R software version 3.4.0 (R Core Team, 2017).
Red-vented bulbul densities
Red-vented bulbul density was measured using a
distance sampling method (Thomas, et al., 2010) in four
sites located within the core of the red-vented bulbul
distribution range. This method relies on three key
assumptions: i) individuals at zero metres distance are
detected with certainty, ii) individuals are detected once
and at their initial location, and iii) distance measurements
are exact.
Sites were selected in man-modified habitats, along a
distance gradient from Nouméa to Tomo, a village located
about 50 km farther north. Three transects of 1 km were
established at each site and sampled between October
and December 2015. A pair of observers walked along
each transect for 30 minutes and counted the number of
individuals seen on both sides. The distance of observed
individuals from each transect was recorded with a laser
telemeter. Transects were sampled three times between
0500 and 0900 hours and data from the three sessions were
used independently and pooled to prevent a potential bias
due to time of day.
Data were analysed with the “Distance” package (Miller,
2016) using R software version 3.4.0 (R Core Team, 2017).
This method considers potential missed observations in
the estimated bird densities thanks to the calculation of
310
a detection probability curve. We first estimated the bird
density at each site using data from the three sessions
separately. Then, we estimated densities at each site using
data of the three sessions together, considering the nine
transects at each site as independent. Finally, we chose to
present the estimates from the pooled dataset as it provided
a smoothed estimation of densities regarding the influence
of time of day on bird detection.
Red-vented bulbul diet analysis
Gut content analysis was conducted on 40 dead redvented bulbuls provided by local hunters. There is no
morphological dimorphism between male and female redvented bulbul, so we were only able to determine the sex
of sexually mature individuals, using anatomical analysis.
Gastrointestinal tracts were excised and the contents
removed and washed with tap water through a 0.2 mm
sieve. The retained contents were placed in a Petri dish
filled with 70% alcohol and examined under a dissecting
microscope at 10× magnification (Olympus SZ61). Items
were photographed (Toupcam UCMOS camera and
Toupview software) for subsequent identification (Lopes,
et al., 2005).
Fruit colour selection
According to the literature, damage to cultivated plants
is the most frequently reported impact of the red-vented
bulbul in its alien range (Thibault, et al., 2018a). This is
also the impact category most often reported locally both
by professionals (Caplong & Barjon, 2010) and nonprofessionals. We tackled this issue through two distinct
experiments, a colour preference test and an open-field test.
We conducted an experiment on fruit colour selection
to test whether the red-vented bulbul was attracted by
some fruit colours more than others. We trapped eight
adult individuals, maintained them in an aviary for at least
a month, and in individual cages for three days. We created
false-coloured fruits of four distinct colours, following
the method presented in Duan & Quan, (2013). Artificial
fruits were made of banana, chicken grain and water, and
three quarters of the fruits were coloured with red, green
and yellow food colouring. Ten fruits of each colour were
placed in four different petri dishes in cages with bulbuls
held individually and observed for 25 minutes from a hidden
position. Each bird was tested once during five consecutive
days, following either two hours or six hours of fasting.
For each repetition, the colour of the first fruit eaten as well
as the total number of fruit eaten per colour were recorded.
ANOVA tests were conducted in R software version 3.4.0
(R Core Team, 2017) with hypothesis H0 being that each
fruit colour had the same probability of being eaten first.
Damage to crops
In 2016, we conducted an open field test to explore the
range of damage caused by red-vented bulbuls to tomato
crops. We planted eight tomato plants inside each of 20
square plots spaced by one metre, and randomly covered
half of the plots with bird netting during the flowering
stage. During the fruiting period in August and September,
each plot was monitored twice a week. Ripe and damaged
fruit were harvested and separated in three categories; i)
marketable; ii) pecked fruits; and iii) other damage. For
each category, the colour, size, and sugar levels (in Brix
degrees; Bates, 1942) of fruit were recorded. Tomatoes
that were pecked by the birds were easily recognizable
by beak marks, and the mark’s size together with direct
observations were used to determine the fruits that were
damaged by red-vented bulbuls. The relative economic
loss in marketable tomatoes due to bulbul damage was then
Thibault, et al.: Red-vented bulbul in New Caledonia
calculated as the total weight of pecked tomatoes divided
by the total weight of tomatoes harvested in ‘unprotected
plots’. This percentage was then extrapolated to the national
production recorded during the month of our experiment.
Data were analysed with the “nlme” package (Pinheiro, et
al., 2017) using in R software version 3.4.0 (R Core Team,
2017).
RESULTS
Red-vented bulbul range expansion
The 2008–2016 red-vented bulbul biannual distribution
map (Fig. 1) shows a continuous increase in the distribution
range occupied by the red-vented bulbul in New Caledonia.
Coloured polygons contain all points where red-vented
bulbul individuals were observed in 2008, 2012, 2014 and
2016. Conversely, green dots represent all points where redvented bulbul were not detected either during point-counts
or during playback calls. The green triangles and diamonds
represent, respectively, absence points located in natural
dry forest patches within the city of Nouméa, and in humid
forest, which represents the northern border of the capital
and its suburbs. This absence data suggests that the species
is not yet spreading into natural forest. Indeed, over the
60 point counts conducted at frontiers between urban and
forest habitats, we detected red-vented bulbul individuals
at 16 points in urban habitats and one point in dry forest
habitat. We also received testimonies from local people
about red-vented bulbul sightings. These testimonies were
rarely confirmed by further observations but sometimes led
to new detections. Figure 1 shows a continuous distribution
of the red-vented bulbuls with range expansion particularly
along main roads. It also presents absence data from
another study (Thibault, et al., 2018b) which are consistent
with this hypothesis. The two road axes from Nouméa to
La Foa (100 km north) and Yaté (95 km south) appeared to
be the main dispersal pathways. In 2012, 25 years after its
introduction in the city of Nouméa, the red-vented bulbul
had reached Tontouta, 42 kilometres north. From 2012 to
2016 the species travelled 35 kilometres north (Fig. 2).
Nowadays the red-vented bulbul occupies at least 1,350
km2 (8% of the New Caledonia territory), mostly restricted
to the west coast of the southern province.
Fig. 2 Rate of red-vented bulbul dispersal toward the North
and South of Nouméa.
Red-vented bulbul densities
Most birds were both heard and seen during our
sampling sessions in inhabited areas. We fitted our data to a
half-normal distribution (Thomas, et al., 2010) to calculate
the detection function (Fig. 3). Density estimates from the
three sessions and from the pooled data set are presented in
Fig. 4. Red-vented bulbul estimated density was six times
higher in the city of Nouméa (d: 204 ± 23 individuals/km2)
than in the village of Tomo which is located 50 kilometres
north (d: 31 ± 11 individuals/km2, Table 1). Estimates from
the two suburban areas, Robinson and Paita, were almost
identical (d: 160 ± 32 individuals/km2 and d: 131 ± 18
individuals/km2, respectively). The density estimates are
corrected by a detection function curve which represents
the probability of an observer detecting a red-vented
bulbul depending on its distance from the transect. In the
four urban habitats we sampled, the average probability of
detecting a red-vented bulbul was 50% when the bird was
approximately 25 metres from the observer.
Red-vented bulbul diet analysis
We extracted and analysed the gut contents of 40 redvented bulbuls. Results of the diet study are presented in
Fig. 1 Map of the expanding distribution of the red-vented
bulbul between 2008 and 2016 according to the biannual
monitoring. Stars represent observations from local
people. Green dots represent point absence data (point
counts) from the distribution monitoring. Green triangles
and diamonds represent absence data (point counts
along transects) in natural forests surrounding the
distribution core. Grey dots show absence data (point
counts) from another study (Thibault, et al., 2018b).
Fig. 3 Probability of detecting a red-vented bulbul individual
as a function of distance from the transect in inhabited
areas.
311
Island invasives: scaling up to meet the challenge. Ch 2B Other taxa: Birds
Table 1 Sampling statistics and density estimates at four urban sites within the current distribution
range of the red-vented bulbuls according to distance from the introduction point. (n) total
number of individuals over the three sessions.
Site
Distance (km) Habitat
Nouméa
Paita
Robinson
Tomo
0
10
25
50
urban
suburban
suburban
rural
Table 2 Occurrences (n) and frequency (%) of
food items identified in the gut content of 40
bulbul individuals.
Fruit parts
Whole fruit
Seeds
Fruit skin
Fruit flesh
Plant families
Myrtaceae
Passifloraceae
Sapindaceae
Solanaceae
Insects
Coleoptera
Diptera
Hemiptera
Hymenoptera
Odonata
n
F (n=40)
16
22
7
17
33
20
1
2
4
22
8
1
13
3
1
40
55
17.5
42.5
82.5
50
2.5
5
10
55
20
2.5
32.5
7.5
2.5
Area
(m2)
787.3
816.3
492.8
993.3
n
117
66
65
15
Density estimate Standard
(ind/km2)
error
204
± 23
160
± 32
131
± 18
31
± 11
Table 2. Mean weight of mature individuals was 38.3 ±
4.9 g for females and 44.1 ± 5.6 g for males. We found
plant remains in the gut content of 33 individuals (82.5%)
and animal items in 22 (55%). Among plant items, seeds
(55%) and fruit flesh (42.5%) were the most frequent. The
most frequent plant family was Myrtaceae (20 individuals),
and the most consumed insect orders were Hemiptera (13
individuals) and Coleoptera (8 individuals). Identification
of the remains highlighted the consumption of one endemic
plant species (Myrtastrum rufopunctatum), two cultivated
species (Syzygium cumini and Lichi chinensis) and two
invasive alien species (Passiflora foetida and Solanum
torvum). Exoskeleton parts from cicada individuals were
frequent in this sample (F=32.5%). No vertebrate remains
were found during this analysis.
Fruit colour selection
Colour selection tests were replicated 102 times. The
first pecked fruit was red in 77% of samples, followed
by green (10% of samples). The average number of
consumed fruits per colour is presented in Fig 5. Red
fruits were the most often consumed (5 ± 0.3), and yellow
ones were consumed five times less often (0.9 ± 0.16).
Colour explained the consumption of fruits significantly
(ANOVA: F: 8.3; p<0.001). In our analysis, fasting period
did not contribute to explain the choice of coloured fruits
(ANOVA: F: 2.7; p=0.1).
Damage to crops
Fig. 4 Densities of red-vented bulbuls at each site
calculated from the three sampling sessions, and from
the pooled dataset.
312
On our 20 plots, we produced a total of 2,310 tomatoes
(345.5 kg). Unfortunately, three plots with nets were
damaged by feral dogs just before the beginning of the
fruiting season, and were thus considered to be unprotected.
Red-vented bulbuls were the only birds that fed on tomato
fruit during the experiment. Results are presented in Fig 6.
On average, production per plot was homogenous in netprotected pots (18.5 ± 2.1 kg) compared to ‘unprotected’
ones (16.6 ± 2.3 kg). Losses due to bird damage were
recorded almost exclusively in ‘unprotected’ plots and
corresponded to 2.95 ± 0.24 kg per plot (17.5%), as only
three tomatoes were pecked at the edge of protected plots
(0.5% in weight). These losses were similar to those caused
by other pests: 2.63 ± 0.3 kg in unprotected plots, and 3.9
± 0.3 kg in protected plots. Pecked fruits were mainly
red (ANOVA F: 7.6; p=0.009), between 50 and 70 mm
in size and with high sugar levels (5°Bx, ANOVA: 5.95;
p=0.016). Considering that 34 tons of tomato were sold at
3.18 USD/kg in September 2016 in New Caledonia, the
17.5% loss we recorded because of bird damage would
have corresponded to an economic loss of approximately
$18,355 USD for September 2016 alone.
Thibault, et al.: Red-vented bulbul in New Caledonia
south of the main island of New Caledonia is dominated
by ultramafic soils and the dominant vegetation type is the
“maquis minier”, a shrubland characterised by xerophytic
plants (Jaffré, et al., 2003; Jaffré, et al., 2004) which may
be less attractive in terms of food source for red-vented
bulbuls. Considering the dispersal speed, we know that
the red-vented bulbul’s range expanded 40 kilometres
toward the north of its introduction point in 25 years. Its
range expansion then increased more quickly, extending a
further 35 kilometres in just four years. This is consistent
with findings of Aagaard & Lockwood (2014) on growth
lag in alien bird populations and suggests that this range
expansion could continue to accelerate. Our observation of
a lagged expansion in the red-vented bulbul could thus be
explained both by a demographic time-lag, inter-specific
relationships, or by the carrying capacity of the different
habitats.
Fig. 5 Result of 102 colour preference tests with redvented bulbuls. The y-axis represents the average
number of fruits consumed by tested individuals during
one session.
DISCUSSION
Dispersal along urban corridors
The red-vented bulbul has continuously increased its
distribution range in New Caledonia since its introduction
25 years ago. The distribution map suggests that roads
and urban habitats are the main dispersal pathways for the
species. The dispersal rates we estimated were different
depending on the direction. One reason for this may be
differences in habitat to the north and south of Nouméa. The
Study of red-vented bulbul occurrence at the
frontiers between urban and forest habitats confirmed the
association of the species with man-modified habitats. Our
results suggest that the red-vented bulbul is not spreading
from invaded urban areas into either dry forest patches or
into native rainforests. This is consistent with previous
observations of Watling (1979) in Fiji. However, in Tahiti
(French Polynesia) the red-vented bulbul is able to colonize
native tropical forests with major impacts on native
avifauna (Blanvillain, et al., 2003). Further monitoring of
the distribution is thus crucial to anticipate potential shifts
in the habitat occupancy and resulting threats on forest
bird communities. A specific effort could be dedicated
by managers to prevent future establishment of pioneer
individuals out of the current range, toward the north, the
Loyalty Islands or specific areas of high conservation/
agricultural value. Quick detection coupled with control
actions at the edges of the red-vented bulbul range will
reduce the colonisation speed and prevent future negative
effects.
Fig. 6 Result of the open field test conducted with tomato plants. (a) Represents the global production weights for “netprotected” and “unprotected” treatments, (b) and (c) represent the mean weight of fruit damaged per plot by birds and
other pests. (d), (e) and (f) represent the average number of damaged fruit depending on fruit colour, size, and sugar
content, respectively.
313
Island invasives: scaling up to meet the challenge. Ch 2B Other taxa: Birds
Density gradient
Dispersal of bird species is partly related to population
densities (Matthysen, 2005), so the anticipation of future
dispersal events may be facilitated by the knowledge of bird
density in specific locations. Our density estimates showed
a density gradient in the red-vented bulbul, depending
on the degree of urbanisation and the distance from the
introduction point. This has been frequently observed in
alien bird populations (Chace & Walsh, 2006). The density
level we estimated in the rural village of Tomo was similar
to those reported by Radhakrishanan & Asokan (2015) in
two villages of the Cauvery delta region in Southern India,
to which the red-vented bulbul is native. However, our
estimates for the centre of Nouméa and suburbs are similar
to those found for common bird species in European/
American urban centres (Clergeau, et al., 1998). High
bird densities in urban habitats are often associated with
low bird-community species richness (Matthysen, 2005).
Regarding its density, the red-vented bulbul is already a
predominant species in Nouméa. Monitoring the change in
red-vented bulbul densities over time will contribute to a
better understanding of the species’ dynamics. It will also
allow the estimation of the density-impact relationship in
further management programmes (Yokomizo, et al., 2009),
as management of invasive alien species populations often
relies on abundance/density reductions (Genovesi, 2005;
Simberloff, et al., 2005). For example, control operations
could be feasible at low densities, whereas mitigation
of specific impacts could be more cost-efficient at highdensity levels.
Predation and frugivory
Results of the diet analysis were consistent with
previous observations elsewhere in both the alien and
native range of the species (Watling, 1978; Bhatt & Kumar,
2001; Brooks, 2013 Bates, et al., 2014). The diet comprised
mostly fruits and a significant part of animal remains. We
observed several red-vented bulbul individuals feeding on
house geckos (Hemidactylus frenatus) and skinks in the
field, but we did not find any reptile or gastropod remains
in the gut contents we analysed. Such food items have
been reported in the red-vented bulbul native range (Bhatt
& Kumar, 2001). Much of the gut contents we analysed
(n=13, F=32.5%) contained remains from cicadas.
Considering the periodic lifecycle of these insects (May,
1974), this observation suggests that red-vented bulbuls
can adapt their diet to this temporary resource. Levels of
endemism are high in New-Caledonia, with approximately
92% of reptiles and nearly 100% of cicadas (Smith, et al.,
2007; Grandcolas, et al., 2008; Delorme, et al., 2016) being
endemic. Predation by alien species such as the red-vented
bulbul could thus represent an additional threat for these
species of high conservation value.
Seeds and whole fruits were found in 50% of
individuals. This observation emphasizes the redvented bulbul’s capacity to participate in seed dispersal,
particularly in association with invasive alien plant species
like Miconia calvescens or Lantana camara (Meyer, 1996;
Spotswood, et al., 2012; Spotswood, et al., 2013). In our
diet study, we identified several candidates for red-vented
bulbul-mediated dispersal. Most of them were invasive
or cultivated species, but we also identified one endemic
(Myrtastrum rufopunctatum) that is used for mining-site
restoration (Lemay, et al., 2009). Consumption of native
species by the red-vented bulbul could result in a service,
by improving the dispersal capacity of some species
(Kawakami, et al., 2009). However, it can also lead to
competitive interactions with native avifauna (Sherman
& Fall, 2010; Thibault, et al., 2018b) which can turn
into a conservation issue (Blanvillain, et al., 2003). New
314
Caledonia is considered a biodiversity hotspot (Myers, et
al., 2000) thanks to its plant diversity, with 3060 species
of flowering plants recorded (78% endemic; Munzinger,
et al., 2016). Exploring variations in the red-vented bulbul
diet over different habitat, seasons and maturity stages will
contribute to better prediction of the dispersal of both alien
and native plants as well as potential negative interactions
with endemic species. At a wider scale, such quantitative
and qualitative data will contribute to the assessment of
impacts caused by red-vented bulbuls (Thibault, et al.,
2018a).
Colour selection and damages on crops
Diet and preference for specific resources plays a key
role in impacts caused by vertebrate pest species (Herrero,
et al., 2006; Gebhardt, et al., 2011). Sometimes, these
preferences can be strong enough to aid bait selection for
both hunters and environment managers. In our experiment,
the red-vented bulbul preferred red, consistent with colour
preference in the red-whiskered bulbul (Pycnonotus
jocosus) (Duan & Quan, 2013). In a French Polynesian
study, authors concluded that preference may sometimes
be stronger than abundance in fruit selection by birds,
including the red-vented bulbul (Spotswood, et al., 2013).
Such preference for specific fruits implies that redvented bulbuls are likely to disperse or damage the fruit of
some species more than others, and that predictions can be
made about species that are likely to be most vulnerable.
Observations made during our open field experiment were
consistent with this hypothesis, with red tomatoes being
damaged more than orange or yellow ones. In unprotected
plots, damage caused by birds was equivalent to that of
all the parasites and corresponded to 17.5% loss in weight
of marketable fruit. This corresponds to the average losses
presented in Oerke (2006) in their global estimation of
economic losses due to animal pests over 11 production
types including tomato, between 2001 and 2003. In this
study, recorded losses attributed to animal species and other
pathogens on unprotected crops were of 18% and 15%,
respectively. Oerke suggested that pest control operations
allowed a 39% reduction in losses due to animal pests.
Here we showed that protecting tomato plants with nets
efficiently protected 99% of fruits, reducing by 97% the
loss in weight of marketable fruit. This early assessment of
colour selection and damage on production suggests that
red and sweet fruit/flower crops could be more sensitive
to red-vented bulbul damages. Such information is already
used in the development of trapping systems dedicated to
this species. Indeed, fruit and fresh vegetables represented
5115 and 6292 tons, respectively, of production in New
Caledonia in 2012, corresponding to 25% and 30% of the
total plant production that year (ISEE, 2012). The redvented bulbul is currently restricted to suburban areas
in a limited range, but up to 35% loss has already been
recorded on fruit production there (Caplong & Barjon,
2010). Future establishment of the species in cultivated
areas of the main island could thus represent an additional
risk to crop productivity.
CONCLUSION
The global distribution and population trends of redvented bulbul have been poorly reported, relative to many
other tropical invasive birds. The potential overlap in the
impacts associated with tropical passerine species from
south Asia, suggested by Kumschick et al. (2015), has not
been explored either. Authors have claimed that introduced
populations of red-vented bulbuls were harmless (Watling,
1979), while in other locations their role in noxious seed
dispersal (Meyer, 1996), competition with native birds
Thibault, et al.: Red-vented bulbul in New Caledonia
(Blanvillain, et al., 2003, Thibault, et al., 2018b) and
damage to crops (Walker, 2008) was suggested. New
Caledonia must deal with the current dispersal of this
species on its territory with only a few quantitative data
available from the literature (Thibault, et al., 2018a).
However, the establishment, on-going dispersal, and
impacts of the red-vented bulbul deserve attention from
conservation biologists, environment managers and local
people. Perceptions of this invasive species differ across
groups of people (Fischer, et al., 2014), but a coordinated
joint effort is required to improve our knowledge of
invasion mechanisms for the red-vented bulbul in the
New Caledonia archipelago. New Caledonia recently
produced a list of priority invasive species for management
actions, and the studies we presented here contributed to
the consideration of the red-vented bulbul among the six
species on this list.
ACKNOWLEDGEMENTS
We thank Institut Agronomique Néo-Calédonien
(IAC) staff members for their support in the field and their
assistance in the determination of food remains. Thanks
to M. Dubreuil, Y. Ititiaty, L. Bordez, V. Hecquet and H.
Vandrot for the determination of plant remains. We thank L.
Demaret for his contribution to the open field experiment.
Thanks to N. Heurard-Cueato and P. Ajahpunya for
their help in the field for the distribution monitoring and
density estimations. We also thank the local hunters for
their participation in the collection of red-vented bulbul
cadavers. All the financial support was provided by IAC.
REFERENCES
Aagaard, K. and Lockwood, J. (2014). ‘Exotic birds show lags in
population growth’. Diversity and distributions 20(5): 547–554.
Bates, F.J. (1942). Polarimetry, Saccharimetry and the Sugars.
Washington, DC: National Bureau of Standards of the US Government.
Bates, J.H., Spotswood, E.N. and Russell, J.C. (2014). ‘Foraging
behaviour and habitat partitioning in sympatric invasive birds in French
Polynesia’. Notornis 61: 35–42.
Bellard, C., Leclerc, C., Leroy, B., Bakkenes, M., Veloz, S., Thuiller, W.
and Courchamp, F. (2014). ‘Vulnerability of biodiversity hotspots to
global change’. Global Ecology and Biogeography 23(12): 1376–1386.
Bhatt, D. and Kumar, A. (2001). ‘Foraging ecology of red-vented bulbul
Pycnonotus cafer in Haridwar, India’. Forktail 17: 109–110.
Blanvillain, C., Salducci, J.M., Tutururai, G. and Maeura, M. (2003).
‘Impact of introduced birds on the recovery of the Tahiti flycatcher
(Pomareanigra), a critically endangered forest bird of Tahiti’. Biological
Conservation 109(2): 197–205.
Brochier, B., Vangeluwe, D. and Van den Berg, T. (2010). ‘Alien invasive
birds’. Revue scientifique et technique 29(2): 217–226.
Brooks, D.M. (2013). ‘Ecology, behavior, and reproduction of an
introduced population of red-vented bulbuls (Pycnonotus cafer) in
Houston, Texas’. Wilson Journal of Ornithology 125: 800–808.
Cabral, J.S., Wiegand, K. and Kreft, H. (2017). ‘Interactions between
ecological, evolutionary, and environmental processes unveil complex
dynamics of island biodiversity’. bioRχiv. doi :10.1101/099978.
Davis, M.A. and Chew, M.K. (2017). ‘‘The denialists are coming!’ Well,
not exactly: A response to Russell and Blackburn’. Trends in Ecology
and Evolution 32(4): 229–230.
Delorme, Q., Mille, C. and Jourdan, H., (2016). ‘A review of the genus
Kanakia Distant, 1892 (Insecta: Hemiptera, Cicadoidea, Cicadidae)
from New Caledonia’. Zootaxa 4092: 301–338. doi:10.11646/
zootaxa.4092.3.1.
Duan, Q. and Quan, R.C. (2013). ‘The effect of color on fruit selection in
six tropical Asian birds’. The Condor 115(3): 623–629.
Dyer, E.E., Redding, D.W. and Blackburn, T.M. (2017). ‘The
global avian invasions atlas, a database of alien bird distributions
worldwide’. Scientific Data 4: doi:10.1038/sdata.2017.41.
Early, R., Bradley, B.A., Dukes, J.S., Lawler, J.J., Olden, J.D., Blumenthal,
D. M., Gonzales, P., Grosholz, E.D., Ibañez, I., Miller, L.P., Sorte, C.J.B.,
and Tatem, A.J. (2016). ‘Global threats from invasive alien species
in the twenty-first century and national response capacities’. Nature
Communications 7: doi:10.1038/ncomms12485.
Fischer, A., Selge, S., van der Wal, R. and Larson, B.M. (2014). ‘The
public and professionals reason similarly about the management of nonnative invasive species: a quantitative investigation of the relationship
between beliefs and attitudes’. PLOS one 9(8): e105495.
Garcia, R.A., Cabeza, M., Rahbek, C. and Araújo, M.B. (2014).
‘Multiple dimensions of climate change and their implications for
biodiversity’. Science 344(6183): 1247579.
Gerard, A., Jourdan, H., Millon, A. and Vidal, E. (2016). ‘Knocking
on heaven’s door: Are novel invaders necessarily facing naïve native
species on islands?’. PLOS one 11(3): e0151545.
Gebhardt, K., Anderson, A.M., Kirkpatrick, K.N. and Shwiff, S.A. (2011).
‘A review and synthesis of bird and rodent damage estimates to select
California crops’. Crop Protection 30(9): 1109–1116.
Genovesi, P. (2005). ‘Eradications of invasive alien species in Europe: a
review’. Biological Invasions 7: 127–133.
Gill, B.J., Hunt, G.R. and Sirgouant, S. (1995). ‘Red-vented bulbuls
(Pycnonotuscafer) in New Caledonia’. Notornis 42: 214–215.
Grandcolas, P., Murienne, J., Robillard, T., Desutter-Grandcolas, L.,
Jourdan, H., Guilbert, E. and Deharveng, L. (2008). ‘New Caledonia:
a very old Darwinian island?’. Philosophical Transactions of the Royal
Society of London B: Biological Sciences 363(1508): 3309–3317.
Haddad, N.M., Brudvig, L.A., Clobert, J., Davies, K.F., Gonzalez, A.,
Holt, R.D., Lovejoy, T.E., Sexton, J.O., Austin, M.P., Collins, C.D.,
Cook, W.M., Damschen, E.I., Ewers, R.M., Foster, B.L., Jenkins, C.N.,
King, A.J., Laurance, W.F., Levey, D.J., Margules, C.R., Melbourne,
B.A., Nicholls, A.O., Orrock, J.L., Song, D. and Townshend, J.R.
(2015). ‘Habitat fragmentation and its lasting impact on Earth’s
ecosystems’. Science Advances 1(2): e1500052.
Harris, J.B.C., Green, J.M., Prawiradilaga, D.M., Giam, X., Hikmatullah,
D., Putra, C.A. and Wilcove, D.S. (2015). ‘Using market data and
expert opinion to identify overexploited species in the wild bird
trade’. Biological Conservation 187: 51–60.
Herrero, J., García-Serrano, A., Couto, S., Ortuño, V.M. and GarcíaGonzález, R. (2006). ‘Diet of wild boar Sus scrofa L. and crop
damage in an intensive agroecosystem’. European Journal of Wildlife
Research 52(4): 245–250.
ISEE (2012). ‘Bilan economique et social Nouvelle Calédonie’. Nouméa,
New Caledonia: ISEE - Bilan economique et social, pp. 93–40.
Jaffré, T., Dagostini, G. and Rigault, F. (2003). ‘Identification, typologie et
cartographie des groupements végétaux de basse altitude du Grand Sud
Calédonien et de la vallée de la Tontouta’. Nouméa, New Caledonia:
ORSTOM.
Caplong, P. and Barjon, F. (2010). ‘Le Bulbul à ventre rouge, une star
méconnue en Nouvelle Caledonie’. La Caledonie Agricole 123: 22–25.
Jaffré, T., Dagostini, G., Rigault, F. and Coic, N. (2004). ‘Inventaire
Floristique des Unités de Végétation de la Zone d’Implantation des
Infrastructures Minières et Industrielles de Goro Nickel’. Nouméa, New
Caledonia: Institut de Recherche pour le Developpement.
Cardador, L., Lattuada, M., Strubbe, D., Tella, J.L., Reino, L., Figueira,
R. and Carrete, M. (2017). ‘Regional bans on wild‐bird trade modify
invasion risks at a global scale’. Conservation Letters. doi:10.1111/
conl.12361.
Kawakami, K., Mizusawa, L. and Higuchi, H. (2009). ‘Re-established
mutualism in a seed-dispersal system consisting of native and
introduced birds and plants on the Bonin Islands, Japan’. Ecological
Research 24(4): 741–748.
Chace, J.F. and Walsh, J.J. (2006). ‘Urban effects on native avifauna: a
review’. Landscape and urban planning 74(1): 46–69.
Kumschick S., Blackburn T.M. and Richardson D.M. (2015). ‘Managing
alien bird species: Time to move beyond ‘100 of the worst’ lists?’ Bird
Conservation International 26(2): 1–10.
Clergeau, P., Savard, J.P.L., Mennechez, G. and Falardeau, G. (1998).
‘Bird abundance and diversity along an urban-rural gradient: a
comparative study between two cities on different continents’. Condor
100(3): 413–425.
Lemay, V., Gâteblé, G., and McCoy, S. (2009). ‘Vegetative propagation
of two endemic species of Cloezia Brongn. & Gris for conservation and
mining revegetation activities in New Caledonia’. New Forests 37(1) :
1–8.
315
Island invasives: scaling up to meet the challenge. Ch 2B Other taxa: Birds
Lopes, L.E., Fernandes, A.M. and Marini, M.Â. (2005). ‘Diet of some
Atlantic forest birds’. Ararajuba 13(1): 95-103.
Lowe, S., Browne, M., Boudjelas, S. and De Poorter, M. (2000) ‘100 of
the World’s Worst Invasive Alien Species A Selection from the Global
Invasive Species Database’. Auckland, New Zealand: The Invasive
Species Specialist Group (ISSG).
Matthysen, E. (2005). ‘Density‐dependent dispersal in birds and
mammals’. Ecography 28(3): 403–416.
May, R.M. (1974). ‘Biological populations with nonoverlapping
generations: stable points, stable cycles, and chaos’. Science 186(4164):
645–647.
Meyer, J-Y. (1996). ‘Status of Miconia calvescens (Melastomataceae),
a dominant invasive tree in the Society Islands (French Polynesia)’.
Pacific Science 50: 66–79.
Myers, N., Mittermeier, R.A., Mittermeier, C.G., Da Fonseca,
G.A. and Kent, J. (2000). ‘Biodiversity hotspots for conservation
priorities’. Nature 403(6772): 853–858.
Miller, D.L. (2016). Distance: Distance Sampling Detection Function
and Abundance Estimation. R package version 0.9.6. <https://CRAN.Rproject.org/package=Distance>.
Munzinger, J., Morat, P., Jaffré, T., Gâteblé, G., Pillon, Y., Tronchet, F.,
Veillon, J-M. and Chalopin, M. (2016). FLORICAL: Checklist of the
vascular indigenous flora of New Caledonia. vers. 22.IV.2016. <http://
www.botanique.nc/herbier/florical>.
Nijman, V. (2010). ‘An overview of international wildlife trade from
Southeast Asia’. Biodiversity and conservation 19(4): 1101–1114.
Oerke, E.C. (2006). ‘Crop losses to pests’. The Journal of Agricultural
Science 144(01): 31–43.
Pinheiro, J., Bates, D., DebRoy, S., Sarker, D. and R Core Team (2017).
nlme: Linear and Nonlinear Mixed Effects Models. R package version
3.1–131. 2017.
Quantum GIS Development Team (2016). Quantum GIS Geographic
Information System. Open Source Geospatial Foundation
Project. <http://qgis.osgeo.org>.
R Core Team (2017). R: A language and environment for statistical
computing. Vienna, Austria: R Foundation for Statistical Computing.
<https://www.R-project.org/>.
Radhakrishanan, P. and Asokan, S. (2015). ‘Population density of redvented bulbul Pycnonotus cafer in a portion of Cauvery Delta Region,
Southern India’. Journal of Life Sciences Research 2(1): 1–4.
Ralph, C.J., Sauer, J.R. and S. Droege. (1995). ‘Managing and monitoring
bird populations using point counts: standards and applications’. In: C.J.
Ralph, J.R. Sauer and S. Droege (eds.) Monitoring Bird Populations by
Point Counts, pp. 161–168. USDA Forest Service General Technical,
USA: Report PSW-GTR-149.
Ricciardi, A., Blackburn, T.M., Carlton, J.T., Dick, J.T., Hulme, P.E.,
Iacarella, J.C., and Pyšek, P. (2017). ‘Invasion science: A horizon scan
of emerging challenges and opportunities’. Trends in Ecology and
Evolution 32(6): 464–474.
Russell, J.C. and Blackburn, T.M. (2017). ‘The rise of invasive species
denialism’. Trends in Ecology and Evolution 32(1): 3–6.
Russell, J.C., Meyer, J.Y., Holmes, N.D. and Pagad, S. (2017).
‘Invasive alien species on islands: impacts, distribution, interactions
and management’. Environmental Conservation: doi:10.1017/
S0376892917000297.
Sherman, J. and Fall, P. (2010). ‘Observations on feeding frequencies
among native and exotic birds and fruit bats at Erythrina variegata and
Dysoxylum trees on American Samoa’. In: S. Haberle, J. Stevenson, and
M. Prebble (eds.) Altered Ecologies: fire, climate and human influence
on terrestrial landscapes, pp.101–116. Canberra, Australia: Australian
National University Press.
Simberloff, D., Parker, I.M., and Windle, P.N. (2005). ‘Introduced species
policy, management, and future research needs’. Frontiers in Ecology
and the Environment 3(1): 12–20.
Smith, S.A., Sadlier, R.A., Bauer, A.M., Austin, C.C. and Jackman, T.
(2007). ‘Molecular phylogeny of the scincid lizards of New Caledonia
and adjacent areas: evidence for a single origin of the endemic skinks of
Tasmantis’. Molecular Phylogenetics and Evolution 43(3): 1151–1166.
Spotswood, E.N., Meyer, J-Y. and Bartolome, J.W. (2012). ‘An invasive
tree alters the structure of seed dispersal networks between birds and
plants in French Polynesia’. Journal of Biogeography 39(11): 2007–
2020.
316
Spotswood, E.N., Meyer, J-Y. and Bartolome, J.W. (2013). ‘Preference for
an invasive fruit trumps fruit abundance in selection by an introduced
bird in the Society Islands, French Polynesia’. Biological invasions
15(10): 2147–2156.
Su, S., Cassey, P. and Blackburn, T.M. (2016). ‘The wildlife pet
trade as a driver of introduction and establishment in alien birds in
Taiwan’. Biological invasions 18(1): 215–229.
Thibault, M., Vidal, E., Potter, M.A. and Brescia, F. (2018a). ‘The
red-vented bulbul (Pycnonotus cafer): serious pest or understudied
invasive?’. Biological invasions 20(121): 121–136.
Thibault, M., Vidal, E., Potter, M.A., Sanchez, T. and Brescia F. (2018b).
‘The invasive red-vented bulbul (Pycnonotus cafer) outcompetes native
birds in a tropical biodiversity hotspot’. PLOS one 13(2): e0192249.
Thomas, L., Buckland, S.T., Rexstad, E.A., Laake, J.L., Strindberg, S.,
Hedley, S.L., Bishop, J.R.B., Marques, T.A. and Burnham, K.P. (2010).
‘Distance software: design and analysis of distance sampling surveys
for estimating population size’. Journal of Applied Ecology 47(1): 5–14.
Walker, R. (2008). ‘The red-vented bulbul, superbird?’. ‘Elepaio 68:
71–78.
Watling, D. (1978). ‘Observation on the naturalised distribution of the
red-vented bulbul in the Pacific, with special reference to the Fiji
Islands’. Notornis, 25: 109–117.
Watling, D. (1979). ‘The bulbul gets a clean bill’. New Scientist 81: 963–
965.
Williams, K., Ford, A., Rosauer, D.F., De Silva, N., Mittermeier, R.,
Bruce, C. and Larsen, F.W. (2011). ‘Forests of East Australia: The 35th
biodiversity hotspot’. In: F.E. Zachos and J.C. Habel (eds.) Biodiversity
Hotspots, pp. 295–310. Berlin Heidelberg: Springer.
Yokomizo, H., Possingham, H.P., Thomas, M.B. and Buckley, Y.M.
(2009). ‘Managing the impact of invasive species: the value of knowing
the density–impact curve’. Ecological Applications 19(2): 376–386.
C.R. Abrahão, J.C. Russell, J.C.R. Silva, F. Ferreira and R.A. Dias
Abrahão, C.R.; J.C. Russell, J.C.R. Silva, F. Ferreira and R.A. Dias. Population assessment of a novel island invasive: tegu (Salvator merianae) of Fernando de Noronha
Population assessment of a novel island invasive: tegu
(Salvator merianae) of Fernando de Noronha
C.R. Abrahão1,2, J.C. Russell3, J.C.R. Silva4, F. Ferreira2 and R.A. Dias2
National Center of Conservation of Reptiles and Amphibians, Instituto Chico Mendes de Conservação da
Biodiversidade, Brazilian Ministry of Environment, Brazil. <carlos.abrahao@icmbio.gov.br>. 2Laboratory of
Epidemiology and Biostatistics, Department of Preventive Veterinary Medicine and Animal Health, School of Veterinary
Medicine, University of São Paulo, Brazil. 3School of Biological Sciences and Department of Statistics, University of
Auckland, Private Bag 92019, Auckland, New Zealand. 4Department of Veterinary Medicine, Federal Rural University
of Pernambuco, Brazil.
1
ABSTRACT Fernando de Noronha is an oceanic archipelago in the Atlantic Ocean, 345 km offshore from the Brazilian
coast. It comprises 21 islands and islets, of which the main island (FN) is 17 km2 with a rapidly growing tourism industry
in the last decades. Despite being a protected area and bearing Ramsar and UNESCO World Heritage site status, it is
threatened by multiple terrestrial invasive species since its colonisation in the early 16th century. Invasive species and
the increasing tourism contributes to a list of at least 15 endangered or critically endangered species according to IUCN
criteria. The black and white tegu (Salvator merianae) is the largest lizard in South America, occurring in most of the
Brazilian territory and reaching up to 8 kg and 1.6 m from head to tail. As an omnivorous and opportunistic lizard, it feeds
on a variety of available items, including smaller vertebrates and eggs. The introduction of the tegu to FN as well as its
immediate impact on local fauna were not recorded; however, its ongoing impact is expected to be high. We captured
and marked 103 tegu in FN during the months of February and November of 2015 and 2016. We also counted animals by
line-transect census in a sparsely inhabited and an uninhabited area of FN. Body size affected the capture probabilities,
while season and sex had little or no effect. Densities estimated by capture-recapture in the sparsely inhabited area varied
from 2.29 to 8.28 animals/ha according to sampling season. Line transect census in the same area revealed a density of
3.98 (±1.1) animals/ha and in the uninhabited area 13.83 (±3.9) animals/ha. Home range was 10.54 ha, ranging from 7.36
to 15.33 hectares. Tegu activity decreased in the months of July and August of 2015. Results from this study can assist
conservation managers and decision makers to implement a science-based tegu management programme in the future.
Keywords: conservation, invasive species, lizard, oceanic island, reptile, Salvator, Teiidae, Tupinambis
INTRODUCTION
Islands are simplified ecosystems where each species
plays an important role in its functioning (Simberloff,
1974). In these environments, the loss of a species and
its functional role are not easily replaced, as would be
the case in more species-rich ecosystems such as on
continents. Despite corresponding to only about 5%
of land area globally, islands contain more than 15% of
terrestrial biodiversity (Tershy, et al., 2015). A lack of
certain behaviour or life-history traits makes native insular
species more vulnerable to the impacts of invasive species
(Vitousek, 1988; Tershy, et al., 2015).
Introduction of invasive species is one of the major
causes of contemporary biodiversity loss (Vitousek, et al.,
1997; Chapin, et al., 2000). On islands, it is probably the
major cause (Veitch & Clout, 2002; Reaser, et al., 2007).
Direct and indirect competition, predation and introduction
of diseases are some of the negative influences that invasive
species can bring to native populations (Wyatt, et al., 2008;
McCreless, et al., 2016; Russell, et al., 2017). Invasive
predators are implicated in at least 58% of the worldwide
contemporary extinctions for birds, mammals and reptiles
(Doherty, et al., 2016). The insular ecosystem frailty
combined with invasive species results in islands bearing
37% of all critically endangered species and 61% of all
recorded extinct species, according to the IUCN Red List
(Tershy, et al., 2015). Furthermore, the impact of invasive
species is not constrained to local biodiversity, but also
affects the economy, agriculture, health and human culture
(Russell, et al., 2017)
Some invasive species, such as rodents, are globally
widespread and their impacts on islands have been well
described (Reaser, et al., 2007; Russell, et al., 2017).
However, some invasive predators are only found
regionally or locally and their impacts and management
are not fully understood (see Eales, et al., 2010; Powell,
et al., 2011). Those less well-known species must not be
overlooked, as their impact might be equal to, if not larger
than, common widespread invaders (Phillips, et al., 2007;
Simberloff, 2009; Dorcas, et al., 2012; Neves, et al., 2017;
Russell, et al., 2017).
Fernando de Noronha
Fernando de Noronha archipelago consists of 21 islands
and islets, 340 km offshore from the northeast Brazilian
coast. The total land area of the archipelago is 18 km2
where the main island, also called Fernando de Noronha
(FN) is about 16.7 km2. The archipelago is a UNESCO
world heritage site (since 2001) and has recently been
named as a Ramsar site. Fernando de Noronha archipelago
is an important breeding site for several species of birds,
sea turtles and reptiles, some endemic and threatened with
extinction (Sazima & Haemig, 2012; Reis & Hayward,
2013). At the moment, at least 22 invasive species of
plants and animals are known in the archipelago (Sampaio
& Schmidt, 2014).
The local economy is fundamentally based on
tourism, with minimal production of goods and other
services. The number of inhabitants on FN has increased
substantially within the last decade due to a lack of control
from Pernambuco State and the opportunities created by
the growing tourism (Gasparini, et al., 2007). The total
number of human inhabitants is debateable, with available
information varying from two to five thousand people,
with an additional up to three thousand tourists per year
in the peak seasons (Andrade, et al., 2009; Marinho, 2016;
IBGE, 2017; Pernambuco, 2017).
Urbanised areas are restricted to the main island, in
the environmental protected area (APA), a protected area
with sustainable use of natural resources – IUCN category
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 317–325. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
317
Island invasives: scaling up to meet the challenge. Ch 2C Other taxa: Herpetofauna
VI – of approximately 8 km2. The remainder of the main
island, including the other islands and islets from the
archipelago, is uninhabited and constitutes the National
Park (PARNAMAR), where only indirect use is permitted
– IUCN category II.
Tegu lizard
The black and white tegu lizard (Salvator merianae
syn. Tupinambis merianae) (Fig. 1), hereby referred to as
tegu, is the largest lizard in South America, up to 160 cm
in total length and weighting up to 8 kg in its native range
(Lopes & Abe, 1999; Andrade, et al., 2004). In their natural
distribution in South America, tegu are commonly seen
living and feeding close to inhabited areas, as well as forested
areas (Oren, 1984; Sazima & Haddad, 1992; Bovendorp, et
al., 2008; Winck, et al., 2011; Klug, et al., 2015; Muscat, et
al., 2016). This omnivorous, opportunistic species feeds on
fruits, vegetables, insects, small vertebrates, garbage and
even carcasses when available (Sazima & Haddad, 1992;
Kiefer & Sazima, 2002; Manes, et al., 2007; Bovendorp,
et al., 2008; da Silva, et al., 2013; Muscat, et al., 2016).
In South America, they can be found from south of the
Amazon River to Argentina (Presch, 1973; Lanfri, et al.,
2013; Passos, et al., 2013). In most areas where the tegu
occurs, they are hunted for their skin and meat (Oren, 1984;
Alves, et al., 2012), which has warranted the inclusion of
the species on the CITES II appendix (UNEP-WCMC,
2014). In South America, adult females can lay up to 54
eggs per year (Donadio & Gallardo, 1984) and in captivity
this species can possibly live up to 20 years (Brito, et al.,
2001). The tegu is also considered an invasive species in
Florida, where it is suspected to have a large impact on
the already impacted local fauna (Pernas, et al., 2012;
Mazzotti, et al., 2015).
Available data indicate that the tegu was deliberately
introduced to the main island of Fernando de Noronha at the
beginning of the 20th century (Santos, 1950), despite other
publications suggesting a different period of introduction
(e.g. Oren, 1984; Silva-Jr., et al., 2005). Whether to serve
as hunting game or to help the control of rodents and toads,
reasons for the introduction of tegu are speculative (Oren,
1984; Gasparini, et al., 2007; Ramalho, et al., 2009).
Descriptions of FN fauna prior to the 20th century don’t
mention the tegu, despite mentioning the other endemic
reptiles on the archipelago (Branner, 1888; Ridley, 1890).
In the last century, very little was done to study the tegu
population and impacts on the island ecosystem. Control
or eradication methods were also never attempted, despite
the management of the tegu being considered important to
Fig. 1 Juvenile of Salvator merianae at Sancho Beach,
Fernando de Noronha (photo: Vinicius Gasparotto).
318
promote the conservation of endangered species living on
the island (Brasil, 2004).
We provide up to date information on the tegu population
size and structure on Fernando de Noronha to contribute to
an informed control programme to be undertaken by island
conservation managers in the future.
METHODS
Study areas
To access the tegu population in the archipelago we
selected two representative areas from the main island and
visited the main vegetated islets that are used as nesting
sites by resident birds. Land use in FN was simplified
into three types, according to human usage: i) Densely
inhabited areas, including hotels, houses and commercial
buildings, paved streets and dense traffic, also with a
higher density of uncontrolled dogs and cats; ii) Sparsely
inhabited areas, including: rural areas similar to those
found on the continent, and small villages with unpaved
roads and sparse houses surrounded by crops and livestock
animals. These two inhabited areas constitute most of
the APA land; iii) Uninhabited areas, including areas of
natural vegetation and secondary regeneration, with a few
abandoned buildings and sporadic tourist usage. This area
constitutes most of the PARNAMAR land (Fig. 2).
Within the inhabited areas, we chose the Boldró village
that is a good representation of a sparsely inhabited area,
with tourist visits, a small amount of commerce, paved and
unpaved roads and houses of local workers. It is common
to find domestic animals (dogs, cats, chickens), and crops
and fruit trees in backyards. In the PARNAMAR we chose
the southwestern Capim-açu region that represents the
most intact? area of native vegetation on the main island
(Mello & Adalardo de Oliveira, 2016). In Boldró village
we performed a mark-recapture study and a line transect
census study. In Capim-açu we performed a line transect
census only.
Mark-recapture
To apply this method we chose the Boldró village
located in a sparsely inhabited area of FN. This area is
representative of the most common vegetation types on
the main island and is subject to various levels of human
interference while leaving space for native vegetation.
Sampling seasons occurred during the years of 2015 and
Fig. 2 Map of the protected areas of Fernando de Noronha
Archipelago. Note: 1 is Boldró village transect, and 2 is
Capim-açu transect (Land use layer by Vívian Uhlig –
RAN/ICMBio).
Abrahão, et al.: Population assessment of tegu, Fernando de Noronha
2016, with 14 to 19 days of consecutive sampling in the
beginning (Jan–Feb) and end (Oct–Nov) of the dry season.
We opted to sample in the summer-spring as this species
has been known to hibernate during the autumn-winter
seasons on the continent (Andrade, et al., 2004; de Souza,
et al., 2004).
We used ten funnel traps made out of PVC pipes
(150 mm × 1 m) with one end closed. Those traps were
placed in shaded spots next to vegetation borders, next to
habitations, restaurants and areas that a tegu could use for
hiding or foraging (Fig. 3). Each trap was placed in a stable
position over trunks or stones in order to maintain at least
a 20 degree angle to the closed end. The inclined position
and lack of friction provided by the PVC material prevents
animals from leaving the trap, where they remain until
release. Raw chicken was used as baits and replaced every
two days. Tegu locates the bait through smell (Yanosky, et
al., 1993) and enters through the higher open entrance of
the trap to get the bait that rests in the closed lower end of
the pipe.
Traps were checked at the end of each day, when the
individuals become inactive. Every animal was then
restrained and marked with a transponder implanted
subcutaneously. Snout vent length (SVL) was measured
to the nearest 0.5 cm, with the use of a tape measure.
The weight was taken using a Pesola® scale with a 10 g
precision. Animals recaptured in the same season (e.g. less
than 30 days interval) were considered to have the same
weight and length, thus these data were collected only on
the first capture of the season.
To estimate density (D) through mark-recapture
data, we used the maximum-likelihood spatially explicit
capture-recapture (ML SECR) package from R (Team,
2000; Borchers & Efford, 2008). We assumed a Poisson
distribution of range centres (i.e. random) with a halfnormal curve detection function parameterised by g0
(probability of detection when trap and range centre
coincide) and σ (spatial scale of the detection function).
Removals from the population (i.e. poaching or death) are
assigned known capture histories of 0 with probability
equals 1 following death. The conditional likelihood was
used to derive density, incorporating individual covariates
of SVL and sex. Models were compared using an AIC
framework, but due to sparse data, subsets of models on
σ and then g0 were considered independently. The area
of capture exposure, which usually would be related to
an individual’s home-range, was approximated by a 95%
circular probability density area of capture as:
Line-transect census
Two tracks were chosen to undertake the census
counting (Fig. 2). One in the Boldró village, 1,820 m in
length, to make possible comparisons between density
methods in the same area, another in the Capim-açu track,
2,000 m in length, to make possible a comparison between
a sparsely inhabited area and uninhabited area. A trained
volunteer walked each track counting tegu in the high
activity hours (10 a.m. to 2 p.m.). For six days, the Capimaçu track was walked in one direction and after a 30 min
break at the farthest point, it was walked back. Atypical
days with rain, temperatures below 25°C or excessive wind
were avoided to prevent weather interference on abundance
data. Counting along Boldró track was repeated nine times
and Capim-açu 35 times during this study. When a tegu
was sighted, the observer took the perpendicular distance
of the animal from the centre of the track using a scale
tape, to the nearest 0.5 m and up to 20 m distance. Any
tegu sightings over 20 m of distance were discarded, but
the thick vegetation in this region prevents seeing animals
in the vegetated area on the transect borders.
We calculated the density of animals along the transect
using distance sampling analysis, but zero spiking in the
data (excessive observations close to the line) violated basic
premises, likely due to a much higher level of detection,
and potentially tegu abundance, along the clear open tracks
in the dense forest. We subsequently used the line-transect
census methodology (Burnham, et al., 1980; McDiarmid,
et al., 2012) on a subset of the data, for observations
directly on the open track only. Total number of individuals
observed along the line-transect were used to represent the
abundance on the track area, assuming every individual
within the transect was observed. The area was calculated
by using the average width of the track (measured every
100 m) and then multiplied by its length. Open areas were
not measured and were assumed to have the same average
width as the forested areas. Only animals observed within
the established width of the track (e.g. clear area) were
considered for such analysis.
We used a two-tailed t-test with unequal variances to
compare daily density data between Boldró and Capimaçu. Only the high activity months (Feb–Jun, Sep–Nov)
were used for this comparison, since we did not have data
from the dry season in the Boldró area. The same method
was used to compare densities observed on Capim-açu in
the high activity months and low-activity months (Jul–
Aug). To coarsely calculate the total abundance of tegu
in FN, we stratified the map according to three main land
uses: i): densely inhabited areas (226 ha); ii) sparsely
inhabited areas (960 ha), and iii) uninhabited areas (417
ha) (Fig. 2). Average density and ranges from Capim-açu
line-transect counts were used to estimate the abundance
of tegu in the uninhabited areas of FN. The same method
was used in Boldró to estimate the abundance of tegu in
the sparsely inhabited areas of FN. Densely inhabited areas
and areas with no vegetation (e.g. beaches, sand dunes and
rocky areas – 97 ha) were excluded from the abundance
calculations for they were not represented in the study area
and were considered poor tegu habitat.
Islet surveys
Fig. 3 PVC Funnel trap to catch tegu mounted near a tree
at the edge of a clearing.
We visited seven of the larger vegetated offshore islets
of the archipelago (Rata, Rasa, do Meio, Conceição, Morro
Dois Irmãos, Morro da Viuvinha and Morro do Chapéu)
at least once during the study period (Fig. 2). We spent
from one to twelve hours actively searching on each islet,
searching for sightings or indirect signs of tegu presence
(tracks or burrows). We also inquired with local inhabitants
and other researchers for records of tegu presence on the
319
Island invasives: scaling up to meet the challenge. Ch 2C Other taxa: Herpetofauna
other islands, since tegu can swim and also could have
been brought to other islands intentionally in the past.
RESULTS
In the mark-recapture study we had a total of 190
captures over 69 trapping days in the Boldró village. From
the 190 captures, we captured 103 unique individuals with
87 recaptures. Of the ten traps installed, two had to be
moved in the last sampling season to avoid interference by
people. These traps remained a total of 55 days in the first
location and 14 days in the second location, less than 50 m
away from where they were previously placed. Since tegu
weight and size (SVL) were highly correlated (R2=0.84),
we have chosen only SVL as a covariate on σ and g0.
SVL also provides a better measure than total length,
for it excludes the tail that can be lost or be regenerated
to a variable size. Given the relatively low number of
recaptures, we had to specify reasonable starting values for
the likelihood maximisation with starting values of g0 = 0.1
and sigma = 50 from a preliminary inspection of the data.
Ranging behaviour and probability of capture
We first fitted and ranked models combining the
influence of sex and size on the ranging behaviour (σ) of
the animals, while keeping a fixed capture probability (g0).
The simplest model, with fixed probability of capture and
fixed ranging behaviour had 91% support, showing that
size has no effect on the ranging behaviour of animals,
while sex has little effect (Table 1).
Based on the best adjusted model for ranging behaviour,
we kept σ constant across sessions and tested the influence
of sampling period, sex and size on the probability of
capture of the individuals. As seen in Table 2, the model
including SVL had 44% support showing that body size
as a continuous variable is the most important of the tested
covariates to affect the probability of capture. Session also
showed some importance in explaining the variation as
seen in models 2 and 3.
Home-range
To produce real estimates for capture probabilities (g0)
and ranging behaviour (σ), we took the model including
the most important covariates (session and size), for
probability of capture and fixed ranging behaviour. The
average size (SVL) used in the estimates was 30.2 cm
(Table 3).
period of study. The calculated density for Boldró is 3.98
(±1.1) animals/ha. The Capim-açu transect (0.492 ha)
was surveyed 35 times from February 2015 to February
2016, with a total linear effort of 70 km. In this transect,
260 animals were sighted within the established average
width of 2.46 m during the study. The calculated density
for Capim-açu is 13.83 (±3.9) animals/ha.
Densities calculated using the line transect method
were different between Boldró village and Capim-açu
transects (t=6.45, P≤0.00001). There were no surveys in
the Boldró transect during the low-activity months, thus,
only densities from high-activity months in both transects
were used to compare the densities averages from different
areas. In Capim-açu, densities also differed between highactivity months and low-activity months (t=3.29, P≤0.01).
The number of sightings on each occasion for Capim-açu
transect is shown in Fig. 4 where a decline in number of
sightings can be seen in the months of July and August.
To estimate the abundance of tegu in FN we used the
calculated uninhabited area of FN as being 417 ha and
the total sparsely inhabited area of FN as being 960 ha
(see Fig. 2). Considering Capim-açu transect densities,
calculated abundances range from 4,141 to 7,393 tegu in the
uninhabited area. Using densities from Boldró transect for
the sparsely inhabited areas, we estimated abundance from
2,765 to 4,877 tegu in that area. Total number of animals
estimated for both calculated areas is from 6,906 to 12,270
tegu. High-density inhabited areas and non-vegetated areas
of the island (463 ha) were excluded from this calculation
for they were not represented in the samples; however, tegu
are expected to be using those areas in a lower rate, thus
abundance results should be taken as an underestimation of
the whole population.
Population parameters
Males constituted the majority of the sampled
population in all but the first sampling period. Males were
also larger and heavier than females in all sampled periods.
Male weight ranged from 400 g to 2,450 g and female
weight ranged from 600 g to 1,940 g. Snout–vent (i.e.
body) length for males ranged from 24 to 40 cm and for
females from 26 to 36 cm. Averages and range by season
and sex are given in Table 5.
Based on real parameters obtained from the chosen
model, we calculated 95% home ranges (HR95) for average
size and both sexes as 10.54 ha, ranging from 7.26 to 15.33
ha.
Density, abundance and activity
Finally, we estimated densities and sampled areas for
each sampling season over the chosen model (Table 4).
In the line transect study, the Boldró transect (0.419
ha) was surveyed six times in the high-activity months
(Nov 2015 and Feb 2016), with a total linear effort of
10.92 km. Only ten animals were sighted in this transect
within the established transect width of 2.3 m during the
Fig. 4 Number of sightings of tegu in the Capim-açu
transect during the 2015 sampling period. The line
represents a moving average of three samples.
Table 1 Model results of tegu detection function for covariates of the scale parameter (σ) and the
probability of capture equal to the home range centre (g0).
σ Models
sigma~1
sigma~sex
sigma~SVL
320
Detection function
Half normal
Half normal
Half normal
Npar
2
3
3
Log likelihood
-609.159
-610.382
-766.307
AICc
1,222.411
1,226.950
1,538.800
Rank
1
2
3
Weight %
91%
9%
0%
Abrahão, et al.: Population assessment of tegu, Fernando de Noronha
Table 2 Best adjustment on models tested for constant probability of capture (g0) with covariates as sampling
size (SVL), sex, season (session) and ranging behaviour (sigma).
g0 Models
g0~svl
g0~session + svl
g0~session
g0~1
g0~sex + svl
g0~session + sex + svl
g0~sex
g0~session + sex
Detection function
Half normal
Half normal
Half normal
Half normal
Half normal
Half normal
Half normal
Half normal
Npar Log likelihood
3
-605.896
6
-602.833
5
-605.242
2
-609.159
4
-608.020
7
-605.046
3
-610.016
6
-608.184
Table 3 Estimates of real parameters for σ and g0 in each
sampling season, using average size of 30.2 cm SVL.
Given standard errors and 95% confidence intervals
(lower class and upper class).
Real
parameters
SVL=30.2
g0 Feb/2015
g0 Nov/2015
g0 Feb/2016
g0 Nov/2016
σ
Estimate
0.012
0.035
0.032
0.039
74.780
SE
0.006
0.010
0.011
0.010
7.148
lcl
ucl
0.004
0.019
0.016
0.023
62.030
0.033
0.062
0.062
0.064
90.150
Islets
From the seven visited islets of the archipelago,
only Rata Island had indirect records of the presence of
tegu. There was an effort of 58.5 person-hours of active
searching, plus 72 trap-hours divided among three visits
to Rata Island, but no direct sights or captures were made.
Tracks, faeces and burrows were found, indicating the
presence of tegu, possibly at a lower density than the main
island.
DISCUSSION
Policy makers, managers and the general public need
to be informed of the consequences of invasive species
in order to manage their impacts. Understanding the
population biology of an invasive species is a first step to
acquire essential information for management decisions
that may alleviate impacts. Despite Fernando de Noronha
being inhabited since the 16th century, very little has been
done to understand or prevent the impact of invasive
species on endangered and endemic species that struggle
to coexist in the archipelago (Sampaio & Schmidt, 2014;
Mello & Adalardo de Oliveira, 2016; Dias, et al., 2017).
Ranging behaviour and probability of capture
Spatial detection models show that size and sex had
little influence on tegu ranging behaviour on FN. Klug,
et al. (2015) found that size differences were not likely
AICc
1,217.978
1,218.333
1,220.957
1,222.411
1,224.353
1,224.988
1,226.219
1,229.034
Rank
1
2
3
4
5
6
7
8
Weight %
44.26
37.06
9.98
4.82
1.83
1.33
0.72
0
to be contributing to movement differences for tegu. In a
subtropical coastal region on southern Brazil, Winck, et al.
(2011) found tegu to be more active when temperatures
start rising by the end of spring and early summer. They
also related peaks of activity while males were dispersing
and after the emergence period, to be due to the beginning
of foraging and sexual activity. The present study does not
capture full seasonal variation because of time-constrained
sampling, but a drop in activity was observed in July
and August, as observed in other tegu studies (Winck &
Cechin, 2008; Tattersall, et al., 2016). This small window
of low activity of tegu on FN may not promote a significant
variation in relation to the impacts it causes to other species.
On the main island, only masked booby (Sula dactylatra)
still nest on the ground in a small peninsula next to the end
of Capim-açu track. Their eggs are laid in the first months
of the year as observed by e Silva & Neves (2008) on
secondary islands of the archipelago. The Noronha skink
(Trachylepis atlantica) is also a common prey item in the
tegu diet. Despite being relatively abundant, nothing is
known about its reproduction. It is thought to reproduce
throughout the year as for Trachylepis sechellensis on the
Seychelles, another tropical archipelago (Bringsøe, 2008).
Sea turtle nests are also preyed upon by tegu, as recorded
by TAMAR project for Chelonia mydas on FN (Bellini &
Sanches, 1996; e Silva & Neves, 2008), including predation
of hatchlings (Ayrton K. Péres-Jr, pers. comm.). Turtles on
FN nest from January to June, when tegu are active.
For the probability of capture, size was an important
determinant, but population studies using traps often fail
to collect a broad representative sample of the population
as seen in Carter, et al. (2012). A hole of 3 cm diameter
was made in the closed end of the pipe to avoid flooding
of the trap and unwanted capture of native lizards. This
safety measure may bias the sample as it allows small
animals to escape. These animals would possibly not be
able to be marked by transponder implant and thus were
of less importance for this study in any case. Behavioural
traits such as niche separation due to intraspecific
competition could also explain a size interference on
capture probabilities (Herrel, et al., 2006; Siqueira &
Rocha, 2008). The observed small influence of season on
capture probabilities is primarily in the first session and
possibly due to adjustments of methodology following that
Table 4 Densities and estimated sampling areas in Boldró village for each season sampled derived from
the best adjustment models.
Period
Feb/2015
Nov/2015
Feb/2016
Nov/2016
Density/ha
4.19
4.45
3.59
5.07
Std. Error
0.93
0.94
0.84
1.05
Min (95%)
2.72
2.95
2.29
3.39
Max (95%)
6.44
6.71
5.64
7.59
ESA
7.40
7.42
7.52
8.28
321
Island invasives: scaling up to meet the challenge. Ch 2C Other taxa: Herpetofauna
Table 5 Tegu sex, size and weight averages with ranges in each sampling season.
Period
Feb 2015
Sex
n (%)
x̅ SVL (cm)
SVL range
Weight range
M
15 (42)
33.58
29–37
1,491.00
875–2,175
F
21 (58)
31.22
28–36
1,051.67
640–1,940
M
21 (60)
32.26
28–39
1,347.14
740–2,240
F
14 (40)
29.35
26–33
965.00
600–1,550
Feb 2016
M
18 (55)
33.28
28–37
1,368.89
660–1,930
F
15 (45)
30.40
27–34
914.67
600–1,590
Nov 2016
M
47 (68)
32.53
24–40
1,294.26
400–2,450
F
22 (32)
30.67
26–35
1,030.45
600–1,560
Nov 2015
first sampling season. A variation in capture probabilities
is not expected once sampling seasons were chosen within
the high-activity periods for tegu.
Home-range
In the tegu natural distribution, older males have larger
territories, while juvenile males and females have smaller
territories with higher overlap. A peak of activity in males
was observed at the end of the low-activity period (Winck,
et al., 2011). A decrease in home range after the mating
season was also observed in the El Palmar National Park,
in Argentina (Fitzgerald, et al., 1991). Results from this
study suggest little influence of sex on tegu capture in
traps. Since SECR only estimates spatial exposure area to
traps, sex could be affecting the ranging behaviour of tegu
in FN but this method is not precise enough to detect such
variation. This result may also have been affected by biases
in the capture probability of certain tegu size classes (i.e.
juveniles).
Lirio, et al. (2004) tracked six radio-implanted tegu in
FN and estimated home-ranges varied from 0.73 to 7.8 ha
(3.3 ha on average). The authors also comment that a gravid
female was used in the study, representing the smallest
home-range, and that the activity centre was usually close
to the shelter. Winck, et al. (2011), found home-ranges from
0.05 to 26.4 ha for a continental population in southern
Brazil. Home-ranges as measured in the present study are
within the previous findings for the natural distribution of
tegu and are a little higher than those described by Lirio,
et al. (2004).
Since σ did not differ between sexes, estimated homeranges were considered the same for males and females.
Home-ranges can provide necessary information to set
management on invasive species, such as the density of
control devices (Hays & Conant, 2007; Howald, et al., 2007;
Anderson, et al., 2016). For continental tegu, behavioural
traits such as season, age and reproductive status can be
implicated in home-range variation (Winck, et al., 2011).
In FN, factors such as the lack of competitors, predators
and resource availability could be also influencing tegu
home ranges (Ballinger, 1977; Shine, 1987; Novosolov, et
al., 2016). With an average home range (HR95) of 10.54
ha, tegu on FN are quite mobile. This behaviour allows
them to look for resources in a vast area and feed even
when resources are not abundant (e.g. dry season). We also
noticed an overlap of territories throughout the year, as
juveniles forage together and coexist with adult males and
females in the same area. Only youngsters seem to avoid
larger tegu, having more secretive habits. In general a large
home range also increases the probability of a species being
exposed to a control method (Howald, et al., 2007). That
means managers might need fewer traps (e.g. one every
few ha) in order to control tegu on FN.
322
x̅ Weight (g)
Density, abundance and activity
In Boldró village, density estimates from capturerecapture ranged from 2.29 to 7.59 animals/ha while
estimates from the line transect census ranged from 2.88
to 5.08 animals/ha. Those densities are much higher than
the 0.83 animals/ha observed for a tegu population living
in Anchieta Island or the 0.63 animals/ha as seen in the
Espírito Santo Atlantic rainforest, both in south-eastern
Brazil (Bovendorp, et al., 2008; Chiarello, et al., 2010).
These higher estimates could be due to a tropical climate in
FN that favours reptiles with a low variation in temperature
over the year. Abundance of resources and the lack of
natural predators can also contribute to the higher density
observed in FN as seen for other island invasive predators
(Pekelharing, et al., 1998; Hays & Conant, 2007; Ferreira,
et al., 2012).
Since density estimates from both methods used in
this study (line transect census and mark-recapture) were
similar, we opted to use values from the line transect
census because it also provided density for the Capim-açu
transect. Density from those transects was applied to the
region represented by each transect to obtain abundance for
both represented areas. There is a possible error associated
with the extrapolation of the transect densities over the
whole area, especially to areas with dense vegetation, as
observers may find a higher number of tegu using the open
areas, causing an overestimation of density. However, a
similar density estimated by two different methods supports
the idea of transect counts being a reliable method, despite
the associated error. An estimate of abundance can help
management decisions in quantifying the effort and costs
required to control or eradicate (Holmes, et al., 2015; Keitt,
et al., 2015). Density estimated in Capim-açu was higher
than that estimated in Boldró and a broad list of factors
could explain such differences, the most important are
discussed here.
Animals are not distributed uniformly in the environment
and they tend to occupy environments that seem more
favourable, while less favourable habitats are occupied
in lower densities (Diaz & Carrascal, 1991; Fraga, et al.,
2013). In FN, presence of predators and competitors, such
as cats, could negatively affect tegu populations by preying
on juveniles and hunting other potential prey of tegu such
as rats and other reptiles. In Boldró village and other highdensity inhabited areas of the island, the influence of cats
is higher, since the cat population is denser when closer
to inhabited areas (Dias, et al., 2017). Dogs also inhibit
presence of tegu by chasing and killing tegu when they
cross territories, making inhabited areas again less suitable
for tegu (C.A. pers. obs.).
Tegu are appreciated for their meat in the northeast of
Brazil, where the species can be a delicacy and an important
source of protein in poor communities (Mendonça, et al.,
2011; Nóbrega Alves, et al., 2012). Poaching of tegu in FN
is driven by different reasons, with tegu being commonly
Abrahão, et al.: Population assessment of tegu, Fernando de Noronha
hunted by poultry farmers when they break into henhouses
to eat eggs and chicks. Hunting in FN is done with fishing
line and hooks, baited with fish or chicken, in the areas
close to residences (C.A. pers. obs.). Tegu abdominal fat
is also widely known as a medicine and is used by locals
to treat sore throat, earache and other ailments (Nóbrega
Alves, et al., 2012). Those properties are scientifically
based since the anti-inflammatory properties of tegu fat has
been proven (Ferreira, et al., 2010).
says tegu were abundant there during that time. It seems
that after his family left Rata, the population of tegu has
decreased. However, the island seems to be big enough to
maintain a small population of tegu. Some animals might
also occasionally swim to other islets, but even a single
animal could hardly live for long on the scarce resources
available on those smaller islands, forcing them back to the
main island.
Tegu are generalists and feed on any available
resources, including vegetation, fruits, insects, vertebrates
and eggs (Vanzolini, et al., 1980; Kiefer & Sazima,
2002; Mourthé, 2010). Those adaptations do not restrict
resources for the tegu population in FN, where it possibly
lives with plenty of food throughout the year. A reduction
in the tegu population is more likely to be present in human
altered environments such as densely inhabited areas,
with negative effects of domestic animals and poaching,
despite a possible higher availability of food (crops, fruit
trees and rubbish). Another factor that could be affecting
the results is of behavioural origin. The negative impact
of human presence seems to make the tegu population
shift towards uninhabited areas that offer better habitat
with less interference and still plenty of resources. Despite
density underestimation being a possibility when failing to
observe all animals on the transect (e.g. behaviour to avoid
human contact in inhabited areas), the more intensive
mark-recapture study showed similar estimates of density
thereby disproving a possible methodological interference.
Future steps
Population parameters
Size in this study was inferred by SVL and was also
closely correlated to weight. Although, size can be affected
by external factors when trying to infer individuals’ ages
(Halliday & Verrell, 1988; Adolph & Porter, 1996), weight
can also reflect body condition and be influenced by the
loss of the tail, a common finding in the FN population.
Size can be related to sexual maturity (Fitzgerald, et al.,
1993), while movement and home-ranges can be affected
by sex and reproductive status (Winck, et al., 2011).
Size is also related to reproductive capacity of females
(Fitzgerald, et al., 1993). Tegu on FN seem to be smaller
than those found in continental South America, thus, the
female reproduction index in FN should be lower than in
the continent (Fitzgerald et al., 1991; Winck et al., 2011).
The smaller size on FN can also be related to a much
higher density caused by lower competition and predation
rates than the ones found in the continent (Novosolov, et
al., 2016).
Males seem to be a higher fraction of the population
on FN, which might influence reproduction and population
growth (Le Galliard, et al., 2005). Sex ratio can be affected
by average temperature (e.g. natality rates) or by any factor
that increases mortality rates in only one of the sexes.
Populations of tegu in Paraguay were consistently malebiased (Mieres & Fitzgerald, 2006), possibly leading to a
higher fecundity rate of females or having a negative effect
on lizard populations as observed by Le Galliard, et al.
(2005).
Islets
Tegu are good swimmers and there are various
sightings from local residents of tegu swimming or diving
near to the main island. A video made by Elias Pereira and
Nelly Burella shows a juvenile tegu voluntarily swimming
across Baia dos Golfinhos, on the main island. Other than
swimming, tegu could have been taken to other islands in
the past for the same reason they were taken to the main
island (either to control rats or serve as a food supply).
Manoel P. dos Santos, who lived on Rata Island until 1986,
The reasons why the tegu was introduced to Fernando
de Noronha, when it happened and the impacts this predator
has caused to the archipelago were not documented and
remain unknown. However, the understanding of impacts
caused by invasive predators in islands worldwide provides
sufficient evidence that management is required in order to
protect local biodiversity. Eradication is usually the best
option when the tools are available, but when they are
absent, control measures may be better than the do-nothing
approach (Fletcher, et al., 2015; Russell, et al., 2017).
On Fernando de Noronha, managing the impacts of
tegu over native fauna is already on the list of priorities,
as documented in the management plan of the APA (Brasil,
2004). However, providing up to date information on
tegu population structure and biology in FN is expected
to contribute to the implementation of a science-based
invasive species programme in the future. Based on results
from this work and field experience of the authors in FN,
our contribution to this programme is offered here as a
suggestion to local managers and decision-makers.
Measures of tegu control in FN should be placed
in strategic locations where impacts on native fauna are
considered higher, such as ground nesting sites for birds,
nesting beaches for turtles and most preserved vegetated
areas for other reptiles, crabs and even invertebrate
fauna. Live or kill traps could be used, depending on the
destination identified for the animals. Traps like the ones
used in this study proved to be very efficient for adult tegu
and seem very cost-effective. Considering the relatively
high probability of capture observed, live traps needs to be
checked at least once a day. Traps also need to be placed in
the shade as lizards are easily prone to overheating in the
tropics. Traps can be baited with eggs, bacon, chicken, fish
or any other scent-driven attractant, since smell is the main
sense for area exploration of tegu.
Considering the very high density of animals, an
equally high number of traps should be required (one per
ha or more). Control areas can be fenced by a tegu-proof
fence to prevent quick reestablishment of the population
by recruitment of juveniles. Traps should be placed
preferentially in transition areas between vegetated and
clear areas, where tegu transit to control body temperature
during times of higher activity. Management effort should
be stronger after the low-activity period, up to the end of
the reproduction season (expected to be from September
to March in FN). However, since there are animals active
throughout the year, effort should also be made according
to the reproduction of potential prey species such as the
ground-nesting birds, sea turtle nests and crab spawning
period. Control effort is expected to be up to four times
higher in the uninhabited areas than in the inhabited areas
of FN, given tegu density variation between those areas.
There are no specific tools available to control tegu and
poison should not currently be considered as an option,
since it would also threaten other endemic reptiles in FN.
Hunting also requires special firearm permits and doesn't
seem to be an option when in a tourist location like FN. For
the moment, only fencing and trapping seem to be feasible
solutions to manage tegu impacts on the archipelago’s
biodiversity.
323
Island invasives: scaling up to meet the challenge. Ch 2C Other taxa: Herpetofauna
CONCLUSION
Some invasive species are not commonly widespread
and attract little attention of researchers. However, once
established, those species can pose a real threat to native
biodiversity (Simberloff, 2009; Neves, et al., 2017).
Tegu have been established on FN for a century (Santos,
1950), but their population structure and impacts on native
fauna remained understudied. This assessment provides
focal information for a future control programme of tegu
on Fernando de Noronha archipelago. We also aim to
contribute to a larger ongoing process in Brazil, where
invasive species move towards being a primary problem
to be addressed for biodiversity conservation. Finally, we
call on researchers worldwide to focus on other neglected
invasive insular species as they represent a challenge and a
frontier for island conservation.
AKNOWLEDGEMENTS
The authors would like to thank the Administration
of the State District of Fernando de Noronha (ADM-FN),
Instituto Chico Mendes de Conservação da Biodiversidade
(ICMBio), National Marine Park of Fernando de Noronha
and its volunteers, Area of Environmental Protection of
Fernando de Noronha, National Center for Research and
Conservation of Amphibians and Reptiles (RAN/ICMBio),
Fire Department of Fernando de Noronha, Brazilian
Institute for Conservation Medicine (Tríade) and Programa
de Pó-graduação em Epidemiologia Experimental Aplicada
às Zoonoses (VPS-FMVZ-USP). ICMBio and Capes
(PVE Project Number 88881.065000/2014-1) provided
financial support to this study. This study was taken under
SISBIO permit no. 41682 and USP ethics committee no.
2724150515.
REFERENCES
Adolph, S.C. and Porter, W.P. (1996). ‘Growth, seasonality and lizard life
histories: Age and size at maturity’. Oikos 77: 267–278.
Alves, R.R.N., Vieira, K.S., Santana, G.G., Vieira, W.L.S., Almeida,
W.O., Souto, W.M.S., Montenegro, P.F.G.P. and Pezzuti, J.C.B. (2012).
‘A review on human attitudes towards reptiles in Brazil’. Environmental
Monitoring and Assessment 184: 6877–6901.
Anderson, D.P., McMurtrie, P., Edge, K.A., Baxter, P.W.J. and Byrom,
A.E. (2016). ‘Inferential and forward projection modeling to evaluate
options for controlling invasive mammals on islands’. Ecological
Applications 26: 2546–2557.
Andrade, D.V, Sanders, C., Milsom, W.K. and Abe, A.S. (2004).
‘Overwintering in Tegu Lizards’. In: B.M. Barnes and H.V. Carey (eds.)
Life in the Cold: Evolution, Mechanisms, Adaptation and Application.
Twelfth International Hibernation Symposium, pp. 339–348. Fairbanks:
University of Alaska.
Andrade, L., Gomes, V. and Dias, M. (2009). ‘Desafios para o futuro
sustentável da Ilha de Fernando de Noronha: A visão ecossistêmica da
ocupação urbana’. Oculum Ensaios 9: 86–105.
Ballinger, R.E. (1977). ‘Reproductive strategies: Food availability as a
source of proximal variation in a lizard’. Ecology 58: 628–635.
Bellini, C. and Sanches, T.M. (1996). ‘Reproduction and feeding of
marine turtles in the Fernando de Noronha Archipelago, Brazil’. Marine
Turtle Newsletter 74: 12–13.
Borchers, D.L. and Efford, M.G. (2008). ‘Spatially explicit maximum
likelihood methods for capture-recapture studies’. Biometrics 64: 377–
385.
Bovendorp, R.S., Alvarez, A.D. and Galetti, M. (2008). ‘Density of
the tegu lizard (Tupinambis merianae) and its role as nest predator at
Anchieta Island, Brazil’. Neotropical Biology and Conservation 3:
9–12.
Branner, J.C. (1888). ‘Notes on the fauna of the islands of Fernando de
Noronha’. The American Naturalist 22: 861–871.
Brasil. (2004). ‘Plano de manejo da APA de Fernando de Noronha - Rocas
- São Pedro e São Paulo’. Mma/Ibama/Arcadis/Tetraplan S.a 1206.
Bringsøe, H. (2008). ‘Trachylepis sechellensis (Seychelles skink):
Reproduction’. Herpetological Bulletin 103: 42–43.
Brito, S.P., Abe, A.S. and Andrade, D.V. (2001). ‘Tupinambis merianae
(tegu lizard) longevity’. Herpetological Review 32: 260–261.
Burnham, K.P., Anderson, D.R. and Laake, J.L. (1980). ‘Estimation of
density from line transect sampling of biological populations’. Wildlife
Monographs 72: 938–942.
Carter, A. J., Heinsohn, R., Goldizen, A. W., and Biro, P. A. (2012).
‘Boldness, trappability and sampling bias in wild lizards’. Animal
Behaviour 83(4): 1051–1058.
324
Chapin, F.S., Zavaleta, E.S., Eviner, V.T., Naylor, R.L., Vitousek,
P.M., Reynolds, H.L., Hooper, D.U., Lavorel, S., Sala, O.E., Hobbie,
S.E., Mack, M.C. and Díaz, S. (2000). ‘Consequences of changing
biodiversity’. Nature 405: 234–242.
Chiarello, A.G., Srbek-Araujo, A.C., Del-Duque Jr., H.J., Coelho, E.D.R.
and Rocha, C.F.D. (2010). ‘Abundance of tegu lizards (Tupinambis
merianae) in a remnant of the Brazilian Atlantic forest’. Amphibia
Reptilia 31: 563–570.
da Silva, E.A.P., Santos, T.D., Leite, G.N. and Ribeiro, L.B. (2013).
‘Tropidurus hispidus (Squamata: Tropiduridae) and Leptodactylus cf.
fuscus (Anura: Leptodactylidae) as prey of the teiid lizards Salvator
merianae and Ameiva ameiva’. Herpetology Notes 6: 51–53.
de Souza, S.C.R., de Carvalho, J.E., Abe, A.S., Bicudo, J.E.P.W. and
Bianconcini, M.S.C. (2004). ‘Seasonal metabolic depression, substrate
utilisation and changes in scaling patterns during the first year cycle of
tegu lizards (Tupinambis merianae)’. Journal of Experimental Biology
207: 307–318.
Dias, R.A., Abrahão, C.R., Micheletti, T., Mangini, P.R., de Oliveira
Gasparotto, V.P., de Jesus Pena, H.F., Ferreira, F., Russell, J.C. and
Silva, J.C.R. (2017). ‘Prospects for domestic and feral cat management
on an inhabited tropical island’. Biological Invasions 19: 2339–2553.
Diaz, J.A. and Carrascal, L.M. (1991). ‘Regional distribution of a
Mediterranean lizard – influence of habitat cues and prey abundance’.
Journal of Biogeography 18: 291–297.
Doherty, T.S., Glen, A.S., Nimmo, D.G., Ritchie, E.G. and Dickman, C.R.
(2016). ‘Invasive predators and global biodiversity loss’. Proceedings
of the National Academy of Sciences 113: 11261–11265.
Donadio, O.E. and Gallardo, J.M. (1984). ‘Biología y conservación de
las especies del género Tupinambis (Squamata, Sauria, Teiidae) en
la República Argentina’. Revista del Museo Argentino de Ciencias
Naturales “Bernardino Rivadavia” Zoologia 13: 117–127.
Dorcas, M.E., Willson, J.D., Reed, R.N., Snow, R.W., Rochford, M.R.,
Miller, M.A., Meshaka, W.E., Andreadis, P.T., Mazzotti, F.J., Romagosa,
C.M. and Hart, K.M. (2012). ‘Severe mammal declines coincide with
proliferation of invasive Burmese pythons in Everglades National Park’.
Proceedings of the National Academy of Sciences 109: 2418–2422.
e Silva, R. S. and Neves, F. O. C. (2008). Aves de Fernando de Noronha:
Birds of Fernando de Noronha. Vinhedo: Avis Brasilis.
Eales, J., Thorpe, R.S. and Malhotra, A. (2010). ‘Colonization history and
genetic diversity: adaptive potential in early stage invasions’. Molecular
Ecology 19: 2858–2869.
Ferreira, F.S., Brito, S. V., Saraiva, R.A., Araruna, M.K.A., Menezes,
I.R.A., Costa, J.G.M., Coutinho, H.D.M., Almeida, W.O. and Alves,
R.R.N. (2010). ‘Topical anti-inflammatory activity of body fat from
the lizard Tupinambis merianae’. Journal of Ethnopharmacology 130:
514–520.
Ferreira, R.B., Beard, K.H., Peterson, S.L., Poessel, S.A. and Callahan,
C.M. (2012). ‘Establishment of introduced reptiles increases with the
presence and richness of native congeners’. Amphibia-Reptilia 33:
387–392.
Fitzgerald, L.A., Chani, J.M. and Donadio, O.E. (1991). ‘Tupinambis
Lizards in Argentina: Implementing Management of a Traditionally
Exploited Resource’. In: J. Robinson and K. Redford (eds.) Neotropical
Wildlife: Use and Conservation, pp. 303–316. Chicago: University of
Chicago.
Fitzgerald, L.A., Cruz, F.B. and Perotti, G. (1993). ‘Reproductive cycle
and the size at maturity of Tupinambis rufescens (Sauria : Teiidae) in the
dry chaco of Argentina’. Journal of Herpetology 27: 70–78.
Fletcher, C. S., Westcott, D. A., Murphy, H. T., Grice, A. C., and Clarkson,
J. R. (2015). ‘Managing breaches of containment and eradication of
invasive plant populations’. Journal of Applied Ecology 52(1): 59–68.
Fraga, R. De, Magnusson, W.E., Abrahão, C.R., Sanaiotti, T. and Lima,
A.P. (2013). ‘Habitat Selection by Bothrops atrox (Serpentes: Viperidae)
in Central Amazonia, Brazil’. Copeia 2013: 684–690.
Gasparini, J.L., Peloso, P.L. and Sazima, I. (2007). ‘New opportunities
and hazards brought by humans to the island habitat of the skink
Euprepis atlanticus’. Herpetological Bulletin 100: 30–33.
Halliday, T.R. and Verrell, P.A. (1988). ‘Body size and age in amphibians
and reptiles’. Journal of Herpetology 22: 253–265.
Hays, W.S.T. and Conant, S. (2007). ‘Biology and impacts of Pacific
island invasive species. 1. A worldwide review of effects of the small
Indian mongoose, Herpestes javanicus (Carnivora: Herpestidae)’.
Pacific Science 61: 3–16.
Herrel, A., Joachim, R., Vanhooydonck, B. and Irschick, D.J. (2006).
‘Ecological consequences of ontogenetic changes in head shape and
bite performance in the Jamaican lizard Anolis lineatopus’. Biological
Journal of the Linnean Society 89: 443–454.
Holmes, N.D., Campbell, K.J., Keitt, B.S., Griffiths, R., Beek, J., Donlan,
C.J. and Broome, K.G. (2015). ‘Reporting costs for invasive vertebrate
eradications’. Biological Invasions 17: 2913–2925.
Howald, G., Donlan, C.J., Galván, J.P., Russell, J.C., Parkes, J.,
Samaniego, A., Wang, Y., Veitch, D., Genovesi, P., Pascal, M., Saunders,
A. and Tershy, B. (2007). ‘Invasive rodent eradication on islands’.
Conservation Biology 21: 1258–1268.
IBGE (2017). Demografia de Fernando de Noronha [WWW Document]
Instituto Brasileiro de Geografia e Estatística. <http://cod.ibge.gov.
br/68D>.
Keitt, B., Griffiths, R., Boudjelas, S., Broome, K., Cranwell, S., Millett,
J., Pitt, W. and Samaniego-Herrera, A. (2015). ‘Best practice guidelines
for rat eradication on tropical islands’. Biological Conservation 185:
17–26.
Abrahão, et al.: Population assessment of tegu, Fernando de Noronha
Kiefer, M.C. and Sazima, I. (2002). ‘Diet of juvenile tegu lizard
Tupinambis merianae (Teiidae) in southeastern Brazil’. AmphibiaReptilia 23: 105–108.
Klug, P.E., Reed, R.N., Mazzotti, F.J., McEachern, M.A., Vinci, J.J.,
Craven, K.K. and Yackel Adams, A.A. (2015). ‘The influence of
disturbed habitat on the spatial ecology of Argentine black and white
tegu (Tupinambis merianae), a recent invader in the Everglades
ecosystem (Florida, USA)’. Biological Invasions 17: 1785–1797.
Lanfri, S., Di Cola, V., Naretto, S., Chiaraviglio, M. and Cardozo,
G. (2013). ‘Understanding the ecological niche to elucidate spatial
strategies of the southernmost Tupinambis lizards’. Amphibia-Reptilia
34: 551–565.
Le Galliard, J.F., Fitze, P.S., Ferrière, R. and Clobert, J. (2005). ‘Sex ratio
bias, male aggression, and population collapse in lizards’. Proceedings
of the National Academy of Sciences 102: 18231–18236.
Lirio, M.R., Araujo, A.F.B., Surita, L. and Péres-Jr., A.K. (2004).
‘Monitoramento do Lagarto Tupinambis merianae em Fernando de
Noronha Através de Rádiotelemetria’. In: Congresso Brasileiro de
Zoologia – Reptilia –anais. Brasília, p. 413.
Lopes, H.R. and Abe, A.S. (1999). ‘Biologia Reprodutiva e
Comportamento do Teiú, Tupinambis merianae (Reptilia, Teiidae), em
Cativero’. In: Manejo Y Conservación de Fauna Silvestre En América
Latina, pp. 259–272. La Paz, Bolivia: Editorial Instituto de Ecología.
Manes, M.E., Parry, H.V. and Manlla, A., (2007). ‘Growth and feeding
conversion responses to different protein and energy supplies in
Tupinambis merianae (Squamata: Teiidae)’. Applied Herpetology 4:
279–285.
Marinho, A.C. (2016). Qual a população de Fernando de Noronha? Portal
G1. <http://g1.globo.com/pernambuco/blog/viver-noronha/post/qualpopulacao-de-fernando-de-noronha.html>.
Mazzotti, F.J., McEachern, M., Rochford, M., Reed, R.N., Eckles, J.K.,
Vinci, J., Edwards, J. and Wasilewski, J. (2015). ‘Tupinambis merianae
as nest predators of crocodilians and turtles in Florida, USA’. Biological
Invasions 17: 47–50.
McCreless, E.E., Huff, D.D., Croll, D.A., Tershy, B.R., Holmes, N.D.,
Spatz, D.R., Butchart, S.H.M. and Wilcox, C. (2016). ‘Past and
predicted future impact of invasive alien mammals on island threatened
vertebrate populations’. Nature Communications 7: 1–11.
McDiarmid, R.W., Foster, M.S., Guyer, C., Whithfield Gibbons, J.,
Chernoff, N. and Dodd Jr, C.K. (2012). Reptile Biodiversity: Standard
Methods for Inventory and Monitoring. Los Angeles: Copeia.
Mello, T.J. and Adalardo de Oliveira, A. (2016). ‘Making a bad situation
worse: An invasive species altering the balance of interactions between
local species’. PLOS ONE 11: 1–17.
Mendonça, L.E.T., Souto, C.M., Andrelino, L.L., Souto, W.M.S., Vieira,
W.L.S. and Alves, R.R.N. (2011). ‘Conflitos entre pessoas e animais
silvestres no semiárido paraibano e suas implicações para conservação’.
Sitientibus série Ciências Biológicas 11: 185–199.
Mieres, M.M. and Fitzgerald, L.A. (2006). ‘Monitoring and managing the
harvest of tegu lizards in Paraguay’. Journal of Wildlife Management
70: 1723–1734.
Mourthé, Í.M.C. (2010). ‘Tupinambis merianae (common tegu) feeding
behavior’. Herpetological Review 41: 232–233.
Muscat, E., Olmos, F. and Rotenberg, E. (2016). ‘Salvator merianae (black
and white tegu) scavenging around the nests of Fregata magnificens
(frigatebird)’. Herpetological Bulletin 135: 36–37.
Neves, V.C., Nava, C., Monteiro, E. V, Monteiro, P.R. and Bried, J. (2017).
‘Depredation of Monteiro’s storm-petrel (Hydrobates monteiroi) chicks
by Madeiran wall lizards (Lacerta dugesii)’. Waterbirds 40: 82–86.
Nóbrega Alves, R., Pereira Filho, G., Silva Vieira, K., Silva Souto, W.,
Mendonça, L.E., Montenegro, P., Almeida, W. de and Silva Vieira, W.
(2012). ‘A zoological catalogue of hunted reptiles in the semiarid region
of Brazil’. Journal of Ethnobiology and Ethnomedicine 8: 27.
Novosolov, M., Rodda, G.H., Feldman, A., Kadison, A.E., Dor, R. and
Meiri, S. (2016). ‘Power in numbers. Drivers of high population density
in insular lizards’. Global Ecology and Biogeography 25: 87–95.
Oren, D.C. (1984). ‘Resultados de uma nova expedição zoológica a
Fernando de Noronha’. Boletim do Museu Paranaense Emilio Goeldi
Zoologia 1: 19–44.
Passos, D.C., Lima-Araujo, F., Melo, A.C.B. and Borges-Nojosa, D.M.
(2013). ‘New state record and distribution extension of the golden
tegu Tupinambis teguixin (Linnaeus, 1758) (Squamata: Teiidae) to the
Caatinga biome, northeastern Brazil’. Check List 9: 1524–1526.
Pekelharing, C.J., Parkes, J.P. and Barker, R.J. (1998). ‘Possum
(Trichosurus vulpecula) densities and the impacts on fuchsia (Fuchsia
excorticata) in South Westland, New Zealand’. New Zealand Journal of
Ecology 22: 197–203.
Pernambuco. (2017). Arquipélago de Fernando de Noronha, Informações
Turísticas. Site oficial do governo de Pernambuco. <http://www.
prontotecnologia.com.br/noronha2/turInfo.php>.
Pernas, T., Giardina, D.J., McKinley, A., Parns, A. and Mazzotti, F.J.
(2012). ‘First observations of nesting by the Argentine black and white
tegu, Tupinambis merianae, in South Florida’. Southeastern Naturalist
11: 765–770.
Phillips, B.L., Brown, G.P., Grennlees, M., Webb, J.P. and Shine, R.
(2007). ‘Rapid expansion of the cane toad (Bufo marinus) invasion front
in tropical Australia’. Austral Ecology 32: 169–176.
Powell, R., Henderson, R.W., H., Farmer, M.C., Breuil, M., Echternacht,
A.C., van Buurt, G., Romagosa, C.M. and Perry G. (2011). ‘Introduced
Amphibians and Reptiles in the Greater Caribbean: Patterns and
Conservation Implications’. In: A. Hailey, B.S. Wilson and J. Horrocks
(eds.) Conservation of Caribbean Island Herpetofaunas Volume 1:
Conservation Biology and the Wider Caribbean, pp. 63–143. Leiden
and Boston: Brill.
Presch, W. (1973). ‘A Review of the tegus, lizard genus Tupinambis
(Sauria : Teiidae) from South America’. Copeia 1973: 740–746.
Ramalho, A.C.O., da Silva, R.J., Schwartz, H.O. and Péres-Jr, A.K.
(2009). ‘Helminths from an introduced species (Tupinambis merianae),
and two endemic species (Trachylepis atlantica and Amphisbaena
ridleyi) from Fernando de Noronha Archipelago, Brazil.’. The Journal
of Parasitology 95: 1026–1028.
Reaser, J.K., Meyerson, L.A., Cronk, Q., De Poorter, M., Eldrege,
L.G., Green, E., Kairo, M., Latasi, P., Mack, R.N., Mauremootoo, J.,
O’Dowd, D., Orapa, W., Sastroutomo, S., Saunders, A., Shine, C.,
Thrainsson, S. and Vaiutu, L. (2007). ‘Ecological and socioeconomic
impacts of invasive alien species in island ecosystems’. Environmental
Conservation 34: 98–111.
Reis, A. and Hayward, P. (2013). ‘Pronounced particularity: A comparison
of governance structures on Lord Howe Island and Fernando de
Noronha’. Island Studies Journal 8: 285–298.
Ridley, H.N. (1890). ‘Notes on the zoology of Fernando de Noronha’.
Linnean Journal of Zoology 20: 473–570.
Russell, J.C., Meyer, J.-Y., Holmes, N. and Pagad, S. (2017). ‘Invasive
alien species on islands: Impacts, distribution, interactions and
management’. Environmental Conservation 44: 1–13.
Sampaio, A.B. and Schmidt, I.B. (2014). ‘Espécies exóticas invasoras em
unidades de conservação federais do Brasil’. Biodiversidade Brasileira
2: 32–49.
Santos, A.T. dos (1950). Fernando de Noronha, Ilha de Cenários
Múltiplos, 1st ed. Rio de Janeiro, RIO.
Sazima, I. and Haddad, C.F.B. (1992). ‘Répteis da Serra do Japi: Notas
sobre História Natural’. In: L.P.C. Morellato (ed.) História Natural
da Serra do Japi: Ecologia E Preservação de Uma Área Florestal do
Sudeste do Brasil, pp. 212–236. Campinas: Unicamp e FAPESP.
Sazima, I. and Haemig, P.D.P. (2012). ‘Aves, mamíferos e répteis de
Fernando de Noronha’. Ecologia.Info 17. < http://ecologia.info/
fernando-de-noronha.htm>.
Shine, R. (1987). ‘Ecological comparisons of island and mainland
populations of Australian tigersnakes (Notechis: Elapidae)’.
Herpetologica 43: 233–240.
Silva-Jr, J. M., Pérez-Jr, A. K., and Sazima, I. (2005). ‘Euprepis atlanticus
(noronha skink). Predation’. Herpetological review 36: 62–63.
Simberloff, D. (2009). ‘Rats are not the only introduced rodents producing
ecosystem impacts on islands’. Biological Invasions 11: 1735–1742.
Simberloff, D.S. (1974). ‘Equilibrium theory of island biogeography and
ecology’. Annual Review of Ecology and Systematics 5: 161–182.
Siqueira, C.D.C. and Rocha, C.F.D. (2008). ‘Predation by lizards as a
mortality source for juvenile lizards in Brazil’. South American Journal
of Herpetology 3: 82–87.
Tattersall, G.J., Leite, C.A.C., Sanders, C.E., Cadena, V., Andrade, D.V,
Abe, A.S. and Milsom, W.K. (2016). ‘Seasonal reproductive endothermy
in tegu lizards’. Science Advances 2: 1–8.
Team, R.C. (2000). ‘R language definition’. Vienna, Austria: R foundation
for statistical computing.
Tershy, B.R., Shen, K.W., Newton, K.M., Holmes, N.D. and Croll, D.A.
(2015). ‘The importance of islands for the protection of biological and
linguistic diversity’. BioScience 65: 592–597.
UNEP-WCMC. (2014). Checklist of CITES species [WWW Document].
CITES Secretariat, United Nations Environmental Programme. <http://
checklist.cites.org>.
Vanzolini, P., Ramos-costa, A.M.M. and Vitt, L.J. (1980). ‘Répteis das
caatingas’, 1st ed., FNDCT-FINEP. Rio de Janeiro, Academia Brasileira
de Ciências.
Veitch, C.R. and Clout, M.N. (eds.) 2002. Turning the tide: the eradication
of invasive species. Occasional Paper SSC no. 28. IUCN SSC Invasive
Species Specialist Group. IUCN, Gland, Switzerland and Cambridge,
UK.
Vitousek, P.M. (1988). ‘Diversity and Biological Invasions of Oceanic
Islands’. In: E.O. Wilson, F.M. Peter (eds.) Biodiversity, pp. 181–189.
Washington, DC: National Academy Press.
Vitousek, P.M., D’Antonio, C.M., Loope, L.L., Rejmánek, M. and
Westbrooks, R. (1997). ‘Introduced species: A significant component
of human-caused global change’. New Zealand Journal of Ecology 21:
1–16.
Winck, G.R. and Cechin, S.Z. (2008). ‘Hibernation and emergence pattern
of Tupinambis merianae (Squamata: Teiidae) in the Taim Ecological
Station, southern Brazil’. Journal of Natural History 42: 239–247.
Winck, G.R., Blanco, C.C. and Cechin, S.Z. (2011). ‘Population ecology
of Tupinambis merianae (Squamata, Teiidae): home-range, activity and
space use’. Animal Biology 61: 493–510.
Wyatt, K.B., Campos, P.F., Gilbert, M.T.P., Kolokotronis, S.O., Hynes,
W.H., DeSalle, R., Daszak, P., MacPhee, R.D.E. and Greenwood, A.D.
(2008). ‘Historical mammal extinction on Christmas Island (Indian
Ocean) correlates with introduced infectious disease’. PLOS ONE 3:
e3602.
Yanosky, A.A., Iriart, D.E. and Mercolli, C. (1993). ‘Predatory behavior in
Tupinambis teguixin (Sauria: Teiidae). Island Tongue-flicking responses
to chemical food stimuli’. Journal of Chemical Ecology 19: 291–299.
325
S.R. Fisher, R.N. Fisher, S. Alcaraz, R. Gallo-Barneto, C. Patino-Martinez, L.F. López Jurado and C.J. Rochester
Fisher, S.R.; R.N. Fisher, S. Alcaraz, R. Gallo-Barneto, C. Patino-Martinez, L.F. López Jurado and C.J. Rochester. Lifehistory comparisons between the native range and an invasive island population of a colubrid snake
Life-history comparisons between the native range and an invasive
island population of a colubrid snake
S.R. Fisher1, R.N. Fisher2, S.E. Alcaraz3, R. Gallo-Barneto4, C. Patino-Martinez5, L.F. López Jurado6
and C.J. Rochester2
Southwestern College, Chula Vista, CA, USA. <samrfisher7@gmail.com>. 2U.S. Geological Survey, Western
Ecological Research Center, San Diego, CA, USA. 3Department of Biology, California Polytechnic State University,
San Luis Obispo, CA, USA. 4Área de Medio Ambiente, Gestión y Planeamiento Territorial y Ambiental (GesPlan S.A.),
Las Palmas, Spain. 5Asociación para el Desarrollo Sostenible y Conservación de la Biodiversidad (ADS), Las Palmas,
Spain. 6Universidad de Las Palmas de Gran Canaria, Las Palmas, Spain.
1
Abstract Invasive snakes can lead to the rapid extinction of endemic vertebrates on insular ecosystems, usually because
snakes are an efficient and novel predator. There have been no successful (i.e. complete) eradications to date of invasive
snakes on islands. In this study we assess a novel invasion on Gran Canaria in the Canary Islands. The invader, the
California king snake (Lampropeltis californiae), arrived from California via several generations in the pet trade. King
snakes are captive bred for various phenotypes, and first were detected in the wild on Gran Canaria in the 1990s. Because
very little natural history data exist from within their native range, we focused on developing datasets from native habitats
to compare with similar data for the introduced snakes in the Canary Islands. We found that most aspects of the snake’s
life history have not changed since invasion, except that there appears to be a lower level of juvenile recruitment along
with an increase in the length and body mass of adult snakes on Gran Canaria. We identified environmental parameters
for when capture/trapping could be completed to reduce effort and maximize success. Additionally, we show different
trap success on the various life stages of the snakes. Risk assessments could be required prior to permitting pet trade or
allowing captive bred snakes into regions where they are not native.
Keywords: California, Canary Islands, detection, Lampropeltis californiae, morphology, pet trade
INTRODUCTION
Invasive species on islands drive high levels of
extinction globally (Jones, et al., 2016). No examples of
eradications of invasive snakes are known from islands
(DIISE, 2015). Unlike mammals, where successful
methods of eradication exist and great conservation
success has been achieved (Jones, et al., 2016), snakes
continue to invade cryptically, often with dramatic impacts
(Willson, 2017). The accidental introduction of the brown
tree snake (Boiga irregularis) to Guam has led to the loss
of almost all the bird and much of the lizard diversity of the
island (Rodda & Savidge, 2007). When this invasion was
recognised, major changes in the biodiversity of the island
had already taken place (Savidge, 1987; Rodda & Fritts,
1992). The brown tree snake, like several other snake
invaders, is poorly known biologically in its native range,
and thus any biological changes to the invader during the
invasion cannot be easily detected (Rodda & Savidge,
2007).
One of the main pathways for introductions of reptiles
is the pet trade, which is linked to many invasive species
issues globally (Krysko, et al., 2016; McFadden, et al.,
2017). Little is known about the effects of having captive
raised snakes released into the wild. In addition, there is
little information regarding the biology (morphology,
reproduction, behaviour, etc.) of non-native snakes when
they are introduced to islands. California king snakes
(Lampropeltis californiae; CKS) were originally caught
and bred for the pet trade, and many are from San Diego
County, California. The CKS has been a major element of
the international pet trade since the 1980s (Hubbs, 2009).
They have been artificially selected for certain coloration
and pattern phenotypes in captivity, including albino,
striped, and banded. They were originally imported to the
Canary Islands as well as many other places to be bred in
captivity and sold as pets. They were released accidentally
or escaped into the wild and have subsequently been on
the Canary Islands as an invasive species since the late
1990s, adversely affecting the native wildlife and currently
occurring in two discrete populations (Cabrera- Pérez,
et al., 2012; Monzón-Arguello, et al., 2015). There have
been perceived morphological changes in the snakes, and
their expansion could be exponential as they irrupt without
competition or predation (Cabrera- Pérez, et al., 2012).
When trying to compare the invasive snakes with those in
their natural habitat, we found that there is little known of
the life history of CKS from their native range, especially
southern California, and most references cite only the
regional field guides, without much primary literature
to support this information. Recently for the first time,
movement data, which is very useful for understanding
the invasion process, has been published for this species
(Anguiano & Diffendorfer, 2015).
The Canary Islands are isolated oceanic islands off
the coast of West Africa. They have low biodiversity, but
high endemism, with some species that have important
adaptations (Rando, et al., 2008; Fernandez-Palacios, et
al., 2011). These include endemic lizards, of which the
lacertids (Gallotia spp.) are herbivorous and are important
seed dispersers (Valido, et al., 2003). The islands contain
no native species of snake. On the Canary Islands, the
invasive CKS have become a major predator for all of
the native lizard species and are therefore threatening
this island’s biodiversity (Cabrera- Pérez, et al., 2012;
Monzón-Arguello, et al., 2015). As with other invasive
species, CKS on the Canaries have gone after the most
abundant prey first, so they have been preying on the native
lizards primarily and then secondarily on invasive small
mammals. Birds do not make up a large part of their diet
yet (Cabrera- Pérez, et al., 2012), but there are endangered
birds present that might become snake prey over time as
other prey become exhausted (Carrascal, et al., 2017). In
addition, there are limited control efforts over the spread
of the snakes on the Canary Islands and potentially all of
Macaronesia (Azores, Madiera, and Cape Verde Islands).
This could potentially threaten the biodiversity of the
entire area if they are not eradicated. The snakes appear to
have no predators in the Canary Islands.
How snakes invade and the dynamics of the early
invasion process, in particular the changes to their
phenology, phenotype, and reproduction during the
irruption phase, have not been previously studied. Most
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
326
up to meet the challenge, pp. 326–331. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Fisher, et al.: Comparing natural and invasive snake populations
snake invasions are more mature before study. The Canary
Islands offer a unique opportunity to study these issues
as it is a novel environment for snakes, and the snake
invading is a species from the mainland of North America
where numerous museum specimens and other field data
are available. Because CKS are relatively well known,
developing detailed life history parameters should be more
straightforward than for other poorly known tropical species
of snakes, such as brown tree snakes or Burmese pythons
(Python bivittatus). The CKS is widespread from southern
Oregon, south to the tip of the Baja Peninsula in Mexico,
and east to mid-Nevada, southern Utah and the majority
of Arizona; throughout its range it occurs naturally with
many other snake species. The goal of this paper is to use
museum and field datasets to resolve critical life history
traits for this species, which can help to interpret CKS
invasion dynamics within the Canary Islands and may be
useful for optimising eradication/control techniques and
efforts (i.e. trapping timing and placement).
MATERIALS AND METHODS
To document potential biological changes in the snake’s
natural history during the invasion process, we sampled
CKS in their native range across 22.8ᵒN to 40ᵒN and made
comparisons with the invasive snakes. Most samples were
from southern California. Data were collected from 1,538
museum specimens (California Academy of Sciences,
Natural History Museum of Los Angeles, San Diego
Natural History Museum, University of California, Santa
Barbara Cheadle Center for Biodiversity & Ecological
Restoration) and augmented with records from wild
caught CKS delivered to the San Diego Zoo (electronic
supplementary materials). Additionally, we used southern
California field data from 778 CKS captured between
1995 and 2012 in pit-fall and snake trap arrays by USGS
(methods from Fisher, et al., 2008; electronic supplementary
materials). These data from southern California included
all snake species caught in these traps (n=4,708) and were
used to assess the capture rate ranking of the CKS species
compared to the other 24 native snake species for which
we had contemporary capture data from these traps. We
also obtained two different field datasets from the native
range for CKS. One was a citizen science dataset from
HerpMapper (HerpMapper, 2017) which had 1,299 records
for the snakes from which we used capture/detection dates.
The second was an unpublished dataset from Brian Hinds
(BH) which represented 717 detections with associated
observation dates. We compared these four native-range
datasets to the Canary Island dataset, which encompassed
668 snakes (hand and trap caught from 2012 to 2014) on
Gran Canaria Island (28ᵒN), all from the western of the two
populations on the island.
The museum specimens of CKS were measured for
snout-vent length (SVL) and tail length using measuring
tapes. Adults were defined as >600 mm in SVL (Hubbs,
2009). Sex was determined either through dissection or
tail length and width. Some snakes were found dead on
road (DOR) and the sex could not be determined. Many
of the older museum specimens were missing reproductive
systems; therefore, only a subset of data was available from
these. Specimens missing their organs were used for length
comparisons, but not for sex or reproductive status. Dorsal
patterning and evidence of tail breaks were recorded and
tail breaks were documented photographically.
The pit-fall and snake trap samples were collected
from the wild in the native range in southern California
primarily from south of Los Angeles to the Mexican border.
Individuals were sexed, weighed, measured, and released.
Data for colour pattern and tail status were lacking for most
specimens. We also analysed the total capture for all snake
species from these traps to look at the relative capture
success of CKS compared to all other snakes for which we
had data in the native snake community in California. To
further look at activity phenology within their native range,
we used data from HerpMapper (2017) and BH to assess
observations by month as a recent sample to compare against
our older native range data sources. Many of these records
are from active searches under artificial cover (AC), and
others are from night driving. Both of these are techniques
that might have high seasonal biases in detections. This is
because snakes under AC could be non-active, but using
the cover to environmentally thermoregulate; whereas
snakes detected on roads at night would be animals that
are actively moving. These behaviours would change
seasonally based on climatic conditions.
Samples from Gran Canaria Island were collected
by hand or by trap then euthanized and frozen for later
dissection. Sex, weight, SVL, tail breaks or scarring, were
recorded.
Comparisons were made among these five study
population samples for the relevant metrics and controlled
for differences in sampling types. For example, the
museum series is similar to the invasive population in that
animals were collected by hand, trap, or opportunistically,
but no comparison of weight could be completed, as the
preserved weight of the museum snakes is not comparable
to live weight. In contrast, live weight and length of the pitfall and snake trap series could be compared to the invasive
series, but reproductive states could not be compared, as
these data were not available for the trapped and released
snakes from their native range. These trap records are
from snakes that are actively foraging, as they have to be
moving in the landscape to encounter a trap. The last two
field data sets (HerpMapper and BH) could only be used
for detection/capture date comparisons with the other data
sets, as they involved primarily active searches, especially
under artificial cover, and not necessarily surface-active
snakes. They also lacked length/weight measurements
for individual snakes. We used means of the top decile to
highlight comparisons between populations.
RESULTS
Snake community structure in California
Within a community of 25 native snake species captured
via pit-fall and snake trap arrays in southern California,
CKS was found to be the second most abundant species
following the California whipsnake (Masticophis lateralis)
and represented approximately 17% of the 4,708 captures
across these species (Fig. 1). Snakes in this dataset were
captured when snakes entered traps; no active searching
for snakes took place. Thus, these records would be
biased towards species more frequently moving over the
landscape. These data indicate that within its native range
the CKS is one of the most abundant snakes captured with
this technique.
Trap success by size class
Using the USGS pit-fall and snake trap dataset, we were
able to look at the effect of trap type on capture success by
snake length, as a proxy for age (Fig. 2). We found that pitfall trap buckets (18.9 L) buried in the ground were most
successful, capturing snakes less than 500 mm in length.
Wire-mesh snake traps had the greatest success with
snakes exceeding 500 mm in length. Additionally, there
was no trend in body size of CKS incidentally observed
while conducting sampling using these traps.
Snake detections by month
We plotted the monthly detections/captures across
five different datasets to assess variability across months.
327
Island invasives: scaling up to meet the challenge. Ch 2C Other taxa: Herpetofauna
Fig. 2 Body size of Lampropeltis californiae by trap type
for southern California. This figure has 50 mm breaks in
size groups and highlights the different capture success
of the two different trap types (pit-fall versus snake trap).
The museum and Canary Island datasets both had their
peaks in the month of May, and these were identified using
a variety of detection types, including active searching and
traps. Finally, the pit-fall and snake trap dataset, with its
passive traps for detections, had its peak in June. This last
dataset was the only method based solely on active snakes.
From August to January there was <10% per month of total
snake detections across all datasets and from November to
January there was <5% per month of total snake detections
(Fig. 3).
Sex ratios, body size, and tail injury comparisons
between California and the Canary Islands
Fig. 1 All snake captures in southern California from
the USGS pit-fall and snake trap study (n=4,708).
Lampropeltis californiae is the second most common
snake species captured.
Overall, monthly detections across datasets were highest
between March and June, with the various peaks being due
to variance in detection technique used. The citizen science
(HerpMapper) and BH datasets, where they were actively
searching for snakes, had peaks between March and April.
We were able to make more detailed comparisons
across three datasets, two from the native range (museum
and pit-fall/snake trap) and the Canary Islands (Table 1).
We found that there was a greater proportion of adults
captured in the Canary Islands compared to the native range
pit-fall/snake trap captures or museum specimens. There
was no difference between the two native populations in
the percent of juveniles, with about 49% of the samples
representing juveniles; in contrast, only 22% of the
invasive snakes were juveniles (Table 1). Thus, there were
2.3 times more juveniles detected in the native range than
in the Canary Islands regardless of dataset used (museum
or pit-fall/snake trap). For the pit-fall/snake trap and
Canary Island captures, we compared the frequency by 50
mm size classes to see where this juvenile/adult bias was
Table 1 Morphological comparisons between native and invasive populations of Lampropeltis californiae.
Differences between values of the invasive versus native populations were calculated as percentages to
illustrate variance from 100%. Values in parentheses in table are sample sizes for top deciles.
Gran Canaria
Island
Total
Total # adults (>600mm)
Percent non-adults
Mean SVL (top decile) (mm)
Largest SVL (mm)
Mean weight (top decile) (g)
Weight largest (g)
Tail break frequency
668
519
21.9
1,071.7
(n=52)
1,474
412.8
(n=52)
770.3
16.64
670 with measurements that could be used
except non-wild caught ~70 individuals
a
b
328
Southern
California field
780a
335
49.6
1,069.1
(n=33)
1,290
334.9
(n=28)
570
-
Museum
specimens
1,538b
769
48.4
1,032.2
(n=77)
1197
Difference
Gran Canaria Is.
vs California
0.44
1.00
1.14
-
1.23
6.72
1.35
2.48
Fisher, et al.: Comparing natural and invasive snake populations
DISCUSSION
Fig. 3 Monthly percent of total detections of Lampropeltis
californiae across the four datasets from the native
range and the dataset from the invasive snakes on Gran
Canaria.
the greatest. We found the native range had only one size
class (351–400 mm) occurring in greater than 10% of the
sample, whereas four consecutive size classes (701–900
mm) occur in greater than 10% each of the sample from
the Canary Island. Thus, our data from the native range had
a bimodal distribution between juvenile and adult captures
compared to the Canary Island data (Fig. 4).
The invasive group of CKS did not have greater mean
of the top decile compared to snakes in their native range
(Table 1). The longest snake in the Canary Islands was 1,474
mm, 14% longer than the longest snake in the California
sample (1,290 mm) and 21% longer than the next longest
snake in the Canary Islands (1,217 mm). The invasive
snakes had 23% greater average mass within the top decile
compared to the USGS pit-fall/snake trap captures (Table
1). The heaviest snake in the Canary Islands was 770 g,
35% greater than the heaviest snake within the California
sample (570 g).
One of the natural history traits we looked at was
the frequency of tail breaks or scarring, as a proxy for
predation risk. In the museum dataset, 6.7% of the snakes
had broken tails, whereas 16.6% of the CKS on the Canary
Islands had broken tails (2.48 times higher frequency of tail
breaks compared with snakes in the native range) (Table
1). There was no noticeable association between tail break
and colour pattern for either of these datasets.
Fig. 4 Comparison of Lampropeltis californiae by
percentage of size class between the USGS pit-fall
snake trap dataset and the Canary Islands dataset. This
graph highlights the lack of smaller size classes in the
Canary Islands and the greater frequency of the larger
size classes.
Since 2009, field work on control and eradication of
the invasive CKS in the Canary Islands has resulted in the
removal of over 4,500 snakes from the invaded habitats
(<www.lifelampropeltis.com>). There was one population
on Gran Canaria when the snakes were discovered, but now
there are at least three populations on the island, indicating
they are still spreading even with the control activities. We
were able to compare various life history traits for native
range CKS to the invasive range in the Canary Islands.
Overall, we compiled records for 4,404 CKS for various
aspects of their biology from the native range across four
different data sources. These data were compared to 668
records for snakes from the Canary Islands. Below, we
make comparisons on their biology and then suggestions
on how they might be controlled or managed as an invasive
species.
Snake community structure in California
We found that CKS was the second most captured
species across the 25 species detected by the USGS pitfall and snake trap sampling in California (Fig. 1). This
sampling is based on the species actively entering the
traps, and since the traps are passive, they only detect
snakes when the snakes are active. Klauber (1931), using
primarily road-riding for eight years (1923–1930), found
that CKS were the third most detected snake species in his
sample. They comprised 14% of the total record of 6,231
snakes across 24 species he detected for San Diego County,
following the gopher snake (Pituophis catenifer) and the
two-striped garter snake (Thamnophis hammondii). As our
data were collected 70 years after his, this difference could
represent actual changes in the abundance of the snakes
due to habitat shifts over time, but it most likely represents
the different sampling techniques. Both studies found CKS
to be in the top three most captured snakes in the region
across habitat types, indicating that even in a diverse snake
community, CKS is one of the dominant species. This
suggests that as an invasive species, it possibly could be
successful even in regions with native snake communities,
such as mainland Europe. Within the Canary Islands, it
appears to have the ability to broadly utilise the habitats
present on these islands.
Trap success by size class and lack of juvenile snakes
in the Canary Islands
It was a quite striking find that juveniles are not detected
in high numbers in the Canary Islands yet the snake is
clearly expanding its range every year. This is very difficult
to explain. The juvenile detection could be affected by
several factors, including trapping technique, foraging
distances and activity, growth rate, etc., but with the data
we have to date we cannot determine the source of this
issue. We know that sampling techniques to detect snakes
vary in their effectiveness. We found a distinctive pattern
of smaller snakes (<500 mm) being detected primarily
by bucket traps (Fig. 2). This indicated there was a size
bias in the sampling, with the buckets being necessary to
capture the smaller snakes (<500 mm) and the mesh wire
snake traps having greater success with the larger snakes
(>500 mm) (Fig. 2). In the Canary Islands bucket traps
are not being used (Cabrera- Pérez, et al., 2012; MonzónArguello, et al., 2015), and this could possibly explain
the lack of juveniles being collected in the invasive range
(Table 1). However, the museum specimens from animals
captured in the wild in California include juvenile snakes,
suggesting their absence could be due to something implicit
in the Canary Islands. It could be there is some increased
predation within the Canary Islands targeting juveniles, but
if that was the case, the population might not be expanding
as rapidly as it appears to be spreading.
329
Island invasives: scaling up to meet the challenge. Ch 2C Other taxa: Herpetofauna
It is likely there is a greater abundance of naïve prey in
the Canary Islands reducing the need for juvenile snakes to
move long distances to forage, thus limiting their exposure
for detection or as prey. Abundant food resources might
also increase their growth rate so that detecting individuals
while they are still juveniles would be more difficult.
When prey presence in captured snakes was evaluated
for 270 individuals in the Canary Islands, 36% of these
snakes had at least one prey item in their digestive tract
(Monzón-Arguello, et al., 2015). In contrast, within their
native range, a recently published study found only about
8% of the snakes assessed contained prey items in their
digestive tract (Wiseman, et al., 2019). This suggests that
the invasive snakes are finding prey at four times the rate
of snakes within their native range, which could be a proxy
for increased prey abundance in the Canary Islands.
Another possible explanation for lower detection rates
of juveniles might be their activity levels compared to
adults. Juveniles might only be active when foraging and
under cover items between foraging bouts, while adults are
active while foraging and also when searching for mates
for reproduction, thus even though foraging exposure
might be reduced for adults in the Canary Islands, they
are still exposed for capture during mating season. Overall
this could result in the lower detection of juveniles in the
invasive range versus the native range, because the high
food availability which could lead to rapid growth rate in
the Canary Islands might limit detection probability (Pike,
et al., 2008).
Snake detections by month
The effectiveness of detection tools varied with the time
of year. Active searches under artificial cover (HerpMapper
and BH) were more effective early in the year (March and
April) before snakes were fully active as they used cover
to thermoregulate (Fig. 3). We found that pit-fall and snake
traps which are dependent on active snakes to enter the
traps were more effective in May and June. Overall, focused
field effort with various sampling techniques from March
to July would maximise the detection success for CKS
versus other months of the year. November, December,
and January had the lowest detection rates across all five
datasets, indicating that lowering field efforts during that
period of time would be justified.
Body size and tail injury comparisons between
California and the Canary Islands
We found no difference in mean SVL of the top decile
between snakes in the invasive range versus the native
range (Table 1). This result indicates that there has not been
a population shift to longer body size within the invasive
range, although the maximum length of the largest snake
in the Canary Islands was 14% longer than any California
snake, and 21% longer than the next largest snake in
the Canary Islands. This snake was an outlier, as it was
greater than three standard deviations longer than the next
longest snake in the Canary Islands. As this snake was the
second heaviest snake we don’t think this resulted from
measurement or recording error. This lack of population
shift in body size contrasts with what has been observed in
other invasive species, some of which have been shown to
grow larger within their invasive range (Rodda & Savidge,
2007), but this outlier snake indicates that this pattern
could change as the age since invasion gets longer. We did
find that the invasive snakes were 23% heavier for the top
decile, and the heaviest invasive snake was 35% larger than
the heaviest snake from the California trap study (Table 1).
Increased weight in invasive snakes is most likely tied to
their increased predation success on naïve prey.
We observed a higher percentage of tail breaks and
scarring of the snakes in the Canary Islands. This could
be due to incomplete predation from cats (Felis catus)
330
or other predators, from defensive wounds of their prey
(e.g. Gallotia stehlini), or possibly some other unknown
process (Medina & Nogales, 2009; Santos, et al., 2011).
Increased frequency of tail breaks does not necessarily
affect body condition, for some species (Pleguezuelos,
et al., 2013). Within the snakes’ native range, predators
may be more efficient resulting more often in complete
predation, especially by raptors, leaving fewer individuals
with incomplete predation scars.
Trophic cascades
A major concern with novel invasive species is that
their removal of highly specialised endemic species
with unique roles in the island ecosystems may result in
unexpected downstream changes in biodiversity and in the
landscape. The Canary Islands have a small but unique
and ancient biodiversity that could be highly susceptible to
perturbations from invasive species (Fernandez-Palacios,
et al., 2011). One example is the endemic Gallotia lizard
which is an essential part of the trophic cascade/feedback
loop that enables the dispersal of trees on the Canary
Islands (Valido, et al., 2003). The lizards eat the fruit off
the trees and shrubs, effectively spreading the seeds of
the endemic flora. The invasive CKS are consuming these
lizards at a high rate, with complete removal of juveniles
in areas where snakes have invaded, and over time will
impede the proliferation of these native trees and shrubs,
altering the biodiversity and native habitat (Cabrera- Pérez,
et al., 2012; Monzón-Arguello, et al., 2015). Published
examples of trophic cascades tied to snake invasions
include the relationship between spiders and birds in Guam
now caused by the snake irruption, and the dynamics of
python and mid-sized mammals in Florida (Rogers, et al.,
2012; Willson, 2017).
The CKS has a varied diet in its native range, including
venomous snakes and juvenile birds (Morrison & Bolger,
2002). Because there are currently no birds recorded in
the diet of the invasive snakes (Monzón-Arguello, et al.,
2015), initiating intensive sampling of birds in areas with
and without snakes to get an assessment of bird density and
recruitment may be valuable. From the literature it seems
clear that these snakes could target birds, many of which
are endemic and some are currently endangered, as prey
as they exhaust the lizards and rodents present (Morrison
& Bolger, 2002; Carrascal, et al., 2017). This may also be
valuable because the published diet data are five years old,
and there might already be a change in their diet if there is
a depletion of the main reptile and rodent prey.
If it looks like the snakes are going to achieve an islandwide distribution, then one approach is to pre-emptively
safeguard various biologically intact areas around the
island at different elevations. This approach could preserve
biodiversity and create reservoirs of native animals in the
event that the snake control/eradication fails.
Pet trade and captive breeding/selection and then
released into wild
The invasive CKS has a unique history as it came from
several generations of selection in captivity for various
colour morphs and albinism, in addition to rapid growth
and reproduction. Their release to the wild in the Canary
Islands is concerning as this selection might provide
some reproductive advantage versus the release of wild
animals not subjected to selection in captivity. This trade
of potential invasive species is concerning as more and
more reptile species become bred for sale globally in the
pet trade (Robinson, et al., 2015).
Considerations for snake management in the Canary
Islands
Looking at CKS published movement data suggests that
placing snake traps with sterile female snakes, or proxies,
Fisher, et al.: Comparing natural and invasive snake populations
less than every 150 m apart may be effective for snake
management. This distance may be appropriate because
the literature indicates that 98% of the males and 100%
of the females radio-tracked do not move farther than this
(Anguiano & Diffendorfer, 2015). Having a grid of traps
in closer proximity across the snake-occupied parts of the
island would be optimal for a snake removal programme.
There are large ecological and monetary costs to
invasive animals, and costs of control and/or eradication
often exceed the available funding. We suggest (1) stronger
controls on snakes in the pet trade, (2) rapid response to
prevent spread when detection first occurs, and (3) use of
citizen science as a tool to detect early invasions.
CONCLUSION
Our results show that data from the native range of the
snake can inform management and control for CKS within
their invasive range. Also, we found that they flourish
within a diverse native snake community; they have a high
natural abundance, both historically (Klauber, 1931) and
currently (Fig. 1).
We suggest that the continued use of a variety of traps in
addition to active surveys be used to maximise detection of
snakes of all sizes, especially within the months of March
through July. We also suggest that managers consider
protection of natural areas with critical biodiversity on Gran
Canaria from invasion by CKS. In addition, managers may
wish to consider increased controls to prevent spread to
other areas in the Canary Islands.
There is no literature on where the CKS lays its eggs in
its natural habitat or in the Canary Islands. A comparison
of this and other reproductive characteristics may be
important as well as a better understanding of how to detect
juveniles within the invasive range. Greater support for risk
assessments of species, within the pet trade in particular,
could help to identify species of greatest concern which
would help reduce these types of invasions elsewhere.
ACKNOWLEDGEMENTS
We thank the following museum collections and
curators for their support of our research on these snakes:
LACM (Greg Pauly, Nefty Camacho), UCSB CCBER
(Sam Sweet, Mireia Beas-Moix), California Academy
of Sciences (Jens Vindum), San Diego Natural History
Museum (Brad Hollingsworth, Laura Williams). United
States Geological Survey, San Diego Zoo (Kim Lovich),
and Brian Hinds provided unpublished records. Data were
provided by HerpMapper (<www.HerpMapper.org>) and
its network of citizen contributors. Stephen Goldberg
provided invaluable training and Tim Felton, Christine and
Herk Alcaraz, and Marie Fisher provided other support.
We thank Miguel Ángel Cabrera- Pérez, Miguel Ángel
Peña, José Miguel Sánchez, Alejandro Ramírez, Jorge
Fernando Saavedra, Francisco Alarcón for all the help with
the snakes and the research in the Canary Islands. Funding
was provided by the European Project LIFE10 NAT/
ES/000565 LAMPROPELTIS and the USGS Ecosystems
Mission Area. Any use of trade, firm, or product names
is for descriptive purposes only and does not imply
endorsement by the U.S. Government.
REFERENCES
Anguiano, M. and Diffendorfer, J. (2015). ‘Effects of fragmentation on the
spatial ecology of the California kingsnake (Lampropeltis californiae)’.
Journal of Herpetology 49: 420–427.
Cabrera-Pérez, M.Á., Gallo-Barneto, R., Esteve, I., Patiño-Martínez, C.
and López-Jurado, L.F. (2012). ‘The management and control of the
California kingsnake in Gran Canaria (Canary Islands): Project LIFE+
Lampropeltis’. Aliens: The Invasive Species Bulletin Newsletter of the
IUCN/SSC Invasive Species Specialist Group 32: 20–28.
Carrascal, L.M., Moreno, A.C., Delgado, A., Suarez, V. and Trujillo, D.
(2017). ‘Habitat suitability – density relationship in an endangered
woodland species: the case of the blue chaffinch (Fringilla polatzeki)’.
PeerJ 5: e3771.
DIISE. (2015). ‘The Database of Island Invasive Species Eradications,
developed by Island Conservation, Coastal Conservation Action
Laboratory UCSC, IUCN SSC Invasive Species Specialist Group,
University of Auckland and Landcare Research New Zealand’. <http://
diise.islandconservation.org>. Accessed 3 September 2017.
Fernandez-Palacios, J.M., de Nascimento, L., Otto, R., Delgado, J.D.,
Garcia-del-Rey, E., Arevalo, J.R. and Whittaker, R.J. (2011). ‘A
reconstruction of Palaeo-Macaronesia, with particular reference to the
long-term biogeography of the Atlantic island laurel forests’. Journal of
Biogeography 38: 226–246.
Fisher, R.N., Stokes, D., Rochester, C., Brehme, C., Hathaway, S. and
Case, T.J. (2008). Herpetological Monitoring Using a Pitfall Trapping
Design in Southern California. Reston, Virginia: US Geological Survey.
HerpMapper. (2017). ‘HerpMapper – A Global Herp Atlas and Data Hub.
Iowa, U.S.A’. <www.herpmapper.org>. Accessed 1 September 2017).
Hubbs, B. (2009). Common Kingsnakes: A Natural History of
Lampropeltis getula’. Tempe, AZ: Tricolor Books.
Jones, H.P., Holmes, N.D., Butchart, S.H., Tershy, B.R., Kappes, P.J.,
Corkery, I., Aguirre-Muñoz, A., Armstrong, D.P., Bonnaud, E., Burbidge,
A.A., Campbell, K., Courchamp, F., Cowan, P.E., Cuthbert, R.J.1.,
Ebbert, S., Genovesi, P., Howald, G.R., Keitt, B.S., Kress, S.W., Miskelly,
C.M., Oppel, S., Poncet, S., Rauzon, M.J., Rocamora, G., Russell, J.C.,
Samaniego-Herrera, A., Seddon, P.J., Spatz, D.R., Towns, D.R. and
Croll, D.A. (2016). ‘Invasive mammal eradication on islands results in
substantial conservation gains’. Proceedings of the National Academy
of Sciences 113: 4033–4038.
Klauber, L.M. (1931). ‘A statistical survey of the snakes of the southern
border of California’. Bulletin of the Zoological Society of San Diego
8: 1–93.
Krysko, K.L., Somma, L.A., Smith, D.C., Gillette, C.R., Cueva, D.,
Wasilewski, J.A., Enge, K.M., Johnson, S.A., Campbell, T.S., Edwards,
J.R., Rochford, M.R., Tompkins, R., Fobb, J.L., Mullin, S., Lechowicz,
C.J., Hazelton, D. and Warren, A. (2016). ‘New verified nonindigenous
amphibians and reptiles in Florida through 2015, with a summary of
over 152 years of introductions’. International Reptile Conservation
Foundation Reptiles and Amphibians 23: 110–143.
McFadden, M.S., Topham, P. and Harlow, P.S. (2017). ‘A ticking time
bomb: Is the illegal pet trade a pathway for the establishment of corn
snake (Elaphe guttata) populations in Australia?’ Australian Zoologist
38: 499–504.
Medina, F.M. and Nogales, M. (2009). ‘A review on the impacts of feral
cats (Felis silvestris catus) in the Canary Islands: Implications for the
conservation of its endangered fauna’. Biodiversity and Conservation
18: 829–846.
Monzón -Arguello, C., Patino-Martinez, C., Christiansen, F., GalloBarneto, R., Cabrera- Pérez, M.A., Pena-Estevez, M.A., Lopez-Jurado,
L.F. and Lee, P.L.M. (2015). ‘Snakes on an island: Independent
introductions have different potentials for invasion’. Conservation
Genetics 16: 1225–1241.
Morrison, S.A. and Bolger, D.T. (2002). ‘Lack of an urban edge effect
on reproduction in a fragmentation-sensitive sparrow’. Ecological
Applications 12: 398–411.
Pike, D.A., Pizzatto, L., Pike, B.A. and Shine, R. (2008). ‘Estimating
survival rates of uncatchable animals: The myth of high juvenile
mortality in reptiles’. Ecology 89: 607–611.
Pleguezuelos, J.M., Feriche, M. and Santos, X. (2013). ‘Tail-breakage
effects on snake-body condition’. Zoologischer Anzeiger 252: 243–245.
Rando, J.C., Alcover, J.A., Navarro, J.F., Garcia-Talavera, F., Hutterer,
R. and Michaux, J. (2008). ‘Chronology and causes of the extinction of
the lava mouse, Malpaisomys insularis (Rodentia: Muridae) from the
Canary Islands’. Quaternary Research 70: 141–148.
Robinson, J.E., Griffiths, R.A., St. John, F.A.V. and Robert, D.L.
(2015). ‘Dynamics of the global trade in live reptiles: Shifting
trends in production and consequences for sustainability’. Biological
Conservation 184: 42–50.
Rodda, G.H. and Fritts, T.H. (1992). ‘The impact of the introduction of
the brown tree snake, Boiga irregularis, on Guam's lizards’. Journal of
Herpetology 2: 166–174.
Rodda, G.H. and Savidge, J.A. (2007). ‘Biology and impacts of Pacific
island invasive species. 2. Boiga irregularis, the brown tree snake
(Reptilia: Colubridae)’. Pacific Science 61: 307–324.
Rogers, H., Hille Ris Lambers, J., Miller, R. and Tewksbury, J.J. (2012).
‘‘Natural experiment’ demonstrates top-down control of spiders by
birds on a landscape level’. PLOS ONE 8(4): e43446.
Santos, X., Feriche, M., León, R., Filippakopoulou, A., Vidal-García, M.,
Llorente, G.A. and Pleguezuelos, J.M. (2011). ‘Tail breakage frequency
as an indicator of predation risk for the aquatic snake Natrix maura’.
Amphibia-Reptilia 32: 375–383.
Savidge, J.A. (1987). ‘Extinction of an island forest avifauna by an
introduced snake’. Ecology 68: 660–668.
Valido, A. Nogales, M. and Medina, F. (2003). ‘Fleshy fruits in the diet
of Canarian lizards Gallotia galloti (Lacertidae) in a xeric habitat of the
island of Tenerife’. Journal of Herpetology 37(4): 741–747.
Willson, J.D. (2017). ‘Indirect effects of invasive Burmese pythons on
ecosystems in southern Florida’. Journal of Applied Ecology 54: 1251–
1258.
Wiseman, K.D., Greene, H.W., Koo, M.S. and Long, D.J. (2019). ‘Feeding
ecology of a generalist predator, the California kingsnake (Lampropeltis
californiae): why rare prey matter.’ Herpetological Conservation and
Biology (in press).
331
G. Picó, M.J. Fernández, J.E. Moreno and V. Colomar
Picó, G.; M.J. Fernández, J.E. Moreno and V. Colomar. Control of the ladder snake (Rhinechis scalaris) in Formentera using experimental live-traps
Control of the ladder snake (Rhinechis scalaris) on Formentera using
experimental live-traps
G. Picó1, M.J. Fernández1, J.E. Moreno2 and V. Colomar1
COFIB (Consorci per a la Recuperació de la Fauna de les Illes Balears), Ctra. Sineu Km 14.400, 07142, Santa
Eugènia, Spain. <gabrielap.pico@gmail.com>. 2SPE (Servei de Protecció d’Espècies), DG Espais Naturals i
Biodiversitat. Conselleria Med. Ambient, Agricultura i Pesca, Govern de les Illes Balears, Gremi de Corredors,
10, 07009, Palma, Spain.
1
Abstract The ladder snake (Rhinechis scalaris) is a recent alien invasive species found on Formentera (83 km2),
in the Balearic Archipelago (4,492 km2). It has been introduced in the last decade as cargo stowaway hidden within
ornamental olive trees from the Iberian Peninsula, causing negative impacts on native fauna. This paper describes the
methodology used to reduce the ladder snake population as a first attempt since it was detected in 2006. For this purpose,
an experimental live-trap was designed by the wildlife management team of the Consorci per a la Recuperació de la
Fauna de les Illes Balears (COFIB) during the 2016 campaign. As a result, 314 R. scalaris were trapped in an area of 472
ha, achieving an efficiency of up to 0.167 captures per trap and night, and 0.040 captures per unit effort on average. This
outcome encourages the use of the live-trap as a cost-effective method for reducing the snake population in Formentera.
Nonetheless, this method should be considered a starting point toward R. scalaris control.
Keywords: alien species, Balearic Islands, ophidian, population
INTRODUCTION
The accidental transportation of invasive alien species
to new locations is a major cause of biodiversity loss
worldwide. This is of special concern in island ecosystems,
where native species are especially vulnerable to biological
invasions (Quammen, 1996). In this regard, the presence
of reptiles in the Balearic Islands is a paradigmatic case,
with a greater number of alien species (19) than native ones
(2), namely, the Lilford’s wall lizard (Podarcis lilfordi) and
the Ibiza wall lizard (Podarcis pityusensis) (Silva-Rocha,
et al., 2015). The ladder snake (Rhinechis scalaris) is
a Mediterranean species which is present in most of the
Iberian Peninsula (just missing on the Cantabric ledge),
in the south-east of France and the north-east of Italy
(Pleguezuelos & Honrubia, 2002). This ophidian is
also considered an introduced species on other Spanish
islands, namely, Ons, Aurosa (in Pontevedra), Mallorca,
Menorca, Ibiza and Formentera (in the Balearic Islands)
(Pinya & Carretero, 2011), but fortunately not on any of
the surrounding islets (Carretero & Silva-Rocha, 2015).
However, in Menorca R. scalaris has a wide distribution and
is catalogued as a protected species in the Catàleg Balear
d’Espècies Amenaçades (Decret, 2005) due to its presence
on the island dating from the pre-Roman period (Vigne
& Alcover, 1985). Conversely, on Mallorca, Ibiza and
Formentera R. scalaris is a recent introduction (Álvarez,
et al., 2010; Mateo & Ayllón, 2012), so its presence is still
isolated to particular locations and its range is expanding.
In fact, in Ibiza and Formentera this ophidian is catalogued
as an invasive alien species (Real Decreto, 2013).
Until 2006 Formentera was considered snake-free. The
first R. scalaris was detected on 25 May 2008, followed
by another sighting on 17 July 2008; both located near La
Mola. Then, a third specimen, not identified, was recorded
the 20 May 2009 (Álvarez, et al., 2010; Mateo & Ayllón,
2012). It is presumed that the first ophidian was introduced
to the Pityusic islands through the trade of ornamental
olive trees originating from the Iberian Peninsula (Álvarez,
et al., 2010; Carretero & Silva-Rocha, 2015; Montes, et al.,
2015), and genetic studies suggest that the whole R. scalaris
population comes from one introduction event (SilvaRocha, et al., 2015). Nonetheless, it would be expected that
all the snakes spotted in Formentera during the first years
could come from Ibiza, since direct connections between
Formentera and the mainland are rather limited (Álvarez,
et al., 2010; Mateo & Ayllón, 2012).
The naturalisation of this ophidian could result in
important consequences for the Pityusic ecosystem and
also for the demographic stability of the endemic Ibiza
wall lizard (Rodríguez-Pérez, 2009; Álvarez, et al.,
2010). Previous cases of introduction of snakes to island
ecosystems have been terrible in terms of ecological balance
as experienced by the ancient settlement of ophidians
on the neighbouring islands of Mallorca and Menorca
(SPE, 2007), the deliberate release of the Californian
kingsnake (Lampropeltis getula californiae) on Gran
Canaria (Cabrera-Pérez, et al., 2012) and the accidental
introduction of the brown treesnake (Boiga irregularis) to
the island of Guam (Savidge, 1987; Rodda, et al., 1997;
Fritts & Rodda, 1998; Wiles, et al., 2003).
In the last decade, sightings from local people have
increased and as Carretero & Silva-Rocha (2015) stated,
“the area of Formentera where ladder snakes were spotted
in the past, should be checked thoroughly and regularly”.
So, the need to monitor the presence of R. scalaris on
Formentera is a real concern.
The present paper reports the first experience of
trying to catch and reduce the presumed population of R.
scalaris in the vicinity of La Mola in Formentera during
the 2016 campaign. For this purpose, an experimental
live-trap was designed by the wildlife management team
of the Consorci per a la Recuperació de la Fauna de les
Illes Balears (COFIB), along with the Government of the
Balearic Islands. Budget constraints restricted the scope of
this first campaign to confirming and mapping the presence
of the ladder snake in the vicinity of La Mola. So, the aim
was to determine the effectiveness of the trap, as defined
by captures per unit effort (CPUE), in order to establish a
starting point for future campaigns.
MATERIALS AND METHODS
Study area
This study was conducted on Formentera, the smallest
(83 km2) and southernmost island of the Balearic archipelago
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
332
up to meet the challenge, pp. 332–336. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Picó, et al.: Control of ladder snake on Formentera
and the opportunity to glimpse the animals. Therefore, the
trap is not completely opaque. There is just one entrance
with a one-way flap door positioned on the mesh front of
the trap, with a diameter of 3.5 cm. The flap only opens
inwards, falling closed behind the snake to prevent escape
and allowing multiple captures. This one-way flap entrance
design has been used on a number of snake trap designs
(reviewed by Rodda, et al., 1999a). Inside the snake
compartment a hide is placed: a 300 mm length of 100 mm
diameter plastic bottle, covered with 40 mm of water to
prevent snake dehydration.
A live mouse, with enough water and food for optimal
welfare, is used as attractant. In this trap, the bait is
contained in a separate compartment to prevent the snake
from ingesting the mouse.
Fig. 1 Map of the Balearic Islands, showing the name of
the main islands.
(Fig. 1). A channel of 3.6 km separates Formentera from
the other Pityusic Island, Ibiza, and it is 100 km away
from the Iberian Peninsula. Vegetation consists of sand
dunes with pine forest, oak groves and brushwood. It is
considered a flat island, with the highest point being La
Mola, at 192 m above sea level, in the south-east of the
island. This is where the R. scalaris population seems to be
concentrated, thus, the study focuses on this area
The live-trap
To conduct this study, live-traps were designed by the
COFIB for the purpose of capturing colubrid snakes. The
trap used on Formentera was the same as those used in the
project “Análisis de la efectividad de métodos de control
de especies exóticas invasoras de la familia colubridae en
islas” (COFIB, 2016) that took place simultaneously in a
parallel campaign on Ibiza.
The trap measures 50.0 × 35.5 × 17.0 cm and is made
of marine plank (1 cm thick) in order to endure inclement
weather conditions (Fig. 2). The box consists of two
compartments separated by a galvanised steel mesh of 5 ×
5 mm, with two large doors on top to allow snake removal
and bait maintenance. These doors are secured with a bolt
in order to prevent escapes. The front side is also made of a
galvanized steel mesh, allowing air flow through the mesh
Fig. 2 Model of the live-trap designed by the COFIB. A:
the two doors of the top, B: galvanized steel mesh that
separates the two compartments, C: galvanized steel
mesh of the front side, D: one-way flap door, E: detail of
the trapping door viewed from the inside.
Trapping method
In the 2016 campaign, trap boxes were placed in the area
near La Mola (Fig. 3), mostly at the limit of pine forests,
near stone walls or at the base of vegetation (Montes, et al.,
2015), all of them at ground-level. We covered a total area
of 472 ha with 64 traps. Fourteen extra traps were placed in
different locations on the island where no snakes had been
spotted in the past, as snake indicators. All the traps located
in the field were georeferenced.
Every effort was made to keep the mice alive during
the whole project, as they are the basis for the operation
of the trap (Mateo & Ayllón, 2012). During the coldest
months of the year dry grass or similar materials were
provided and the boxes were placed in the sunlight,
avoiding hypothermia. Conversely, in summer the boxes
were moved slightly towards the shade, helping the mice
to endure the suffocating heat. Also, for the duration of
the rainy season, traps were placed on stones and covered
with plastic, preventing contact between the bottom of
the box and a waterlogged ground. These measures were
taken not just for humane and economic reasons but also
because they allowed a longer period between inspections.
All traps were checked and bait replaced every nine days,
on average.
Capture and data gathering
When an ophidian was captured, it was identified to
ensure it was a R. scalaris (as opposed to an unknown and
possibly venomous snake), so handling did not require
Fig. 3 Distribution of the ladder snake (Rhinechis scalaris)
in Formentera in 2016. The number of captures per trap
is represented with different shapes (see the key). The
sample area and the capture area are determined by
applying the minimum convex polygon method. All traps
on the field are represented on the map but none of the
indicator traps are included in the sample area.
333
Island invasives: scaling up to meet the challenge. Ch 2C Other taxa: Herpetofauna
DISCUSSION
Fig. 4 Two specimens of ladder snake (Rhinechis scalaris),
a juvenile on the left and an adult on the right, caught
with the same live-trap in the field on Formentera.
anything special other than a pair of gardening gloves
(Fig. 4). We euthanised the snakes using pentobarbital (the
approximate dose was 0.1 ml per 100 g) and all injections
were performed by an experienced veterinarian, who
was the main field technician. Once the specimen was
lifeless, it was placed in a ‘zip’ bag with an identification
label including trap number, species, and date. Each
captured specimen was stored in a freezer for further
investigation; no morphological data was collected on
the field. Afterward, all captures were mapped in order to
estimate the abundance of R. scalaris in La Mola, using the
minimum convex polygons tool of Quantum GIS (1.8.0).
RESULTS
A total of 64 traps where placed on La Mola, remaining
in the same location for the entire sampling period (Fig.
3). The trapping was conducted between May and late
November. The team was based on Ibiza and, for this
reason, both sea conditions and vehicle availability
restricted the number of possible surveys to 19. The
number of traps increased in nearly every survey till the
end of July and August, when we had the total number of
traps placed in the field. The capture area comprised 321
ha from a sampling area of 472 ha. No indicator trap had
any capture.
By the end of the campaign, 314 ladder snakes had
been caught, with a total of 7,906 trap-nights (Table 1). It
is evident that this was a grass-roots effort, using the best
available knowledge to catch as many snakes as possible
while keeping costs as low as possible. Therefore, we did
not have the time to estimate the density of snakes prior
to the trapping. Instead of this, we evaluated the trap
effectiveness as defined by captures per unit effort (CPUE).
By the end of the project we had an average of 0.040 CPUE.
In May, the first month of trapping, we obtained a
trap efficiency of 0.108 captures per trap and night. Next
month, June, we got 0.075 captures per unit effort (CPUE),
even though the number of traps in the field was more than
double. A similar pattern occurred in the following months:
the captures per unit effort continued dropping, until we
got 0.006 CPUE in November. Therefore, preliminary
data seems to indicate an encouraging trap capture decay
rate, with a high CPUE at the beginning and a declining
recovery from traps as the local supply of snakes depleted.
However, seasonal changes in capture success need to be
evaluated.
334
The 2016 trapping campaign is the first attempt
to remove large numbers of snakes as a step towards
controlling the invasion of R. scalaris on the island of
Formentera. Previous attempts on the neighbouring islands
of Ibiza (Montes, et al., 2015) and Mallorca (Mateo, 2015)
have tested different methods to capture R. scalaris and
Hemorrhois hippocrepis. After a thorough review of these
documents, we decided to use a passive method to trap as
many snakes as possible, continuing the work of Montes,
et al. (2015) by adapting the wooden box they used. In this
regard, we followed the advice and recommendations of
previous snake trapping studies. As Rodda, et al. (1999a)
showed, it is possible to have higher capture rates using
live mice as lures, opaque chambers and flap entrances.
Firstly, flap traps have a lower entry rate than open funnel
traps, but the former have a higher capture rate. For this
reason, we replaced the two open-funnels used by Montes,
et al. (2015), with a single frontal flap door, as these are
considered to have a negligible escape rate (Rodda, et al.,
1999a).
Secondly, in contrast with the lack of a mouse’s chamber
in the wooden box by Montes, et al. (2015) and the small
one that housed the mouse inside the funnel trap described
by Mateo & Ayllón (2012), our trap had a proper shelter for
the mouse, which was the second big compartment of the
cage. With this modification we avoided snake ingestion of
the lure and contributed to reducing mouse mortality.
Finally, in order to enhance capture success, refugium
bottles were placed inside the snake’s compartment as
it has been observed that there is a significantly higher
number of entries into traps having hiding places, even if
the possibility of escape is unlikely (Rodda, et al., 1999a).
Our trap optimises previous designs and the positive
capture rates seem to be the result of using both a flap
door and a bottle refuge, as these contribute to reducing
the number of snakes escaping, along with the separate
compartments, which keeps the trap active after a first
successful capture. Indeed, our results (0.040 CPUE)
confirm a higher efficiency when compared with the study
by Montes, et al. (2015) (0.007 CPUE).
All data could have been more accurate had we had
a technician exclusively dedicated to checking the traps
every other day. Then, not only the number of traps per
hectare would have been greater, but the capture rate
probably higher. In this case, re-check intervals were
determined in relation to care and maintenance of live
lures (as the snakes had enough water to avoid death by
dehydration during these intervals) instead of capture rate
increase. This allowed optimising labour and maximising
cost-effectiveness. Even so, trap captures are hypothesised
to be higher if the area of trapping is not disturbed (Rodda,
et al., 1999a), suggesting normal entrance rates if checks
are done within longer intervals. In this study, traps were
checked weekly during the summer season but checks
were done every 12 days in autumn.
Regarding trap location, Rodda, et al. (1999a), argue
that traps should be widely spaced in order to maximise the
capture rate when traps are infrequently checked. However,
there is still a lack of a mathematic equation describing the
relationship between capture rate and trap spacing, as well
as a poorly understood interaction between trap design and
the environment in which it is used. Taking into account
that R. scalaris is an active forager (Pleguezuelos, et al.,
2007), traps were placed as far apart as it was practical for
revisits considering the topography, the trapping area and
the number of traps available, resulting in a wide range
from 50 m to 600 m apart.
Picó, et al.: Control of ladder snake on Formentera
Table 1. Data on ladder snakes (Rhinechis scalaris) caught during the 2016 campaign on Formentera.
Survey
1
2
3
4
Month
May
May
May
May
Average for May
No. Traps
9
18
18
43
22
5
6
7
8
June
June
June
June
Average for June
56
57
57
63
58.25
41
40
36
13
130
306
392
497
532
1,727
0.134
0.102
0.072
0.024
0.075
9
10
11
July
July
July
Average for July
63
63
64
63.33
29
11
2
42
635
420
506
1,561
0.046
0.026
0.004
0.027
12
13
August
August
Average for August
64
64
64
14
9
23
408
408
816
0.034
0.022
0.028
14
September
Average for September
43
43
6
6
352
352
0.017
0.017
15
16
17
October
October
October
Average for October
43
44
44
43.67
14
18
7
39
645
602
396
1,643
0.022
0.030
0.018
0.024
18
19
November
November
Average for November
5
2
7
314
616
572
1,188
7,906
0.008
0.003
0.006
0.040
44
44
44
TOTAL
The first traps distributed on the ground were placed
within view of neighbours (on the south of the road to
the lighthouse of La Mola), and then a consecutive radial
expansion was drawn. As can be seen in Fig. 3, there is a
clear ‘hot spot’ to the south of La Mola, with the highest
number of captures close to the southern coast. The
number of captures decreases the further we move from
this high-density area and the traps on the north and west
boundaries are characterised by no captures (except for
two traps on the west). Considering that the sea is a natural
barrier, potential expansion is only possible to the north
or to the west of the sample area. As mentioned above, no
indicator traps in other parts of the island had any captures.
Therefore, our trap array gives an initial indication about
the range of R. scalaris on Formentera, having a higher
density of snakes in the core of the invasion zone than at
the edges. Still, a larger array of traps around La Mola,
especially on the west boundary, would depict the range of
R. scalaris more accurately.
It is clear that a population of R. scalaris is naturalised
in Formentera. Previous extinctions of endemic birds
and lizards have been documented as a result of the
introduction of an alien snake, such as the well-known case
of the B. irregularis in the Island of Guam (Savidge, 1987;
Rodda, et al., 1997; Fritts & Rodda, 1998). Therefore, the
Guam experience should made us wary of the invasive
potential that R. scalaris could have on the native fauna
No. Captures Trap/night
3
36
15
90
12
157
37
336
67
619
CPUE
0.083
0.167
0.076
0.110
0.108
of the Pityusic islands. It has the potential to affect a wide
range of animals, such as the emblematic Ibiza wall lizard,
the Balearic shearwater (Puffinus maruritanicus), the
Scopoli's shearwater (Calonectris diomedea), the storm
petrel (Hydrobates pelagicus) or the garden dormouse of
Formentera (Eliomys quercinus ophiusae) (Hinckley, et al.,
2016), as few predators are present on Formentera and the
abundant endemic fauna is an easy and vulnerable target
because prey species lack co-evolutionary experience with
snakes (Rodda, et al. 1999b).
Successful control of R. scalaris is Formentera’s
highest conservancy priority (Pleguezuelos, et al., 2015).
This is an early invasion, in chronological terms, and the
area of invasion seems to be relatively small. The use of
this wooden box trap seems to be a useful starting point
towards R. scalaris control. However, more comprehensive
research is required to determine whether the ladder
snake’s expansion on Formentera can be stopped by using
this capture method. In order to assess this question, the
study will continue in future years with a greater trap array.
ACKNOWLEDGEMENTS
This work was supported by the Consell Insular de
Formentera and Eivissa, Agents de Medi Ambient and
IBANAT. The wooden trap would not have been possible
without the help of the Escola d’Arts i Oficis d’Eivissa in
335
Island invasives: scaling up to meet the challenge. Ch 2C Other taxa: Herpetofauna
making the first prototype. This project has been funded
by the FEADER (Fondo Europeo Agrario de Desarrollo
Rural) through the Desenvolupament Rural de les Illes
Balears Programme.
REFERENCES
Álvarez, C, Mateo, J.A., Oliver, J. and Mayol, J. (2010). ‘Los ofidios
ibéricos de introducción reciente en las Islas Baleares’. Boletín de la
Asociación Herpetológica Española 21: 126–131.
Cabrera-Pérez, M.A., Gallo-Barneto, R., Esteve, I., Patiño-Martínez, C.,
and López-Jurado, L.F. (2012) ‘The management and control of the
California kingsnake in Gran Canaria (Canary Islands): Project LIFE+
Lampropeltis’. Aliens: The Invasive Species Bulletin 32: 20–28.
Carretero, A. and Silva-Rocha, I. (2015). ‘La culebra de escalera
(Rhinechis scalaris) en las Islas Baleares’. Boletín de la Asociación
Herpetológica Española 26(2): 84–87.
COFIB (Consorci per a la Recuperació de la Fauna de les Illes Balears)
(2016). Análisis de la Efectividad de Métodos de Control de Species
Exóticas Invasoras de la Familia Colubridae en Islas. Unpublished
report. Palma: COFIB.
Decret (2005). ‘D.75/2005, de 8 de juliol, pel qual es crea el Catàleg
Balear d’Espècies Amenaçades i d’Especial Protecció, les Arees
Biològiques Crítiques i el Consell Assessor de Fauna i Flora de les Illes
Balears’. Boletín Oficial de las Islas Baleares (BOIB) 106. <http://www.
caib.es/sites/institutestudisautonomics/ca/n/decret_752005_de_8_de_
juliol_pel_qual_es_crea_el_cataleg_balear_despecies_amenacades_i_
despecial_proteccio_les_arees_biologiques_critiques_i_el_consell_
assessor_de_fauna_i_flora_de_les_illes_balears_-58273/>.
Fritts, T.H. and Rodda, G.H. (1998). ‘The role of introduced species in
the degradation of island ecosystems: A case history of Guam’. Annual
Review of Ecology, Evolution, and Systematcs 29: 113–140.
Hinckley, A., Montes, E., Ayllón, E. and Pleguezuelos, J.M. (2016). ‘The
fall of a symbol? A high predation rate by the introduced horseshoe
whip snake Hemorrhois hippocrepis paints a bleak future for the
endemic Ibiza wall lizard Podarcis pityusensis’. European Journal of
Wildlife Research 63(1): 13.
Mateo, J.A. (2015). ‘Los Anfibios y los Reptiles Introducidos en Baleares:
un repaso a lo que sabemos y un ejemplo de puerta de entrada’. In: Servei
de Protecció d’Espècies (ed.) Llibre Verd de Protecció d’Espècies a les
Balears, pp. 447–454. Palma: Govern de les Illes Balears. Conselleria
d’Agricultura, Medi Ambient i Territori.
Mateo, J. A. and Ayllón, E. (2012). Viabilidad del Control de Ofidios en
Ibiza y Formentera. Palma: Asociación Herpetológica Española.
Montes, E., Estarellas, J., Ayllón, E., Carretero, M.A., Feriche, M.,
Hernández, P.L. and Pleguezuelos, J.M. (2015). ‘Dades Preliminars
del Projecte Pilot de Control de Serps a l’Illa d’Eivissa’. In: Servei de
Protecció d’Espècies (ed.) Llibre Verd de Protecció d’Espècies a les
Balears, pp. 455–464. Palma: Govern de les Illes Balears. Conselleria
d’Agricultura, Medi Ambient i Territori.
Pinya, S. and Carretero, M.A. (2011). ‘The Balearic herpetofauna: A
species update and a review on the evidence’. Acta Herpetologica 6:
59–80.
Pleguezuelos, J.M. and Honrubia, S. (2002). ‘Elaphe scalaris’. In: J.M.
Pleguezuelos, R. Márquez and M. Lizana (ed.) Atlas y Libro Rojo de
los Anfibios y Reptiles de España, pp. 281–283. Madrid: Organismo
Autónomo de Parques Nacionales.
Pleguezuelos, J.M, Fernández-Cardenete, J.R., Honrubia, S., Feriche,
M., Villafranca, C. (2007). ‘Correlates between morphology, diet and
foraging mode in the ladder snake Rhinechis scalaris (Schinz, 1822)’.
Contributions to Zoology 76(3): 179–186.
Pleguezuelos, J.M., Ayllón, E., Arribas, O., Bertoloero, A., Bosch, J.,
Cabido, C., Carranza, S., Carretero, M.A., Díaz-Paniagua, C., EgeaSerrato, A., Garin-Barrio, I., Giménez, A., Gosá, A., Garciá, E.,
Guicking, D., Llorente, G.A., Martínez-Solano, I., Mateo, J.A., Montori,
A., Palomar, G., Perera, A., Pinya, S., Pretus, J.L., Pujol-Buxó, E., Rato,
C., Recuero, E., Sanz-Azkue, I., Silva-Rocha, I., Vasconcelos, R., VeloAntón, G., Vörös, J., Santos, X. (2015). ‘Conclusiones y propuestas de
gestión de las poblaciones y especies de anfibios y reptiles alóctonos
en España’. Boletín de la Asociación Herpetológica. Española 26 (2):
113–115.
Quammen, D. (1996). The Song of the Dodo: Island Biogeography in an
Age of Extinction. New York: Simon and Schuster.
Real Decreto (2013). ‘RD.630/2013, de 2 de agosto, por el que se Regula
el Catálogo Español de Especies Exóticas Invasoras’. Boletín Oficial
del Estado (BOE) 185. <https://www.boe.es/buscar/pdf/2013/BOE-A2013-8565-consolidado.pdf >.
336
Rodda, G.H., Fritts, T.H., Clark, C.S., Gotte, S.W. and Chiszar, D., (1999a)
‘A State-of-the-Art Trap for the Brown Treesnake’. In: G.H. Rodda, Y.
Sawai, D. Chiszar, and H. Tanaka (ed.) Problem Snake Management:
The Habu and the Brown Treesnake, pp. 268–305. Ithaca and London:
Cornell University Press.
Rodda, G.H.; Fritts, T.H.; McCoid, M.J. and Campbell, E.W. III. (1999b).
‘An Overview of the Biology of the Brown Treesnake, Boiga irregularis,
a Costly Introduced Pest on Pacific Islands’. In: G.H. Rodda, Y. Sawai,
D. Chiszar, and H. Tanaka (ed.) Problem Snake Management: The
Habu and the Brown Treesnake, pp. 44–80. Ithaca and London: Cornell
University Press.
Rodda, G.H., Fritts, T.H. and Chiszar, D. (1997). ‘The disappearance of
Guam’s wildlife; new insights for herpetology, evolutionary ecology,
and conservation’. BioSience 47: 565–574.
Rodríguez-Pérez, J. (2009). ‘Pérdidas de interacciones planta-animal en
islas: dispersión de semillas por la lagartija balear y sus consecuencias
para la regeneración de un arbusto endémico’. El Indiferente 20: 36–43.
Savidge, J.A. (1987). ‘Extinction of an island forest avifauna by an
introduced snake’. Ecology 68: 660–668.
SPE (Servei de Protecció d’Espècies) (2007). Pla de Recuperació del
Ferreret (Alytes muletensis). Palma: Conselleria de Medi Ambient.
Govern de les Illes Balears.
Silva-Rocha, I., Salvi, D., Sillero, N., Mateo, J.A. and Carretero,
M.A. (2015). ‘Snakes on the Balearic Islands: An invasion tale with
implications for native biodiversity conservation’. PLOS ONE 10(4):
e0121026. doi:10.1371/journal.pone.0121026
Vigne, D. and Alcover, A. (1985). ‘Incidence des relations historiques
entre l’homme et l’animal dans la composition actuelle du peuplement
amphibien, reptilien et mammalien des îles de méditerranée occidentale’.
Actes 110e Congrès national des Sociétés savantes, Section Sciences:
79–91.
Wiles, G.J., Bart, J., Beck, Jr., R.E., and Aguon, C.F. (2003). ‘Impacts
of the brown tree snake: patterns of decline and species persistence in
Guam’s avifauna’ Conservation Biology 17: 1350–1360.
A.N. Pili, C.E. Supsup, E.Y. Sy, M.L.L. Diesmos and A.C. Diesmos
Pili, A.N.; C.E. Supsup, E.Y. Sy, M.L.L. Diesmos and A.C. Diesmos. Spatial dynamics of invasion and distribution of alien frogs in a biodiversity hotspot archipelago
Spatial dynamics of invasion and distribution of alien frogs in a
biodiversity hotspot archipelago
A.N. Pili1,7, C.E. Supsup2,7, E.Y. Sy3,7, M.L.L. Diesmos4,5,7 and A.C. Diesmos 1,6,7
The Graduate School, University of Santo Tomas, España, 1015 Manila, The Philippines. <armannorciopili@gmail.
com> 2Biology Department, College of Science, De La Salle University, 2401 Taft Avenue, Manila, The Philippines.
3
Philippine Center for Terrestrial and Aquatic Research, Tondo, Manila, The Philippines. 4Department of Biological
Sciences, College of Science, University of Santo Tomas, España, 1015. Manila, The Philippines. 5Research Center for
the Natural and Applied Sciences, University of Santo Tomas, España, 1015. Manila, The Philippines. 6 Herpetology
Section, Zoology Division, National Museum of the Philippines, Ermita, 1000. Manila, The Philippines. 7Herpwatch
Pilipinas, Inc., Tondo, Manila, The Philippines.
1
Abstract The endemic-rich amphibian fauna of the Philippine Archipelago (ca. 350,000 km2) includes six alien frogs: the
American bullfrog (Lithobates catesbeianus), Asiatic painted toad (Kaloula pulchra), cane toad (Rhinella marina), Chinese
bullfrog (Hoplobatrachus rugulosus), green paddy frog (Hylarana erythraea), and greenhouse frog (Eleutherodactylus
planirostris). The chronological history of their invasion across the Philippines was reconstructed based on historical and
geographic data. Subsequently, we estimated their current and potential distribution through species distribution modelling
and Gaussian kernel density smoothing species distribution data. Seven known and potential pathways of introduction into
and spread throughout the Philippines were identified, namely, intentional introduction as a (1) biocontrol agent and (2)
food source; contamination of (3) agriculture trade, (4) aquaculture trade, and (5) ornamental plant trade; (6) stowaway of
cargo; and (7) through the exotic pet trade. Spatio-temporal patterns of distribution showed a stratified diffusion process
of spread wherein human-mediated jum dispersal is the primary mode followed by diffusion dispersal. The status of the
American bullfrog in the Philippines is unresolved, whether it has successfully established. Meanwhile, the other five alien
frogs have established populations in the wild, typically the dominant species in both artificial and disturbed habitats, and
are continuously spreading throughout the Philippines. Estimates of current and potential distribution indicate that none
of the alien frogs has realised its full potential distribution and that the cane toad is the most widespread, occurring in
almost all major islands of the Philippines (ca. 85%), while the greenhouse frog is the least distributed, being found so far
in eight provinces and on seven islands. In light of these findings, we provide policy and management recommendations
for responding to current and future alien frog invasions.
Keywords: frogs, geographic risk assessment, invasion history, invasive alien species, policy and management
INTRODUCTION
The Philippines (Fig. 1) is the second largest archipelago
in the world, with ca. 7,641 islands, and is recognised as a
megadiverse nation and a global biodiversity conservation
hotspot (Heaney & Mittermeier, 1997; Heaney, et al.,
1999; Myers, et al., 2000a). A compelling example of its
rich biodiversity is exhibited by the country’s amphibian
assemblage, which is among the most important faunas in
the Indomalayan Region in terms of diversity and endemism
(Bain, et al., 2008; Diesmos, et al., 2014). Currently, there
are 110 native species of amphibians known from the
Philippines, 97 of which (ca. 91%) are endemics (Diesmos,
et al., 2015). However, ca. 45% of Philippine amphibians
are threatened with extinction: the major threats include
habitat loss and deforestation, invasive alien species,
emerging infectious diseases, and climate change (Alcala,
et al., 2012; Brown, et al., 2012; Diesmos, et al., 2014).
Included in the Philippine amphibian fauna are
six introduced frogs, namely, the American bullfrog
(Lithobates catesbeianus [Shaw, 1802]), the Asiatic
painted toad (Kaloula pulchra Gray, 1831), the cane toad
(Rhinella marina [Linnaeus, 1758]), the Chinese bullfrog
(Hoplobatrachus rugulosus [Wiegmann, 1834]), the green
paddy frog (Hylarana erythraea [Schlegel, 1837]), and the
greenhouse frog (Eleutherodactylus planirostris [Cope,
1862]) (Fig. 2; Diesmos, et al., 2006; Diesmos, et al., 2014;
Olson, et al., 2014; Diesmos, et al., 2015). Preliminary
studies and anecdotal reports indicated that these introduced
species, particularly the cane toad and the Chinese bullfrog,
are harmful invasives, threatening Philippine wildlife
through competitive exclusion and direct predation
(Rabor, 1952; Alcala, 1957; Soriano, 1964; Espiritu, 1985;
Adraneda, et al., 2005; Diesmos, et al., 2006). Diesmos,
et al., (2006) provided the first review on the status and
distribution of alien frogs in the Philippines (then only
five alien frogs were present). However, there remains a
large knowledge gap on their history of invasion and no
recent attempts have been made to synthesise the growing
body of knowledge on their geographic distribution. By
assembling and analysing historical and geographical data
of the six alien frogs in the Philippines, we reconstructed
the chronological history of invasion and updated their
status and distribution. We then estimated their current and
potential distribution by projecting suitable areas based on
two separate species distribution models (“native range
models” and “Philippine models”) and, subsequently,
Gaussian kernel density smoothing distribution data to
delineate occupied suitable areas (“current distribution”)
and unoccupied suitable areas (“potential distribution”).
METHODS
Reconstructing history of invasion
We reconstructed the chronological history of invasion
of the six alien frog species in the Philippines based on
historical and geographical data (“species distribution
data”) obtained from the following sources: (1) Natural
history collections (NHC): data obtained directly from
collections managers of local and international institutions
or through the Global Biodiversity Information Facility
(GBIF); (2) published and (3) unpublished scientific
literature; and (4) personal observations of authors and
fellow experts.
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 337–347. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
337
Island invasives: scaling up to meet the challenge. Ch 2C Other taxa: Herpetofauna
Species distribution modelling
Species Distribution Modelling (SDM) involves the
quantification of species-environment relationships to
define a species’ ecological niche. The ecological niche
models are then projected into geographic space to
visualise and yield an estimate of geographic range or
suitable areas where a species can or cannot persist (Guisan
& Zimmermann, 2000; Guisan & Thuiller, 2005; Elith &
Leathwick, 2009). In studies dealing with invasive alien
species, predictions of suitable areas are typically made
by extrapolating models fitted with data from the species’
native range onto areas that could be invaded (Peterson &
Vieglais, 2001; Venette, et al., 2010; Araújo & Peterson,
2012). However, empirical studies have shown evident
niche shift in invasive populations, suggesting that species
can occupy climatically distinct niche spaces following
their introduction into a new area (Broennimann, et al.,
2007; Beaumont, et al., 2009).
Here, we developed two separate projections of
Philippine-suitable areas for the alien frogs based on models
fitted with species distribution data and environmental data
(1) from the invaded range in the Philippines (hereafter
called “Philippine models”) and (2) from the alien frogs’
native ranges (hereafter called “Native models”). Because
of the limited amount of species distribution data, we did
not develop Philippine models for the American bullfrog
and the greenhouse frog.
Fig. 1 The Philippine archipelago overlaid on a hypsometric
raster shaded-relief. (1) Batanes Island Group, (2) Babyan
Island Group, (3) Luzon, (4) Polilio, (5) Catanduanes, (6)
Mindoro, (7) Marinduque, (8) Busanga, (9) Romblon
Island Group, (10) Masbate, (11) Samar, (12) Palawan,
(13) Panay, (14) Leyte, (15) Guimaras, (16) Cebu, (17)
Negros, (18) Bohol, (19) Dinagat, (20) Siargao, (21)
Siquijor, (22) Camiguin, (23) Mindanao, (24) Basilan, (25)
Samal, (26) Jolo, and (27) Tawi-Tawi. Copyright ArcGIS.
Data collection and calibration. The Philippine
models were fitted using species distribution data from
the Philippines (data used in reconstructing history of
invasion). Meanwhile, Native models were fitted using
species distribution data obtained from the GBIF. Sampling
bias was corrected through systematic subsampling
neighbouring species distribution data to a resolution of
one distribution point per five square kilometres or 2.5
arcminutes and by developing bias files (Elith, et al., 2010;
Fourcade, et al., 2014).
The original set of environmental variables includes 19
bioclimatic datasets (Worldclim – Hijmans, et al., 2005) and
Global Land Cover 2000 (GLC2000) (Fritz, et al., 2003)
Table 1 Calibration of ecological niche models. Shown are the species distribution data used for model training and
testing, model validation approach, number of replicates, and the Maxent features (L – linear; Q – quadratic; P –
product) used in fitting Philippine models (A) and Native range models (B) of the alien frogs. Due to the limited amount
of species distribution data viable for model fitting, Philippine models of the American bullfrog and the greenhouse frog
were not developed.
Species
Training
data
Testing
data
Validation
Replication
Maxent Features
L
Q
P
A. Philippine model
American bullfrog
10
-
-
Asiatic painted toad
23
-
Crossvalidation
Cane toad
-
-
-
-
10
-
114
38
Subsampling
10
Chinese bullfrog
79
10
Subsampling
10
Green paddy frog
101
33
Subsampling
10
6
-
-
-
-
-
3,704
1,234
Subsampling
10
93
31
Subsampling
10
1,582
527
Subsampling
10
83
27
Subsampling
10
Green paddy frog
57
18
Subsampling
10
-
Greenhouse frog
32
-
Crossvalidation
10
-
Greenhouse frog
B. Native range model
American bullfrog
Asiatic painted toad
Cane toad
Chinese bullfrog
338
-
Pili, et al.: Invasion & distribution of alien frogs
Fig. 2 Photographs in life of (a) the American bullfrog,
(b) the Asiatic painted toad, (c) the cane toad, (d) the
Chinese bullfrog, (e) the green paddy frog, and (f) the
greenhouse frog. Photographs copyright Tony Gerard
(a), Arman N. Pili (b), Emerson Y. Sy (c,d,e,f).
with a spatial resolution of 30 arc seconds. Environmental
variables used for fitting the Philippine models had a
spatial coverage from the Philippines only. Meanwhile,
Native models had a spatial coverage equivalent to the
native range of the species, based on a convex hull polygon
of species distribution data. For both Philippine and Native
models of each species, the environmental variables used
for model fitting were pre-selected to only include those
that are ecologically relevant (Austin, 2002; Wells, 2007)
to the species and are not highly inter-correlated (Dormann,
et al., 2013). Correlation between variables were assessed
using pair-wise Pearson’s correlation coefficient (stats R
version v.3.3.0 by R Core Team, 2016) and, subsequently,
we selected only the putatively ecologically most relevant
variable from each group of highly inter-correlated
variables (|r| ≥ 0.7) (Dormann, et al., 2013). The final set
of environmental variables used for model fitting included
(1) diurnal temperature, (2) temperature seasonality, (3)
maximum temperature of warmest month, (4) minimum
temperature of coldest month, (5) annual precipitation,
(6) precipitation seasonality, (7) precipitation of wettest
quarter, and (8) Global Land Cover 2000.
Model fitting. Species distribution modelling was
performed using Maximum Entropy Modelling (Maxent
v.3.3.3k) (Phillips, et al., 2004; Phillips, et al., 2006a).
Maxent is a general-purpose machine learning method
premised on the principle of maximum entropy and with a
simple and precise mathematical formulation for presence
only (i.e., species distribution data) modelling of species
distributions from incomplete information (Phillips, et al.,
2004; Phillips, et al., 2006a). Maxent has been found to
outperform other statistical approaches based on predictive
accuracy (Jeschke & Strayer, 2008; Elith & Graham, 2009).
Maxent settings used for fitting species distribution models
are shown in Table 2. The Maxent features (i.e. linear,
quadratic, product) used for each species’ models were
selected following Phillips (2005), Phillips, et al. (2006b)
and Phillips & Dudik (2008) suggestions and were based on
the number of species distribution points after systematic
subsampling. Developed bias files were incorporated in the
bias function of Maxent (Table 1). A logistic output was
selected to represent the predicted suitable habitats of the
species. Pseudo-absence data or background data were
generated at random within the Philippines for Philippine
models and within the native geographic range of each
species for Native models. All other Maxent settings were
set to default.
Model evaluation. Model performance of the
Philippine models of the cane toad, Chinese bullfrog,
and green paddy frog, and Native Models of all alien
frogs except the greenhouse frog was evaluated using the
area under receiver operating characteristic (ROC) curve
(AUC) by subsampling (randomly splitting presence/
pseudo-absences into two subsets with 70% of the records
used for model fitting and the remaining 30% to evaluate
the models) and was repeated 10 times (Table 1, Pearce
& Ferrier, 2000; Allouche et al., 2006; Araújo & Guisan,
2006). Meanwhile, due to the limited amount of species
distribution data, model performance of the Philippine
models of the Asiatic painted toad and Native Models of the
greenhouse frog was evaluated using the AUC values by
10-fold cross-validation and was repeated five to 10 times
(Pearce & Ferrier, 2000; Allouche, et al., 2006; Araújo &
Guisan, 2006) (Table 1). The AUC values were interpreted
based on Swets (1988) recommendation where 0.5–0.6 =
fail, 0.61–0.7 = poor; 0.71–0.8= fair, 0.81–0.9 = good, and
0.91–1.0 = excellent.
Projection. The models were projected to Philippine
geographic space to predict suitable areas for the species.
The projections were transformed into binary maps of
suitable/unsuitable areas, wherein areas above a minimum
training presence threshold (no omission) are referred to as
“suitable” areas (Liu, et al., 2005).
Table 2 History of invasion and current status and distribution of alien frogs in the Philippines.
Year and locality of
introduction or first
detection
Pathway of introduction
and spread
American bullfrog
Origin of
introduced
populations
Louisiana,
USA
1966 in Luzon Island
Food source
Asiatic painted
toad
Unknown
2003 in Luzon Island
Cane toad
Hawaii, USA
1934 in Luzon Island
Chinese bullfrog
Unknown
1993 in Luzon Island
Green paddy frog
Borneo Islandb
Greenhouse frog
Hawaii, USAb
1800s (unknown
locality); 1908 Panay Is Agricultural trade
2014 in Mindanao Is
Exotic plant trade
Species
Cargo Stowaway, Exotic
Pet Trade, Ornamental Plant
Trade
Biocontrol agent
Food source, Aquaculture
trade
Islands Provinces
Present Present
5
12
6
16
36
53
7
26
20
38
8
7
339
Island invasives: scaling up to meet the challenge. Ch 2C Other taxa: Herpetofauna
Estimating current and potential distribution
We define the current and potential distribution of
invasive alien species as respectively areas occupied and
unoccupied by the alien species conditional on areas of
suitable habitat (Gormley, et al., 2011). The geographic
ranges of the alien frogs in the Philippines were estimated
by two-dimensional Gaussian kernel smoothing assembled
species distribution data (kde2d function of MASS v.7.45 R
package; Ripley, et al., 2015). This method applies a twodimensional Gaussian kernel to compute distribution of an
animal within its home range/geographic range (Worton,
1989; Venables & Ripley, 2002; Gaston & Fuller, 2009).
The solve-the-equation method (width.SJ function MASS
R package; Sheather & Jones, 1991), was used to select the
bandwidth for kernel smoothing, and was defined to include
99.5% of species’ distribution data. Estimated geographic
ranges were then used to delineate the occupied suitable
areas (“current distribution”) and unoccupied suitable
areas (“potential distribution”). Because of the limited
amount of species distribution data, we did not estimate
the geographic range of the American bullfrog and the
greenhouse frog in the Philippines, and, consequently, we
did not delineate their current and potential distribution.
Hawaiian Sugar Planters Association and were brought to
the Philippines in 1934 (Merino, 1936). The toads were
initially reared on Luzon Island. Since then, they have
spread in all directions across islands and onto different
islands throughout the Philippines. Their spread is primarily
mediated by human movement (deliberate release for
biocontrol), as a cargo stowaway, and neighbourhood
diffusion dispersal (Rabor, 1952). Today, the cane toad can
RESULTS
History of invasion
A comprehensive review of the history of invasion,
including an assembled species distribution database,
of the six alien frogs in the Philippines is prepared in a
separate study for future publication. The review provided
below will suffice as a general overview of their history of
invasion.
The American bullfrog
Individuals of the American bullfrog were imported
from Louisiana, United States in 1966 and were first reared
on Luzon Island (Ugale, 1976; Pascual, 1987b). Frogs
were initially bred for the export production of scientific
specimens for biomolecular and medical research and
other educational activities (Pascual, 1987a; Urbanes,
1988; Urbanes, 1990; Matienzo, 1990). Subsequently,
in 1980, through government efforts to boost food
security, the American bullfrog breeding shifted to food
production. Another eight American bullfrog breeding
centres were established across the Philippines (Table 2;
Fig 3a; Ministry of Natural Resources, 1981; Buenviaje,
1983; Inovejas, 1985). Breeding centres ceased operation
in 1985. The current status of the American bullfrog in
the Philippines, whether they were able to successfully
establish populations in the wild, is unknown.
The Asiatic painted toad
The Asiatic painted toad was first reported in the
Philippines in 2003 on Luzon Island (Diesmos, et al.,
2006). It was earlier suggested that the initial introduction
of the Asiatic painted toad was through the exotic pet trade
(Diesmos, et al., 2006). Introduction as a contaminant
of ornamental plant trade or as cargo stowaway is also
plausible. From localities of its initial introduction, the
Asiatic painted toad has spread in all directions throughout
the Philippines and is now recorded in 16 provinces on
six islands (Table 2; Fig. 3b). It is likely that the identified
introduction pathways may have mediated its spread
throughout the Philippines.
The cane toad
The cane toad was intentionally introduced in the
Philippines as a part of a national pest control programme
(Merino, 1936). Cane toads were secured from the
340
Fig. 3 Geographic distribution of the alien frogs in the
Philippines, (a) American bullfrog, (b) Asiatic painted
toad, (c) cane toad, (d) Chinese bullfrog, (e) green paddy
frog, and (f) greenhouse frog. Points indicate areas
where alien frogs were reported present (release sites
or areas where bullfrog breeding centres were formerly
established for the case of the American bullfrog).
Pili, et al.: Invasion & distribution of alien frogs
be found on almost all major islands of the Philippines,
where it is usually the dominant amphibian species in
invaded areas (Table 2; Fig. 3c; Alcala, 1986; Alcala &
Brown, 1998; Diesmos, et al., 2006; Diesmos, et al., 2015).
The Chinese bullfrog
The Chinese bullfrog was first reported in the Philippines
in 1993 on Luzon Island (Diesmos, 1998; Diesmos et al.,
2006). It was speculated that this species was introduced
into and spread throughout the Philippines along with
American bullfrog breeding in the 1980s (Diesmos, et
al., 2006). Other potential pathways of introduction and
spread of the Chinese bullfrog throughout the Philippines
are contamination of agricultural trade, as for the case
of co-occurring alien and native frogs in the Philippines
(Inger, 1954; Kuraishi, et al., 2009), and contamination of
aquaculture trade, as was the case of its congeneric (Indian
bullfrog Hoplobatrachus tigerinus) on Andaman Islands,
India (Surendran & Vasudevan, 2013). The Chinese
bullfrog is now found in 26 provinces on seven islands in
the Philippines (Table 2; Fig. 3d).
especially in gardens (Olson et al., 2014; Sy & Salgo, 2015;
Sy, et al., 2015a, b; Sy, 2017a,b), suggests that the trade in
exotic ornamental plants is the most plausible pathway of
its introduction into and spread throughout the Philippines,
as was documented in Hawaii (Kraus, et al., 1999). The
greenhouse frog has so far been recorded in eight provinces
on seven islands (Table 2; Fig. 3f).
Philippine-suitable areas
Models of the alien frogs indicate fair to excellent
training-AUC values (>70) (Table 3). Based on projections
of Philippine-suitable areas of both the Philippine models
(except American bullfrog and greenhouse frog) and Native
models, the alien frogs are, to varying extents, suitable to
the Philippines. It should be noted that the Native models
consistently projected a broader range of Philippine-suitable
areas (Figs 4 & 5; Table 4). Moreover, both the Philippine
and Native models consistently projected human-modified
and disturbed areas to exhibit typical to high probability of
suitable conditions for these alien species.
Current and potential distribution
The green paddy frog
The earliest valid records of the green paddy frog,
overlooked in previous discussions regarding its history
of invasion (e.g., Inger, 1954), were collections from
Panay Island in 1908 (Orrell & Hollowell, 2017). In the
early 1900s, the green paddy frog was initially thought to
be native to the Philippines with restricted distribution on
the islands of Negros, Panay, Sibuyan and Tablas (Taylor,
1920; Taylor, 1922; Inger, 1954). Inger (1954) suggested
that the green paddy frog was introduced as a contaminant
of agricultural trade owing to its disjunct distribution from
the nearest extra-Philippine populations on Borneo Island.
The green paddy frog is now found in 38 provinces on 20
islands (Table 2; Fig. 3e). Contamination of agricultural
and aquaculture trade may be implicated for its spread
throughout the Philippines.
The greenhouse frog
The greenhouse frog was first detected on Mindanao
Island in 2013 (Olson, et al., 2014). the propensity of the
greenhouse frog to thrive in human-modified environments,
Maps show that the Asiatic painted toad has occupied
ca. 30–40% of projected suitable areas (or ca. 20–30%
of total Philippine land area), particularly most of central
and northern Luzon Island, north-western islands of
central Philippines (Cebu, Marinduque, Mindoro, and
Palawan Islands), and central Mindanao Island. Potential
distribution of the Asiatic painted toad includes islands
north of Luzon Island (Babuyanes and Batanes group of
islands) areas in north-central (Cordillera Administrative
Region), southern (Bicol Region) and most of central
Luzon Island, western Mindanao Island, and Sulu
Archipelago (Table 5; Fig. 6). The cane toad has occupied
almost all projected suitable areas (ca.98–100%) except
those on the islands of Batanes Province (northernmost
group of islands of the Philippines), islands of Palawan
Province (westernmost group of islands), and most of Sulu
Archipelago (southernmost islands) (Table 5; Fig. 6). Maps
showed that Chinese bullfrog has a disjunct distribution
throughout the Philippines, having occupied ca. 40–50%
of suitable areas (or ca. 35–40% of total Philippine land
area), specifically most of Luzon Island, islands of central
Table 3 Evaluation of the prediction of Species Distribution Models of the six alien frogs. Philippine
models (A) and Native range models (B) were evaluated by validation of predictions based on the Area
Under the Receiver Operating Characteristic (ROC) Curve (AUC).
Training AUC
(mean)
Test AUC
(mean)
American bullfrog
Asiatic painted toad
Cane toad
Chinese bullfrog
Green paddy frog
Greenhouse frog
B. Native range model
0.78
0.86
0.72
0.84
0.82
0.97
0.71
0.82
0.79
-
Fair
Good
Fair
Good
Fair
Excellent
0.2886
0.2663
0.0922
0.0858
0.1286
0.1527
American bullfrog
Asiatic painted toad
Cane toad
Chinese bullfrog
Green paddy frog
Greenhouse frog
0.70
0.86
0.76
0.85
0.74
0.78
0.70
0.81
0.75
0.83
0.69
-
Fair
Good
Fair
Good
Fair
Fair
0.0378
0.1243
0.1091
0.0206
0.1225
0.1823
Species
AUC values Minimum training
interpretation presence threshold
A. Philippine model
341
Island invasives: scaling up to meet the challenge. Ch 2C Other taxa: Herpetofauna
Fig. 5 Projected Philippine-suitable areas for
(from left to right) the American bullfrog
and the greenhouse frog based on Native
models.
Fig. 4 Projected Philippine-suitable areas for (from left to right) Asiatic
painted toad, cane toad, Chinese bullfrog, and green paddy frog, based
on (top row) Philippine model and (bottom row) Native model.
Philippines (Mindoro and Panay Island), and central and
eastern Mindanao Island. Most of its potential distribution
are the islands north of Luzon Island (Babuyanes and
Batanes group of islands), some areas in central Philippines
(Central Visayas Region and Eastern Visayas Region),
and most of Mindanao Island including Sulu Archipelago
(Table 5; Fig. 6). Despite being present in the Philippines
for more than a century, the green paddy frog has only
invaded ca. 40–60% of projected suitable areas or 30–40%
of the Philippines. The current distribution of the green
paddy frog is mainly in Central Philippines, southern and
central parts of Luzon Island, disjunct areas in Mindanao
Island, and Basilan Island. Potential distribution of the
green paddy frog includes most of the islands of Palawan
Province, Mindanao Island, and central to northern Luzon
Island (Table 5; Fig. 6). Lastly, due to the limited amount
of species distribution data, the current and potential
distribution of the American bullfrog and the greenhouse
frog were not estimated. Interestingly, projections show
that almost all of the Philippines is suitable for both species
(Fig. 7).
Collectively, maps showed that none of the alien frogs
has fully occupied all projected Philippine-suitable areas,
and that all alien frogs are on Luzon Island and Mindanao
Island, the two largest islands of the Philippines. The
islands of Batanes Province (Northernmost group of
islands of the Philippines) are the only remaining places in
the Philippines with no record of alien frogs.
DISCUSSION
We first discuss here how our study filled knowledge
gaps on the invasion history and the status of the alien frogs
in the Philippines. At the end of this section, we provide
policy and management recommendations.
Invasion history: conceptual background
Our study is the first to reconstruct the invasion history
of the six alien frogs in the Philippines. History of invasion
refers to the historical, demographical and geographical
features of a species’ invasion processes. This may include
information on the source of propagules and propagule
pressure, the dispersal pathways and associated vectors,
and the geographical and demographical dynamics of the
spread of the adventive populations (Dlugosch & Parker,
2008; Estoup & Guillemaud, 2010). Knowledge of the
invasion history forms the foundation of invasion biology
by addressing practical and theoretical questions as well
as testing different hypotheses concerning the ecology
and evolution underlying biological invasions (Estoup
& Guillemaud, 2010). More importantly, elucidating
invasion history can provide invaluable insights for the
Table 4 Estimates of suitable area in the Philippines (PH) for the six alien frogs. Total area
(km2) and percentage (%) of total Philippine land area that is suitable (above minimum
training presence threshold) to the alien frogs.
Species
American bullfrog
Asiatic painted toad
Cane toad
Chinese bullfrog
Green paddy frog
Greenhouse frog
342
Philippine model
(%) of total PH
km2
193,964
55.36
344,317
98.27
272,797
77.86
237,825
67.88
-
Native range model
km2
(%) of total PH
349,107
99.64
280,825
80.15
349,106
99.64
325,179
92.81
349,104
99.64
349,107
99.64
Pili, et al.: Invasion & distribution of alien frogs
Table 5 Estimate of suitable area (current distribution) in the Philippines (PH) occupied by the Asiatic painted
toad, the cane toad, the Chinese bullfrog, and the green paddy frog. Shown are total area (km2), percentage
(%) of total suitable (Minimum Training Threshold) area, and percentage (%) of PH total land area that is
occupied by the alien frogs.
Species
Asiatic painted toad
Cane toad
Chinese bullfrog
Green paddy frog
Philippine model
(%) of
Occupied
(%) total
total
suitable area
PH
2
suitable
(km )
81,262
41.90
23.19
293,061
85.11
83.64
127,185
46.62
36.30
141,927
59.68
40.51
development and implementation of sound strategies
and science-based policies for the management of
invasive alien species, particularly in preventing future
introductions and controlling incursions (Hulme, et al.,
2008; Estoup & Guillemaud, 2010; Kulhanek, et al.,
2011). Here, we reconstructed the chronological history
of invasion, identified known and potential pathways
involved in introduction, and updated the current status
and distribution of invasive frog species in the Philippines.
Below we discuss further the dynamics and mechanisms
underlying their spread based on spatio-temporal patterns
of species distribution.
Invasion history: pathways of introduction of the alien
frogs
Identifying the geographical origin, causative
pathways, and associated vectors of past introductions can
help guide the development of preventive measures, such
as monitoring and quarantine schemes, which are most
effective when specifically targeted to ports of entry and
trade of commodities associated with identified pathways
of introduction (Hulme, 2006; Hulme, 2009; Hulme, et al.,
Native range model
Occupied
(%) of
(%) total
suitable area
total
PH
2
(km )
suitable
106,894
296,938
141,797
160,700
38.064
85.057
43.606
46.032
30.51
84.75
40.47
45.86
2008). Six principal pathways are involved in the global
movement of species into new areas: alien species may
be commodities (intentionally released and escapees),
contaminants of commodities, stowaways on vectors,
opportunists exploiting corridors resulting from transport
infrastructures, or they may spread naturally (Hulme, et al.,
2008). It is noteworthy that the number of total introductions
through each pathway may vary among taxonomic groups.
For instance, global alien amphibian introductions are most
frequently through intentional release as biocontrol agent
and food source, contaminant of ornamental plant trade,
stowaway of cargo, and escapees from exotic pet trade
(Kraus, 2009). In addition to these, herein we identified
two other pathways by which alien frogs were introduced
into the Philippines: as a contaminant of agricultural trade
and aquaculture trade.
Invasion history: dynamics and mechanisms of spread
of the alien frogs
Understanding the pattern and rate of spread of
invasions are essential components of risk assessment
of invasive alien species (Stohlgren & Schnase, 2006;
Fig. 7 Current and potential distribution in
the Philippines of (from left to right) the
American bullfrog and the greenhouse frog
based on overlaid species distribution data
over Philippine-suitable areas projected by
Native models. Points indicate areas where
alien frogs were reported present (release
sites or areas where bullfrog breeding
centres were formerly established for the
Fig. 6 Current and potential distribution in the Philippines of (from left
case of the American bullfrog).
to right) the Asiatic painted toad, cane toad, Chinese bullfrog, and
the green paddy frog based on estimates of geographic range and
Philippine-suitable areas projected by (top row) Philippine models and
(bottom row) Native models.
343
Island invasives: scaling up to meet the challenge. Ch 2C Other taxa: Herpetofauna
Stohlgren & Jarnevich, 2009). For invading organisms, a
stratified diffusion process of spread “seems to be the rule
rather than the exception” (Higgins & Richardson, 1999).
In a stratified diffusion process, initial range expansion
occurs though neighbourhood diffusion and new colonies
are successively created through jump dispersal events by
long-distance migrants, accelerating the rate of overall
invasion (Van der Plank, 1967 as cited in Hengeveld,
1989; Shigesada, et al., 1995; Higgins & Richardson,
1999). Jump dispersal events are particularly common for
species that are closely associated with humans (Suarez,
et al., 2001). For instance, human-mediated jum dispersal
has been documented in Eleutherodactylus spp. in Hawaii
(Kraus & Campbell, 2002), Argentine ants (Linepithema
humile) in the United States (Suarez, et al., 2001), common
ragweed (Ambrosia artemisiifolia) in France (Chauvel, et
al., 2006).
The reconstructed history of invasion showed that
the spread of alien frogs in the Philippines followed a
stratified diffusion process wherein human-mediated jump
dispersal and neighbourhood diffusion dispersal were the
main modes of spread. Given the innate physiological
limitations of frogs to cross marine barriers and the close
affinity of alien frogs with humans (Wells, 2007), humanmediated jump dispersal is the most plausible primary
mode of dispersal of alien frogs inter- and intra-island.
Numerous jump dispersal events throughout the course
of the invasion of alien frogs in the Philippines can be
observed in the spatio-temporal distribution patterns
shown in the generated species’ distribution maps (Fig.
3). This is particularly evident in the invasion of the cane
toad, wherein from founder populations on five islands, it
has invaded almost all major islands in the Philippines in a
matter of decades (Rabor, 1952). The dispersal of the green
paddy frog to Basilan Island, some 300 km from the nearest
introduced population in Negros Island in the 1960s and
350 km from nearest native population on Borneo Island,
demonstrates a good example of either long-distance jump
dispersal or perhaps a secondary introduction event. For the
cases of the Asiatic painted toad, the Chinese bullfrog, and
the greenhouse frog, it is unclear whether their presence
on different islands is caused by jump dispersal events
from a single founder population or the result of multiple,
independent introduction events.
The same pathways implicated for alien frog
introductions may have served as the same pathways that
mediated their jump-dispersal throughout the Philippines.
Spread of the cane toad was primarily human-mediated,
being released deliberately by both government and private
individuals with the belief that the frogs would control insect
and rodent pests in agricultural fields (Merino, 1936; Rabor,
1952; Soriano, 1964). Observations in the Philippines and
on Borneo reported cane toads and the Asiatic painted toads
in cargo and vehicles of transport and trade as stowaways
(Inger, 1966; A.C. Diesmos personal observation).
The greenhouse frog may have spread throughout the
Philippines as a contaminant of ornamental plant trade and
nursery plants, as happened in Hawaii (Kraus, et al., 1999;
Olson, et al., 2012). Similarly, the propensity of the Asiatic
painted toad to seek refuge in greenhouse materials (e.g.
potted plants, soil, etc.) implicate ornamental plant trade
and movement of nursery plants as a potential pathway for
its spread (E.Y. Sy personal observation). The American
bullfrog, despite having an unresolved status in the
Philippines, was dispersed throughout the Philippines as
a food source. It was earlier speculated that the Chinese
bullfrog may have been introduced and spread throughout
the Philippines alongside the proliferation of American
bullfrog breeding centres in the 1980s (Diesmos, et al.,
2006). Moreover, agricultural trade and aquaculture trade
may have served as dispersal pathways for the Chinese
344
bullfrog and as well as the green paddy frog. Agricultural
trade has been attributed to recent range expansion of
some Philippine native species such as the Philippine
common tree frog (Polpedates leucomystax), the common
mud frog (Occidozyga laevis), and the Philippine paddy
frog (Fejervarya vittigera) (Inger, 1954; Brown, et al.,
2010). Meanwhile, the aquaculture trade served as a minor
pathway of global introduction for alien frogs and has been
well documented in some alien frogs on Guam (Christy, et
al., 2007; Kraus, 2007; Kraus, 2009).
Neighbourhood diffusion dispersal also played an
invaluable role in the spread of alien frogs within islands.
For instance, it was observed that the cane toad has diffused
up to 20 km around Dumaguete City, Negros Island in a
matter of 15 years (Rabor, 1952). Though this observation
was not supported by empirical data, in Australia, the
cane toad was observed to travel up to 1.8 km per night,
especially during the rainy months (Phillips, et al., 2006b).
Moreover, short-distance dispersal may be aided by other
“natural” processes such as extensive floods, which are
common in most parts of the Philippines.
Policy and management recommendations
Given the potential negative ecological and economic
implications of alien frogs (Kraus, 2015), policies and
management strategies for alien frog invasions in the
Philippines are urgently needed. Our study filled knowledge
gaps on the invasion of the alien frogs in the Philippines,
which can guide the development and implementation of
sound policies and management strategies, particularly the
Philippines’ National Invasive Species Strategy and Action
Plan (NISSAP; DENR-PAWB, 2013).
Prevention of future alien introductions.
Of the six alien frogs currently occurring in the
Philippines, three were introduced only in the past three
decades, with the greenhouse frog being the most recently
reported. Given the lack of measures to prevent invasions
in the Philippines, future alien frog introductions seem
inevitable. In fact, a recent survey conducted by the
authors reported a seventh alien frog is now present (A.C.
Diesmos, for future publication). A useful preventive
measure are early-warning systems (i.e., black-white lists,
watch lists, etc.). These systems direct border preventive
measures, such as inspection, quarantine, and policies
banning entry, by identifying alien species with the
potential to threaten native biodiversity (Heger & Trepl,
2003; Hulme, 2006; Maynard & Nowell, 2009). A separate
study conducted by the authors for future publication
identified alien amphibians that can potentially threaten
Philippine biodiversity based on three factors of invasion
success, namely history of invasion elsewhere, climate
match, and propagule pressure.
To prevent future alien frog introductions, preventive
measures are best focused on potential pathways and
associated vectors (Perrings, et al., 2005; Hulme, 2006;
Hulme, 2009; Hulme, et al., 2008). Some examples of
preventive measures include (1) prohibition or developing
stricter regulations and standards for the breeding, trading,
and keeping of exotic pets (e.g., Taiwan, Australia, and
New Zealand) and animals for food production (e.g.,
European Union States), (2) post-border inspection,
quarantine, and treatment of imported commodities
such as ornamental plants (e.g., Hawaii and Guam), fish
fingerlings, and agricultural products, standardising risk
assessment of candidate biocontrol species, and (3) early
detection and rapid eradication schemes at ports of entry
such as seaports and airports (reviewed in Hulme, 2009;
preventive measures focusing on alien amphibians and
reptiles are reviewed in Kraus, 2009).
Pili, et al.: Invasion & distribution of alien frogs
Management of spread between islands
Developing measures to control the inter-island spread
of alien frogs is critical in archipelagic systems, such as
the Philippines. Like prevention, measures to control the
inter-island spread of alien species are best focused on the
identified potential pathways of spread and their associated
vectors (Hulme, 2006; Hulme, 2009; Hulme, et al., 2008).
Some examples of control measures include: for the
American bullfrog, Asiatic painted toad, and the Chinese
bullfrog, prohibition of release and implementation of
standards and regulations for possession or breeding either
as pets or for farming (although no bullfrog breeding
centres are operational to date); for the Asiatic painted toad
and greenhouse frog, quarantine, inspection, and treatment
of traded and transported ornamental plants, nursery plants,
and greenhouse material; for the Asiatic painted toad and
cane toad, early detection and rapid eradication schemes on
ports of entry such as seaports and airports and inspection
of cargo; for the Chinese bullfrog and green paddy frog,
inspection of products of agricultural trade and prohibition
of fish fingerling collection for release in novel areas.
These control measures should be focused on unoccupied
but suitable islands (Leung, et al., 2005), especially in the
Batanes Island Group.
It is noteworthy that the Philippines has perhaps
the moral responsibility to contain these exotics from
spreading into neighboring foreign areas. For example,
the southernmost extent of invasion of the cane toad in
the Philippines is on Basilan Island, which is about 100
kilometers from Borneo Island (Malaysia) and where the
species is alien. The spread of the alien frogs to foreign
countries can be prevented by inspection of commodities
for export, especially those associated with pathways of
introduction and spread. In fact, the Philippine common
treefrog was introduced into Ryukyu Archipelago, Japan by
contaminated traded agricultural commodities (Kuraishi,
et al., 2009).
Maps of current and potential distribution as a guide for
management schemes.
Estimating and delineating the potential and current
geographical range of alien species is a critical component
of risk assessment by providing science-based information
that can help guide the strategic allocation of limited
resources for the management of invasive alien species
(Stohlgren & Schnase, 2006; Stohlgren & Jarnevich, 2009;
Venette, et al., 2010). For instance, surveys and monitoring
schemes should be focused in areas with no information on
the status of the alien frogs or areas of high conservation
concern (Wittenberg & Cock, 2001; McGeoch & Squires,
2015), such as in most Protected Areas in the Philippines, on
central Luzon Island (Cordillera Administrative Region),
western Mindanao Island, and islands in the Batanes
Province. More importantly, field surveys are warranted
in areas where bullfrog breeding centers were formerly
established as well as in release sites, so to confirm the
status of the American bullfrog in the Philippines (Fig. 3a).
Control and containment of incursions and mitigation of
impacts should be focused on invaded areas, especially in
areas of high conservation value such as protected areas
and nature reserves (Myers, et al., 2000b; Wittenberg &
Cock, 2001; Parrish, et al., 2003). Early detection and
rapid eradication schemes are best focused on the interface
between the potential and current distribution (Hulme,
2006), such as the invasion front of the green paddy frog
on central Luzon Island (Fig. 3e).
Recognizing the variability in projected suitable
areas between the Philippine and Native models, and that
different modelling techniques yield different results even
if calibrated with same set of data, we developed in a
separate study for future publication projections of suitable
areas based on ensembles of models fitted with data from
the entire range (native range and all invaded range/s) and
using different statistical techniques. Moreover, evaluation
of the accuracy of projections and estimates through
ground truthing are underway.
Recommendations for future research
The following recommendations for future research on
alien amphibian invasions in the Philippines are suggested:
data mining grey literature, conducting interviews, and
targeted field work to populate the assembled species
distribution database and improve reconstructed invasion
history; vector analysis of the pathways so as to understand
their importance to current and future alien amphibian
invasions; identify ‘native exotics’ and understand
their invasion histories (e.g. dynamics and mechanisms
involved in their spatial spread) and impact to ecosystems;
comprehensive risk analysis of the alien frogs, specifically
research on their ecological and socio-economic impacts;
test different hypothesis on the evolution and ecology of
alien species invasions.
ACKNOWLEDGEMENTS
We sincerely thank all museums, both local and
international, that shared their data on the six alien
frogs either directly or through the Global Biodiversity
Information Facility (GBIF); namely, the National Museum
of the Philippines, California Academy of Sciences, Florida
Museum of Natural History, National Museum of Natural
History, University of Kansas Biodiversity Institute, and the
University of Texas at Austin Biodiversity Collections. In
addition, we thank the Agriculture & Fisheries Information
Division (AFID)/ Library Section of the Department
of Agriculture, the Library Section of the Science and
Education Institute of the Department of Science and
Technology, the Library Section of the Environmental
Research and Development Bureau of the Department of
Environment and Natural Resources for allowing us to
access their information facilities. We are truly grateful
for all the institutions that provided financial support for
fieldwork, including support for ANP from the Department
of Science and Technology Science Education Institute
(Accelerated Science and Technology Human Resource
Development Program) and National Geographic Science
and Exploration Asia Young Explorers Grant (ASIA 5716), ACD from the National Museum of the Philippines,
ANP and MLD from the University of Santo Tomas, ANP,
CES, MLD, and ACD from the Department of Environment
and Natural Resources and GEF-UNEP Project 0515
“Removing Barriers to Invasive Species Management
in Production and Protection Forests in Southeast Asia”
(FORIS Project).
REFERENCES
Adraneda, C., Malit, I., Ortiz-Luis, G., Pagaduan, C. and Perez. J. (2005).
Initial Study on the Dietary Preference of Hoplobatrachus rugulosus
(Wiegmann 1834). BSc thesis. Manila, Philippines: Department of
Biological Sciences. University of Santo Tomas.
Alcala, A.C. (1957). ‘Philippine notes on the ecology of the giant marine
toad’. Silliman Journal 6(2): 90–96.
Alcala, A.C. (1986). Guide to Philippine Flora and Fauna. Vol X:
Amphibians and Reptiles. Quezon City, Philippines: Natural Resources
Management Center – Ministry of Natural Resources; Quezon City,
Philippines: University of the Philippines.
Alcala, A.C. and Brown, W.C. (1998). Philippine Amphibians. An
Illustrated Field Guide. Makati City, Philippines: Bookmark, Inc.
Alcala, A.C., Bucol, A.A., Diesmos, A.C. and Brown, R.M. (2012).
‘Vulnerability of Philippine amphibians to climate change’. Philippine
Journal of Science 141(1): 77–87.
345
Island invasives: scaling up to meet the challenge. Ch 2C Other taxa: Herpetofauna
Allouche, O., Tsoar, A. and Kadmon, R. (2006). ‘Assessing the accuracy
of species distribution models: prevalence, kappa and the true skill
statistics (TSS)’. Journal of Applied Ecology 43: 1223–1232.
Araújo, M.B. and Guisan, A. (2006). ‘Five (or so) challenges for species
distribution modelling’. Journal of Biogeography 33: 1677–1688.
Araújo, M.B. and Peterson, A.T. (2012). ‘Uses and misuses of bioclimatic
envelope modeling’. Ecology 93(7): 1527–1539.
Austin, M.P. (2002). ‘Spatial prediction of species distribution: an
interface between ecological theory and statistical modelling’.
Ecological Modelling 157: 101–118.
Bain, R., Biju, S.D., Brown, R., Das, I., Diesmos, A., Dutta, S., Gower,
D., Inger, R., Iskandar, D., Kaneko, Y., Lau, M.W.N., Meegaskumbura,
M., Ohler, A., Papenfuss, T., Pethiyagoda, R., Stuart, B., Wilkinson,
M. and Xie, F. (2008). ‘Amphibians of the Indomalayan Realm’. In:
S.N. Stuart, M. Hoffmann, J.S. Chanson, N.A. Cox, R. Berridge, P.
Ramani and B.E. Young (eds.) Threatened Amphibians of the World, pp.
74–49. Barcelona, Spain: Lynx Ediciones; Gland, Switzerland: IUCN;
Arlington Virginia, United States: Conservation International.
Beaumont, L.J., Gallagher, R.V., Thuiller, W., Downey, P.O., Leishman,
M.R. and Hughes, L. (2009). ‘Different climatic envelopes among
invasive populations may lead to underestimations of current and future
biological invasions’. Diversity and Distributions 15(3): 409–420.
Broennimann, O., Treier, U. A., Müller–Shärer, H., Thuiller, W., Peterson,
A. T. and Guisan, A. (2007). ‘Evidence of climate niche shift during
biological invasion’. Ecology Letter 10: 701–709.
Brown, R.M., Linkem, C.W., Siler, C.D., Sukumaran, J., Esselstyn, J.A.,
Diesmos, A.C., Iskandar, D.T., Bickford, D., Evans, B.J., McGuire, J.A.
and Grismer, L. (2010). ‘Phylogeography and historical demography of
Polypedates leucomystax in the islands of Indonesia and the Philippines:
Evidence for recent human-mediated range expansion?’. Molecular
Phylogenetics and Evolution 57(2): 598–619.
Brown, R.M., Diesmos, A.C., Sanguila, M.B., Siler, C.D., Diesmos,
M.L.D. and Alcala, A.C. (2012). ‘Amphibian conservation in the
Philippines’. FrogLog 20(5), Number 104: 40–43.
Buenviaje, S.V. (1983). ‘Bullfrog culture, anyone?’. Aquascope 3(4): 4.
Chauvel, B., Dessaint, F., Cardinal-Legrand, C. and Bretagnolle, F.
(2006). ‘The historical spread of Ambrosia artemisiifolia L. in France
from herbarium records’. Journal of Biogeography 33: 665–673.
Christy, M. T., Savidge, J. A. and Rodda, G. H. (2007). ‘Multiple pathways
for invasion of anurans on a Pacific island’. Diversity and Distributions
13: 598–607.
Department of Environment and Natural Resources – Protected Areas
and Wildlife Bureau (DENR–PAWB). (2013). The National Invasive
Species Strategy and Action Plan. Quezon City, Philippines: DENR–
PAWB.
Diesmos, A. C. (1998). The amphibian faunas of Mount Banahao, Mount
Cristobal, and Mount Maquiling, Luzon Island, Philippines. MSc
thesis. Laguna, Philippines: University of the Philippines–Los Baños.
Diesmos, A.C., Diesmos, M.L. and Brown, R.M. (2006). ‘Status and
distribution of alien invasive frogs in the Philippines’. Journal of
Environmental Science and Management 9(2): 41–53.
Diesmos, A.C., Alcala, A.C., Siler, C.D. and Brown, R.M. (2014). ‘Status
and Conservation of Philippine Amphibians’. In: H. Heatwole and I.
Das (eds.) Conservation Biology of Amphibians of Asia – Status of
Conservation and Decline of Amphibians: Eastern Hemisphere, pp.
310–336. Kota Kinabalu, Malaysia: Natural History Publications
(Borneo).
Diesmos, A.C., Watters, J.L., Huron, N.A., Davis, D.R., Alcala, A.C.,
Crombie, R.I., Afuang, L.E., Gee-Das, G., Sison, R.V., Sanguila, M.
B., Penrod, M. L., Labonte, M.J., Davey, C.S., Leone, E.A., Diesmos,
M. L., Sy, E. Y., Welton, L. J., Brown, R. F. and Siler, C.D. (2015).
‘Amphibians of the Philippines, part I: Checklist of the species’.
Proceedings of the California Academy of Science 62(3) Number 20:
457–539.
Dlugosch, K.M. and Parker, I.M. (2008). ‘Founding events in species
invasions: Genetic variation, adaptive evolution, and the role of multiple
introductions’. Molecular Ecology 17: 431–449.
Dormann, C.F., Elith, J., Bacher, S., Buchmann, C., Carl, G., Carré, G.,
Marquéz, J.R.G., Gruber, B., Lafourcade, B., Leitão, P.J., Münkemüller,
T., McClean, C., Osborne, P. E., Reineking, B, Schröder, B., Skidmore,
A. K., Zurell, D. and Lautenbach, S. (2013). ‘Collinearity: A review
of methods to deal with it and a simulation study evaluating their
performance’. Ecography 36(1): 27–46.
Elith, J. and Graham, C.H. (2009). ‘Do they? How do they? Why do
they differ? On finding reasons for differing performances of species
distribution models’. Ecography 32: 66–77.
Elith, J. and Leathwick, J.R. (2009). ‘Species distribution models:
ecological explanation and prediction across space and time’. Annual
Review of Ecology, Evolution, and Systematics 40: 677–697.
Elith, J., Kearney, M. and Phillips, S. (2010). ‘The art of modelling rangeshifting species’. Methods in Ecology and Evolution 1: 330–342.
346
Espiritu, L.M. (1985). Population Ecology and Distribution of Bufo
marinus L. in the Province of Isabela, Philippines. MSc thesis. Los
Baños, Laguna: University of the Philippines–Los Baños.
Estoup, A. and Guillemaud, T. (2010). ‘Reconstructing routes of invasion
using genetic data: Why, how and so what?’. Molecular Ecology 19:
4113–4130.
Fourcade, Y., Engler, J. O., Rödder, D. and Secondi, J. (2014). ‘Mapping
species distributions with MAXENT using a geographically biased
sample of presence data: a performance assessment of methods for
correcting sampling bias’. PLOS ONE 9(5): e97122.
Fritz, S., Bartholome, E., Belward, A., Hartley, A., Stibig, H.J., Eva, H.,
Mayaux, P., Bartalev, S., Latifovic, R., Kolmert, S. and Roy, P.S. (2003).
Harmonisation, mosaicing and production of the Global Land Cover
2000 database (Beta Version). Ispra, Italy: Joint Research Center (JRC).
Gaston, K.J. and Fuller, R.A. (2009). ‘The sizes of species’ geographic
ranges’. Journal of Applied Ecology 46(1): 1–9.
Gormley, A.M., Forsyth, D.M., Griffioen, P., Lindeman, M., Ramsey,
D.S., Scroggle, M.P. and Woodford, L. (2011). ‘Using presence-only and
presence-absence data to estimate the current and potential distributions
of established invasive species’. Journal of Applied Ecology 48: 25–34.
Guisan, A. and Zimmermann, N.E. (2000). 'Predictive habitat distribution
models in ecology’. Ecological Modelling 135: 147–186.
Guisan, A. and Thuiller, W. (2005). ‘Predicting species distribution:
Offering more than simple habitat models’. Ecology Letters 8: 993–
1009.
Heaney, L.R. and Mittermeier, R.A. (1997). ‘Philippines’. In: R. A.
Mittermeier, G. Robles, and C. G. Mittermeier (eds.) Megadiversity.
Earth’s Biologically Wealthiest Nations, pp. 236–255. Monterey,
Mexico: CEMEX.
Heaney, L.R., Mittermeier, R.A. and Mittermeier, C.G. (1999).
‘Philippines’. In: R.A. Mittermeier, N. Myers, G. Robles and C.G.
Mittermeier (eds.) Hotspots. Earth’s Biologically Richest and Most
Endangered Terrestrial Ecosystems, pp. 308–317. Monterrey, Mexico:
CEMEX.
Heger, T. and Trepl, L. (2003). ‘Predicting biological invasions’.
Biological Invasions 5: 313–321.
Hengeveld, R. (1989). Dynamics of Biological Invasions. Springer
Science & Business Media.
Higgins, S.I. and Richardson, D.M. (1999). ‘Predicting plant migration
rates in a changing world: The role of long-distance dispersal’. The
American Naturalist 153(5): 464–475.
Hijmans, R.J., Cameron, S.E., Parra, J.L., Jones, P.G. and Jarvis, A.
(2005). ‘Very high resolution interpolated climate surfaces for global
land areas’. International Journal of Climatology 25: 1965–1978.
Hulme, P.E. (2006). ‘Beyond control: wider implications for the
management of biological invasions’. Journal of Applied Ecology 43:
835–847.
Hulme, P. E. (2009). ‘Trade, transport and trouble: Managing invasive
species pathways in an era of globalization’. Journal of Applied Ecology
46: 10–18.
Hulme, P.E., Bacher, S., Kenis, M., Klotz, S., Kühn, I., Minchin, D.,
Nentwig, W., Olenin, S., Panov, V., Pergl, J., Pyšek, P., Roques, A., Sol,
D., Solarz, W. and Vilá, M. (2008). ‘Grasping at the routes of biological
invasions: A framework for integrating pathways into policy’. Journal
of Applied Ecology 45: 403–414.
Inger, R.F. (1954). ‘Systematics and zoogeography of Philippine
amphibia’. Fieldiana Zoology 33: 183–531.
Inger, R.F. (1966). ‘Systematics and zoogeography of the amphibia of
Borneo’. Fieldiana Zoology 52: 1–402.
Inovejas, W.P. (1985). ‘The American bullfrog: Tapped for lucrative
farming in the Philippines’. Habitat Philippines 5(1):30–44.
Jeschke, J. M. and Strayer, D. L. (2008). ‘Usefulness of bioclimatic
models for studying climate change and invasive species’. Annals of the
New York Academy of Science 1134: 1–24.
Kraus, F. (2007). ‘Using Pathway Analysis to Inform Prevention
Strategies for Alien Reptiles and Amphibians’. In: G.W. Witmer, W. C.
Pitt and K. A. Fagerstone (eds.) Managing Vertebrate Invasive Species:
Proceedings of an International Symposium, pp. 94–103. Colorado,
United States: USDA/APHIS Wildlife Services, National Wildlife
Research Center.
Kraus, F. (2009). Alien Reptiles and Amphibians. A Scientific Compendium
and Analysis (Vol. 4). Springer Science & Business Media.
Kraus, F. (2015). ‘Impacts from invasive reptiles and amphibians’. Annual
Review of Ecology, Evolution, and Systematics 46: 75–97.
Kraus, F., Campbell, E.W., Allison, A. and Pratt, T. (1999).
‘Eleutherodactylus frog introductions to Hawaii’. Herpetological
Review 30(1): 21–25.
Kraus, F. and Campbell, E.W. III (2002). ‘Human–mediated escalation
of a formerly eradicable problem: The invasive Caribbean frogs in the
Hawaiian Islands’. Biological Invasions 4: 327–332.
Pili, et al.: Invasion & distribution of alien frogs
Kulhanek, S.A., Ricciardi, A. and Leung, B. (2011). ‘Is invasion history
a useful tool for predicting the impacts of the world’s worst aquatic
invasive species?’. Ecological Applications 21(1): 189–202.
Kuraishi, N., Matsui, M. and Ota, H. (2009). ‘Estimation of the origin
of Polypedates leucomystax (Amphibia: Anura: Rhacophoridae)
introduced to the Ryukyu Archipelago’. Japan Pacific 63: 317–325.
Leung, B., Finnoff, D., Shogren, J.F. and Lodge, D. (2005). ‘Managing
invasive species: Rules of thumb for rapid assessment’. Ecological
Economics 55: 24–36.
Liu, C., Berry, P.M., Dawson, T.P. and Pearson, R.G. (2005). ‘Selecting
thresholds of occurrence in the prediction of species distribution’.
Ecography 28: 385–393.
Matienzo, L.H. Jr. (1990). ‘The bullfrog farmers of San Mateo’.
Greenfields 20(3): 8–13.
Maynard, G. and Nowell, D. (2009). ‘Biosecurity and Quarantine for
Preventing Invasive Species’. In: M.N. Clout and P.A. Williams
(eds.) Invasive Species Management. A Handbook of Principles and
Techniques, pp. 1–18. New York, USA: Oxford University Press Inc.
McGeoch, M.A. and Squires, Z.E. (2015). An Essential Biodiversity
Variable Approach to Monitoring Biological Invasions: Guide for
Countries. GEO BON Technical Series 2.
Merino, G. (1936). ‘Introduction of the giant toad, Bufo marinus, into the
Philippines’. The Philippine Journal of Agriculture 7: 283–286.
Ministry of Natural Resources (1981). Annual Report of the Ministry
of Natural Resources. Quezon City, Philippines: Ministry of Natural
Resources.
Myers, N., Mittermeier, R.A., Mittermeier, C.G., Da Fonseca, G.A. and
Kent, J. (2000a). ‘Biodiversity hotspots for conservation priorities’.
Nature 403: 853–858.
Myers, J.H., Simberloff, D., Kuris, A.M. and Carey, J.R. (2000b).
‘Eradication revisited: Dealing with exotic species’. Trends in Ecology
and Evolution 15(8): 316–320.
Olson, C.A. Beard, K.H. and Pitt, W.C. (2012). ‘Biology and impacts of
Pacific island invasive species. 8. Eleutherodactylus planirostris, the
greenhouse frog (Anura: Eleutherodactylidae)’. Pacific Science 66(3):
255–270.
Olson, C., Diesmos, A. and Beard, K. (2014). ‘Eleutherodactylus
planirotris (greenhouse frog): Philippines, Mindanao Island, Davao
City, Ladislawa Villages’. Herpetological Review 45(4): 652–653.
Orrell, T. and Hollowell, T. (2017). NMNH Extant Specimen Records.
V. 1.7. National Museum of Natural History, Smithsonian Institution.
Occurrence Dataset. <https://doi.org/10.15468/hnhrg3>. Accessed via
GBIF.org on 27 May 2017.
Parrish, J.D., Braun, D.P. and Unnasch, R.S. (2003). ‘Are we conserving
what we say we are? Measuring ecological integrity within protected
areas’. BioScience 53(9): 851–860.
Pascual, R. (1987a). ‘Bullfrogs for export: Not for meat but for laboratory
use’. Agribusiness Weekly Mr & Ms Special Edition (August 28-3): 6–8.
Pascual, R. (1987b). ‘Bullfrogs for meat (part 2): Inside the farm’.
Agribusiness Weekly Mr & M. Special Edition (September 11-17): 8–9.
Pearce, J. and Ferrier, S. (2000). ‘Evaluating the predictive performances
of habitat models developed using logistic regression’. Ecological
Modelling 133: 225–245.
Perrings, C., Dehnen-Schmutz, K., Touza, J. and Williamson, M. (2005).
‘How to manage biological invasions under globalization’. Trends in
Ecology and Evolution 20(5): 212–215.
Peterson, A.T. and Vieglais, D.A. (2001). ‘Predicting species invasions
using ecological niche modeling: New approaches from bioinformatics
attack a pressing problem’. BioScience 51(5): 363–371.
Phillips, S.J. (2005). A Brief Tutorial on Maxent. AT&T Research.
<https://biodiversityinformatics.amnh.org/open_source/maxent/
Maxent_tutorial2017.pdf>.
Phillips, S.J., Dudik, M. and Schapire, R.E. (2004). ‘A maximum entropy
approach to species distribution modeling’. Proceedings of the TwentyFirst International Conference on Machine Learning: 655–662.
Phillips, S.J. Anderson, R.P. and Schapire, R.E. (2006a). ‘Maximum
entropy modeling of species geographic distributions’. Ecological
Modelling 190(3): 231–259.
Phillips, B.L., Brown, G.P., Webb, J.K., and Shine. R. (2006b). ‘Invasion
and the evolution of speed in toads’. Nature 439: 803.
Phillips, S.J. and Dudík, M. (2008). ‘Modeling of species distributions
with Maxent: New extensions and a comprehensive evaluation’.
Ecography 31: 161–175.
R Core Team (2016). R: A language and environment for statistical
computing. Vienna, Austria: R Foundation for Statistical Computing.
<http://www.R–project.org/>.
Rabor, D.S. (1952). ‘Preliminary notes on the giant toad, Bufo marinus
(Linn.), in the Philippine Islands’. Copeia 4: 281–282.
Ripley, B., Venables, B., Bates, D.M., Hornik, K., Gebhardt, A. and Firth,
D. (2015). Support Function and Datasets for Venables and Ripley’s
MASS. R package version 7.3-45.
Sheather, S.J. and Jones, M.C. (1991). ‘A reliable data-based bandwidth
selection method for kernel density estimation’. Journal of the Royal
Statistical Society Series B 53: 683–690.
Shigesada, N., Kawasaki, K., and Takeda, Y. (1995). ‘Modeling stratified
diffusion in Biological Invasions’. American Naturalist 146 (2): 229–
251.
Soriano, P. (1964). ‘Notes on the food habits of the giant toad (Bufo
marinus Linn.) in the Philippines’. Philippine Journal of Plant Industry
29 (3-4): 79–86.
Stohlgren, J.S. and Schnase, J.L. (2006). ‘Risk analysis for biological
hazards: What we need to know about invasive species?’. Risk Analysis
26(1): 163–173.
Stohlgren, T.J. and Jarnevich, C.S. (2009). ‘Risk Assessment of Invasive
Species’. In: M.N. Clout and P.A. Williams (eds.) Invasive Species
Management. A Handbook of Principles and Techniques, pp. 19–35.
New York, United States: Oxford University Press.
Suarez, A.V., Holway, D.A. and Case, T.J. (2001). ‘Patterns of spread
in biological invasions dominated by long-distance jump dispersal:
Insights from Argentine ants’. Proceedings of the National Academy of
Sciences 98(3): 1995–1100.
Surendran, H. and Vasudevan, K. (2013). ‘Recent introduction and spread
of Indian bullfrog Hoplobatrachus tigerinus (Daudin, 1802) into the
Andaman Islands’. Aliens 33: 42–43.
Swets, J.A. (1988). ‘Measuring the Accuracy of Diagnostic
Systems’. Science 240: 1285–1293.
Sy, E.Y. and Salgo, J. (2015). ‘Geographic distribution: Eleutherodactylus
planirostris (greenhouse frog): Philippines, Cebu Island, Cebu City,
Barangay Bulacao’. Herpetological Review 46(2): 212.
Sy, E.Y., Eleazar, B.I., Achacoso, R. and Diesmos, A.C. (2015a).
‘Geographic distribution: Eleutherodactylus planirostris (greenhouse
frog): Philippines, Luzon Island, Metropolitan Manila, Quezon City’.
Herpetological Review 46(1): 56.
Sy, E.Y., Martyr, J.C. and Diesmos, A.C. (2015b). ‘Geographic
distribution: Eleutherodactylus planirostris (greenhouse frog):
Philippines, Negros Island, Negros Occidental, Bacolod City, Barangay
Villamonte’. Herpetological Review 46 (1): 56.
Sy, E.Y. (2017a). ‘Geographic distribution: Eleutherodactylus planirostris
(greenhouse frog): Philippines, Palawan Island’. Herpetological Review
48(3): 583.
Sy, E.Y. (2017b). ‘Geographic distribution: Eleutherodactylus planirostris
(greenhouse frog): Philippines, Mactan Island’. Herpetological Review
48(3): 583.
Taylor, E.H. (1920). ‘Philippine amphibia’. The Philippine Journal of
Science 16 (3): 213–359.
Taylor, E.H. (1922). ‘Additions to the herpetological fauna of the
Philippine Islands, II’. The Philippine Journal of Science 21(2): 257–
304.
Ugale, C.T. Jr. (1976). ‘Big profits from bullfrogs’. Greenfields 6(4): 4–7.
Urbanes, P.O. Jr. (1988). ‘Raising bullfrogs brings on profits by leaps and
bounds’. Animal Husbandry and Agricultural Journal 22(6): 29, 37.
Urbanes, P.O. Jr. (1990). ‘Export of cultured, preserved bullfrogs a
lucrative business’. Animal Husbandry and Agricultural Journal 24(2):
38.
Van Der Plank, J E. (1967). ‘Spread of Plant Pathogens in Space and
Time’. In: P.H., Gregory and P.A. Monteith (eds.) Airborne Microbes,
pp. 227–247. Cambridge, United Kingdom: Cambridge University
Press.
Venables, W.N. and Ripley, B.D. (2002). Modern Applied Statistics with
S. Fourth edition. Springer.
Venette, R.C., Kriticos, D.J., Magaret, R.D., Koch, F.H., Baker, R.H.A.,
Worner, S.P., Gómez Raboteaux, N.N., McKenney, D.W., Dobesberger,
E.J., Yemshanov, D., De Barro, P.J., Hutchinson, W.D., Fowler, G.,
Kalaris, T.M. and Pedlar, J. (2010). ‘Pest risk maps for invasive alien
species: A roadmap for improvement’. BioScience 60(5): 349–362.
Wells, K.D. (2007). The Ecology and Behaviour of Amphibians. Chicago,
United States: The University of Chicago Press.
Wittenberg, R. and Cock, M.J.W. (eds.) (2001). Invasive Alien Species:
A Toolkit of Best Prevention and Management Practices. Wallingford,
Oxon, United Kingdom: CABI International.
Worton, B.J. (1989). ‘Kernel methods for estimating the utilization
distribution in home-range studies’. Ecology 70(1): 164–168.
347
S.R. Siers, W.C. Pitt, J.D. Eisemann, L. Clark, A.B. Shiels, C.S. Clark, R.J. Gosnell and M.C. Messaros
Siers, S.R.; W.C. Pitt, J.D. Eisemann, L. Clark, A.B. Shiels, C.S. Clark, R.J. Gosnell and M.C. Messaros. In situ evaluation of
an automated aerial bait delivery system for landscape-scale control of invasive brown treesnakes on Guam
In situ evaluation of an automated aerial bait delivery system for
landscape-scale control of invasive brown treesnakes on Guam
S.R. Siers1, W.C. Pitt1,4, J.D. Eisemann1, L. Clark1, A.B. Shiels1, C.S. Clark2, R.J. Gosnell2 and M.C. Messaros3
USDA APHIS Wildlife Services National Wildlife Research Center, LaPorte Ave, Fort Collins CO 80521.
<shane.r.siers@aphis.usda.gov>. 2USDA APHIS Wildlife Services, Western Region, Centre Ave, Fort Collins CO
80521. 3Applied Design Corporation, Western Ave, Boulder CO 80301. 4Current affiliation: Smithsonian Conservation
Biology Institute, Remount Road, Front Royal, VA 22630.
1
Abstract After decades of biodiversity loss and economic burden caused by the brown treesnake invasion on the
Pacific island of Guam, relief hovers on the horizon. Previous work by USDA Wildlife Services (WS) and its National
Wildlife Research Center (NWRC) demonstrated that brown treesnake numbers in forested habitats can be dramatically
suppressed by aerial delivery of dead newborn mouse (DNM) baits treated with 80 mg of acetaminophen. However,
manual bait preparation and application is impractical for landscape-scale treatment. WS, NWRC, and the US Department
of the Interior have collaborated with Applied Design Corporation to engineer an automated bait manufacturing and
delivery system. The core technology is an aerially delivered biodegradable “bait cartridge” designed to tangle in the
tree canopy, making the acetaminophen bait available to treesnakes and out of reach of terrestrial non-target organisms.
When mounted on a rotary- or fixed-wing airframe, the automated dispensing module (ADM) unit can broadcast 3,600
bait cartridges at a rate of four per second and can treat 30 hectares of forest at a density of 120 acetaminophen baits
per hectare within 15 minutes of firing time. We conducted the first in situ evaluation of the ADM in July 2016. Initial
acetaminophen bait deployment rates (proper opening of the bait cartridge for canopy entanglement) were low, and
mechanism jams were frequent due to internal friction and wind forces; on-site remedial engineering improved these
performance measures. Bait cartridge placement and spacing were accurate (average 8.9 m along 9 m swaths) under
various flight heights and speeds. Canopy entanglement of properly-deployed acetaminophen baits was high (66.6%).
Although only a small proportion (5.9%) of radio transmitter-equipped acetaminophen baits were confirmed to have been
taken by brown treesnakes, the baiting density was high enough to make it likely that a significant proportion of brown
treesnakes in the area had taken acetaminophen baits. With subsequent improvements in system reliability, the automated
bait cartridge manufacturing and delivery system is poised to transition from research and development to operational
field implementation. Applications include reduction of brown treesnake numbers around transportation infrastructure
and within core habitats for the reintroduction of native birds extirpated by this troublesome invasive predator.
Keywords: invasive species suppression, invasive vertebrate predator, public-private partnership, scaling up, technical
innovation, toxic baits
INTRODUCTION
The brown treesnake (Boiga irregularis) is a nocturnal,
arboreal predator that was probably introduced on the
island of Guam after World War II as a passive stowaway
in cargo from the Admiralty Islands north of New Guinea
(Rodda & Savidge, 2007; Richmond, et al., 2014). Lacking
natural predators on Guam, the population of brown
treesnakes irrupted, reaching as many as 50–100 brown
treesnakes per hectare in some areas (Rodda, et al., 1999).
Brown treesnakes colonised the entire island of Guam
(54,930 ha) in about 20–30 years (Savidge, 1987). The
brown treesnake has been – and continues to be – a threat
to the economy and ecology of Guam, and is currently
the subject of a cooperative programme to control brown
treesnake populations on the island and prevent its spread
throughout the Pacific Basin and other vulnerable locations
(Clark, et al., 2018). Owing to the significant ecological
and economic damages caused by the brown treesnake on
Guam, the potential for the brown treesnake to be spread to
other Pacific Islands, including Hawai`i, is of great concern
(Shwiff, et al., 2010).
Landscape-scale suppression of brown treesnakes is
desirable in habitats adjacent to transportation network
infrastructure (e.g., cargo terminals), to reduce the risk
of accidental transport to other vulnerable ecosystems,
and within key habitats for the recovery of Guam’s native
wildlife. Because of the great amount of inaccessible and
topographically challenging forest habitat on Guam, aerial
delivery of brown treesnake suppression tools is key to the
management of this species on a landscape scale. Dead
newborn mice (DNM) dosed with 80 mg of acetaminophen
have proven to be safe and effective baits for lethal control
of brown treesnakes (Savarie, et al., 2001; Johnston, et al.,
2002; Clark, et al., 2012) and are registered with the US
Environmental Protection Agency (EPA) as an approved
pesticide (Registration No. 56228-24, Revised 06/2018).
To be effectively delivered to the forest canopy where they
are available to foraging brown treesnakes and inaccessible
by terrestrial non-targets, the baits must be coupled with a
‘flotation device’ intended to entangle in foliage (Savarie
& Tope, 2004).
Through a previous project, the US Department of
Agriculture (USDA) Animal and Plant Health Inspection
Service (APHIS) Wildlife Services National Wildlife
Research Center (NWRC) has demonstrated that brown
treesnake abundance in Guam’s forests can be suppressed
via the aerial application of DNM baits adhered to paper
streamers (Dorr, et al., 2016). During this prior study, baits
were hand-prepared and hand-broadcast from a helicopter.
While this method of treatment proved effective on a
small scale (two 55 ha plots), manual bait preparation
and application is economically impractical for larger
landscape-scale treatments. In scaling up to meet the
challenge of landscape-scale control of brown treesnakes,
one of the principal logistical concerns is the obvious need
to automate both bait production and the aerial dispensing of
baits. In response to this need, NWRC, primarily funded by
the US Department of the Interior Office of Insular Affairs,
has partnered with a private, small business engineering
company (Applied Design Corporation, Boulder, Colorado)
to develop a brown treesnake suppression system that offers
the capability to achieve precise distribution of thousands
of baits in a matter of minutes, through a fully-integrated
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
348
up to meet the challenge, pp. 348–355. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Siers, et al.: Automated bait system for treesnakes on Guam
solution that encompasses bait cartridge production, an
aerial bait cartridge dispensing system, and supporting
infrastructure and logistics for practical manufacturing,
storing, and flight-line handling of bait cartridges.
Automated Bait Manufacturing System (ABMS)
Many of the functional details of the three-stage
ABMS are currently considered proprietary information
pending application for US and foreign patent protection.
The descriptions provided below will suffice as a basic
functional explanation.
The first of three bait cartridge manufacturing stations
is the Gluer/Placer Station (Station 1) where the DNM
are distributed on moulded pulp paper trays and an 80 mg
acetaminophen tablet is adhered to the DNM via a hotmelt adhesive. At the final stage of Station 1, the individual
capsules containing the acetaminophen tablet and DNM
are cut from the paper trays and fed into a transport cassette
for transfer to the Assembly/Winder Station (Station 2).
Hereafter, a DNM with an adhered acetaminophen tablet
will be referred to as an “acetaminophen bait”.
Automated Dispensing Module (ADM)
The Automated Dispensing Module (ADM; Fig. 2) is
comprised of three main components: 1) four magazines;
2) an electro-mechanical firing unit on a tilt-plate; and 3) a
frame, which holds the power supply battery, the computer
control module, and integrates the other components into
a single functional ADM. The frame is mounted within the
hold of the aircraft.
Each magazine is comprised of a body with two halves
hinged at the back, allowing the payload area to be fully
exposed, and a faceplate. The opened magazine receives
the contents of one case (900 bait cartridges). Upon
loading, the bait cartridges receive a final inspection for
manufacturing imperfections or shipping damage which
may adversely affect smooth feeding through the magazine
and into the firing unit (Fig. 3).
At Station 2, the capsule is folded and held closed
by pinching at the paper hinge between the two capsule
halves. This pinched paper hinge, hereafter referred to as
the ‘tang,’ is inserted into a slotted pressed pulp paper end
cap. One end of a biodegradable plastic ribbon is adhered
to the endcap and the entire assembly is rotated until the
ribbon is wound around the length of the capsule in a
‘barber pole’ fashion. The terminal end of the ribbon is
then adhered to the paper capsule. An exterior cardboard
tube is placed over the wound assembly, with the end cap
tightly pressed into the tube; this entire resulting assembly,
comprised of the acetaminophen bait, capsule, streamer,
and end cap, enclosed within the external tube, is referred
to as a “bait cartridge”. The entire bait cartridge (Fig. 1) is
biodegradable.
The final manufacturing station is Packaging (Station
3). Completed bait cartridges are automatically fed to the
packaging station, where they are gathered and placed into
a corrugated plastic case (900 bait cartridges per case).
Filled cases are shrink wrapped, placed on a shipping pallet,
and frozen. A complete pallet of 40 cases holds 36,000
bait cartridges, enough to treat 300 ha at the current EPAapproved maximum application rate of 120 acetaminophen
baits/ha.
Fig. 1 When deployed, the bait capsule and outer bait
cartridge tube are joined by a length of unfurled ribbon
intended to entangle in the forest canopy when applied
aerially.
Fig. 2 The ADM is comprised of four firing units and four
900-cartridge magazines along with an onboard battery
and control electronics (not visible).
Fig. 3 Bait cartridges are inspected for manufacturing
imperfections or shipping damage that might impede
smooth feeding and ejection.
349
Island invasives: scaling up to meet the challenge. Ch 2C Other taxa: Herpetofauna
A magazine can be loaded with a case of bait cartridges
and prepared for flight in two to three minutes. Once four
magazines are prepared, they are loaded into the aircraftmounted ADM frame (Figs. 4 and 5). The bait cartridge
exit door on each magazine is then opened, allowing
bait cartridges to flow into the firing unit feed chute. A
full payload of 3,600 bait cartridges is sufficient to treat
30 ha of forest at the maximum application rate. At full
performance, this area can be treated at 120 acetaminophen
baits/ha within 15 minutes of firing time (Fig. 6). An
additional set of magazines allows for the next payload to
be prepared while the current payload is being applied.
A payload manager and the pilot are the only personnel
aboard the aircraft. As directed by the payload management
software, the computer control module engages the firing
units within the ADM to fire bait cartridges at the proper
rate to match the aircraft’s current ground speed and
intended acetaminophen bait application rate. The payload
management software detects when a port is jammed or a
magazine is empty and increases the firing rate of the other
three ports to maintain the desired bait cartridge delivery
rate.
Aerial navigation is achieved by following a
preprogrammed mission plan in the payload management
software, which details the transects to be flown. The pilot
is provided with an LCD display, a “virtual lightbar,” that
provides realtime feedback as to whether the aircraft is on
the prescribed flight path and what corrective movements
are needed to return to the path. The payload manager
manually toggles on bait cartridge firing when over the
treatment area, and toggles it off when the flight path is
complete. After an ‘ag turn’ (an aerial maneuver to quickly
reverse directions) the next flight path is flown in the
opposite direction. This is repeated until the payload is
expended or the treatment area has been fully covered.
Objectives
This report describes the first in situ evaluation of this
system through the experimental treatment of 110 hectares
of forest on Guam. The major objectives were to evaluate:
1) the ground support work flow and performance of the
automated dispensing module in-flight; 2) the precision of
spatial coverage of the treatment area; and 3) the proper
deployment of bait cartridges into the forest canopy and
the fate of acetaminophen baits once distributed into the
environment.
MATERIALS AND METHODS
Study site
The evaluation was conducted over 110 ha of
secondary forest on the Marbo Annex of Andersen Air
Force Base (typically referred to as “Andy South”) in
Yigo, Guam, at approximately 13.508°N, 144.873°E. This
site was selected because: 1) there is low risk to threatened
or endangered species; 2) the habitat is representative of
much of Guam’s forests; and 3) it is on a closed military
facility with restricted public access.
ADM performance
We assessed the performance of the ADM through
two trial applications of acetaminophen baits, simulating
operational applications for brown treesnake control. The
first application was initially scheduled to be completed
on 19 July 2016, during which 13,200 acetaminophen
Fig. 4 Magazines are loaded into the helicopter-mounted
ADM frame.
Fig. 5 Complete ADM with loaded magazines mounted in a
McDonnell-Douglass MD 500D helicopter.
350
Fig. 6 Bait cartridges dispensed in flight.
Siers, et al.: Automated bait system for treesnakes on Guam
baits would be applied over the 110 ha treatment area
(120/ha). A second application was scheduled to occur
three days later. For the purposes of this report, we
define an “application” as a treatment of an area with
aerially-distributed acetaminophen baits within the usage
restrictions described in the EPA label.
bait was properly deployed and the DNM available for
take by a brown treesnake, whether the acetaminophen
tablet was still adhered to the mouse, and other notes about
the circumstances of the condition and location of the
acetaminophen bait and its availability for take by a brown
treesnake.
A McDonnell-Douglas MD 500D (Fig. 5) and pilot were
contracted from Hansen Helicopters (Tamuning, Guam) to
perform aerial bait cartridge delivery. GoPro video cameras
(GoPro, Inc., San Mateo, California) were positioned
at various locations on the helicopter to document and
evaluate bait cartridge ejection and deployment success.
If a DNM became separated from the bait cartridge and
was on the ground but still had the acetaminophen tablet
attached, it was considered intact and available for take by
a brown treesnake. If the acetaminophen bait did not deploy
properly and the DNM was not available to be taken, the
bait cartridge and transmitter were recovered and that trial
was ended. After deployment-day data were collected, the
transmitter-equipped baits were left to determine the fate
of acetaminophen baits over the next 48–72 hours. On each
day following the application, each transmitter was relocated and the following data were recorded: whether the
acetaminophen bait was still present and viable, whether
the acetaminophen tablet was still attached, whether the
acetaminophen bait was consumed by a brown treesnake
or a non-target, whether the brown treesnake or non-target
was alive or dead, whether the transmitter had moved to
a new location, and other notes about acetaminophen bait
location and condition.
On the night prior to flight operations, bait cartridge
cases required for the next day’s application were removed
from the freezer to thaw and were stored overnight in an
air-conditioned workspace to minimise condensation. The
plastic wrapping on the cases were left intact to ensure that
all condensation would occur on the external surface of the
plastic wrap rather than on the paper-based bait cartridges
themselves.
Bait cartridge coverage
Bait cartridge spacing trials were conducted to
determine the accuracy and evenness of bait cartridge
distribution at varying flight heights and airspeeds. Three
lanes of approximately 200 m were delineated with orange
traffic cones within an open grassy area at the treatment
site. The helicopter, traveling at 50 knots, distributed bait
cartridges along each flight line at heights of 25 m, 50 m,
and 100 m above ground level. A ground crew attempted
to locate all bait cartridges and measured their distance
from the ideal flight path and the distance to the next bait
cartridge along that path. A second round of transects was
flown, this time at 60 knots, to determine the effect of
airspeed on accuracy and spacing.
The completeness and the evenness of the spatial
coverage of the treatment area was determined by recording
the GPS flight paths in the payload management software,
and generating coverage maps. Flight path segments were
highlighted where the ADM unit was firing.
Acetaminophen bait fate
Methods for monitoring of radio transmitter-equipped
baits were modified from procedures established by Dorr,
et al. (2016). During each treatment, a subset of baits was
prepared containing small 1.0 g VHF radio transmitters
(Holohil BD-2H with internal helical antennae, Holohil
Systems Ltd., Carp, Ontario, Canada) implanted in the
acetaminophen bait DNM. Transmitter-equipped bait
cartridges were placed directly in the ADM firing port
unit so that they would be deployed simultaneously at
the beginning of the flight path, to be followed by bait
cartridges without transmitters.
An acetaminophen bait is considered properly
“deployed” when the inner capsule assembly slides
out of the outer cardboard tube, unfurling the ribbon to
allow entanglement in the forest canopy. While some
acetaminophen baits may deploy on impact with treetops,
the system is designed for the acetaminophen bait to deploy
in the air immediately upon ejection of the bait cartridge
from the ADM.
Immediately after being aerially distributed, field
technicians with handheld VHF receivers located the
transmitter-equipped baits and recorded: bait cartridge
location, position (in tree/vegetation or on ground), type of
vegetation the bait cartridge was suspended from, height
above ground, whether the bait cartridge was actually seen
or its location was estimated, whether the acetaminophen
If acetaminophen baits were unconsumed and still
viable, they were left for another night and located again
the next day. If acetaminophen baits had been consumed by
a brown treesnake or non-target that was still alive, it was
left undisturbed and relocated daily to establish survival
or time to death. While tracking transmitters, technicians
were alert for carcasses of any dead organisms, including
those that had ingested transmitter-equipped baits.
Global Positioning System (GPS) locations and notes on
the location and condition of carcasses were recorded.
Carcasses were collected and stored frozen for future
analytical chemistry to verify acetaminophen exposure.
RESULTS
ADM performance
The first application of bait cartridges commenced
on schedule on 19 July 2016. Ground operations and
logistical support proceeded according to plan. However,
crew and video observations indicated poor ADM
performance in two primary categories: 1) bait cartridge
feed/ejection reliability and 2) percentage of bait cartridges
properly deploying in flight. These problems with system
performance resulted in frequent flight stoppages to
resolve bait cartridge jams and address other engineering
challenges. As a result, additional flight days on 20, 22, 23,
25, and 26 July were required to achieve the first complete
coverage of the treatment area.
Reliable bait cartridge ejection was hampered in three
primary manners: 1) mechanical jams in the firing unit; 2)
“starvation” of the firing unit feed ramp (bait cartridges not
arriving at the firing position from the magazine); and 3)
impediment of ejection by aerodynamic forces. These are
not distinct processes, with multiple possible interactions
among them. These issues were resolved with a variety of
on-the-fly field improvements, with the causes and effects
noted for future ADM design improvements.
Acetaminophen baits that do not deploy from the bait
cartridge constitute a waste of resources (because the
toxicant is inaccessible to snakes) and a fruitless toxic
input into the environment. While we did not expect 100%
deployment, observations by ground crew and video
camera evidence indicated that initial acetaminophen bait
deployment rates were unacceptably low at far less than
50%. Acetaminophen bait deployment issues generally fell
351
Island invasives: scaling up to meet the challenge. Ch 2C Other taxa: Herpetofauna
into two categories: inadequate rotational energy imparted
by the firing unit to overcome external air resistance effects
and internal friction between the sliding components of the
bait cartridge.
Air resistance effects were largely mitigated by
employing adjustable baffles near the firing unit ejection
ports to disrupt ejection-inhibiting air currents. Internal
friction effects were traced to excessive friction between the
bait cartridge capsule ‘tang’ and end cap. As manufactured,
the tang (folded paper hinge) of the interior clamshell is
seated in the slot of the end cap to prevent rotation of the
internal assembly during manufacturing and unwinding of
the ribbon during shipping and handling. However, it was
discovered that the tension of the ribbon wound around
the clamshell capsule caused the inner assembly to rotate
slightly and the tang to twist against the sides of the slot
in the end cap. This friction, along with the taut wind of
the ribbon, created a ‘locking’ force, holding the entire
assembly together and resisting the available centrifugal
force which would otherwise deploy the acetaminophen
bait properly. We determined that tearing off the paper tang
would relieve the friction against the end cap slot, greatly
increasing the deployment rate. For the second application,
all bait cartridges were prepared by manual removal of the
paper tang.
flight speed (60 knots) and highest flight height (100 m)
resulted in an acceptable distribution pattern. Spacing
between bait cartridges along a given flight path was
highly variable, but the mean overall spacing of 8.9 m was
virtually identical to the target spacing of 9 m.
Due to frequent flight stoppages during the first
application, the full site coverage was achieved piecemeal
over several days, with the entire area being treated by
the 6th flight day. While the appropriate number of bait
cartridges was deployed, the evenness of transect spacing
was of reduced importance compared to overcoming
the engineering challenges. The second application of
acetaminophen baits was relatively uninterrupted. Flight
paths were flown as planned which, along with increased
pilot and payload manager experience, resulted in a much
more even treatment (Fig. 8).
Acetaminophen bait fate
On 26 July 2016, 28 transmitter-equipped baits were
broadcast over the treatment site. On 29 July 2016, an
additional 23 were broadcast, for a total of 51 transmitterequipped baits. The conditions of acetaminophen baits on
the day of deployment are summarised in Table 1. Of the
After system modifications were made, the second
application was re-scheduled for 29 July 2016 (three days
after the completion of the first application in accordance
with EPA label restrictions). During this application, bait
cartridge ejection and acetaminophen bait deployment
were far more reliable. Bait cartridge jams in firing ports
were less frequent and were promptly cleared. The only
significant delay occurred when an ejector unit bearing
broke; a temporary bushing replacement was fabricated
and the ADM was returned to service within a few hours.
Aside from this stoppage, the entire second application
was completed within 2.5 hours.
Even after the above-mentioned modifications, only
37.3% of acetaminophen baits (571 out of a sample of
1,528 bait cartridge ejections observed on video) deployed
immediately, as intended. Bait cartridges could only
reliably be observed for about a third of their trajectory
to the canopy, and some certainly deployed lower in
the airstream. Still more would have deployed upon
impact with the canopy or the ground. Nonetheless, we
determined that improvement is needed in the reliability
of aerial deployment of acetaminophen baits. Though there
is no way to be certain of the actual deployment rates, we
presume the realised acetaminophen bait deployment rate
to be <50% for the overall acetaminophen bait application
period.
Fig. 7 Bait cartridge spacing and placement results.
The centre line for each flight path indicates the target
line, over which the pilot flew and bait cartridges were
dispensed. Green boxes around the centre lines indicate
4.5 m on each side of the centre line, for a 9 m swath
(the ideal flight path spacing for applications at 120
baits/ha). “Spacing” is the average distance from one
bait cartridge to the next one along the flight path (target
spacing was 9 m).
Bait coverage
Bait cartridge placement and spacing was tested on 28
July 2016. Wind conditions during all flights were recorded
at 0 to 1 on the Beaufort scale (0 = < 1 km/h, calm, smoke
rises vertically; 1 = 1–5 km/h, light air, wind motion visible
in smoke). When air movement was detectable, it was
moving north to north-northwest. Flight direction was west
to east or east to west. Bait cartridge distributions over trial
flight paths are depicted in Fig. 7.
Placement along target flight paths and within 9-m
swaths was very accurate and consistent. The one exception
was the run at 100 m flight height at 50 knots airspeed; these
results are inconsistent with the other five, and we consider
this to be an anomalous lapse in pilot flight accuracy.
Results do not appear to be influenced by the difference
between 50 and 60 knots airspeed. Likewise, accuracy of
placement along paths did not appear to be influenced by
flight height. The most challenging combination of higher
352
Fig. 8 Flight paths from the second bait application. Green
swaths (portion of the flight paths where bait firing was
actuated) are depicted at 9 m width, the optimal bait
cartridge spacing for the 120 acetaminophen baits/ha
application rate.
Siers, et al.: Automated bait system for treesnakes on Guam
51 transmitter-equipped baits, 92.2% deployed from the
bait cartridges, with the acetaminophen baits available to
brown treesnakes.
Thirty-four bait cartridges (66.6%) tangled in the canopy
as intended (Fig. 9). Thirteen (25.5%) were on the ground,
but open and available to be taken by ground-foraging
brown treesnakes. Four (7.8%) were not deployed (closed)
on the ground, making the bait and toxicant unavailable
to the brown treesnake. During each application, one
transmitter-equipped cartridge was in the canopy but
could not be confirmed to have deployed; we consider it
unlikely that an unopened bait cartridge would be caught
in the canopy, so assumed that these acetaminophen baits
deployed.
Of the 47 opened bait cartridges, two from each
application had the acetaminophen tablet detached from
the DNM, making it an ineffective acetaminophen bait.
In total, 41 of the 51 acetaminophen baits (80.4%) had
acetaminophen tablets attached to the DNM and were
available for take by a brown treesnake. This should be
viewed as the overall successful bait deployment rate for
this sample of baits.
Of the 51 transmitter-equipped baits, canopy height
and acetaminophen bait height data were available on 33
acetaminophen baits that successfully deployed (18 from
Application 1 and 15 from Application 2). The hanging
height of the acetaminophen bait with respect to the
canopy height is represented graphically in Fig. 10. In a
linear regression, canopy height and acetaminophen bait
height were significantly correlated (p << 0.001, adjusted
R2 = 0.694; the four bait cartridges on the ground were not
included in the regression). These results show that the
majority of deployed acetaminophen baits were entangled
within a few metres of the top of the canopy.
Deployed and intact acetaminophen baits were rechecked daily, with very few confirmed takes by brown
treesnakes or non-target organisms (Table 2). Of the 51
transmitter-equipped baits, three (5.9%) were confirmed
by visual sighting to have been taken by brown treesnakes,
or 7.3% of the 41 transmitter-equipped baits known to
be available and intact. The 95% binomial confidence
interval (logit parameterisation) for the estimated take
rate of 5.9% is 1.9% to 16.7%; given the small number
Table 1 Status of transmitter-equipped bait cartridges
following ejection from ADM. “Deployed” means the
inner capsule completely exited the outer tube and the
acetaminophen bait was available for take by a brown
tree snake. “Intact” means the acetaminophen tablet
was still attached to the bait mouse and available to be
taken by a brown treesnake.
Fig. 9 Desired canopy entanglement and acetaminophen
bait exposure.
Bait
cartridge
status
Opened in
canopy*
Opened on
ground
Not
deployed
Unknown
Total
deployed*
Total known
deployed
and intact**
Application Application TOTAL
1 (n=28)
2 (n=23)
(n=51)
19 (67.9%)
15 (65.2%)
34 (66.6%)
7 (25.0%)
6 (26.1%)
13 (25.5%)
2 (7.1%)
2 (8.7%)
4 (7.8%)
1 (3.6%)
1 (4.3%)
2 (3.9%)
26 (92.3%)
21 (91.3%)
47 (92.2%)
23 (82.1%)
18 (78.3%)
41 (80.4%)
*Assumes that “unknown” bait cartridges in canopy were
deployed; **Does not assume “unknown” baits were intact.
Table 2 Transmitter-equipped acetaminophen baits taken
by target (brown treesnake) or non-target species.
Species
Fig. 10 Hanging height of the bait cartridge (y-axis)
in relation to the height of the canopy at that location
(x-axis).
Brown tree snake
Monitor lizard
Marine toad
Unknown
Application Application Total
1
2
1
2
3
1*
0
1
0
2*
2
0
1
1
*Transmitter recovered in faeces
353
Island invasives: scaling up to meet the challenge. Ch 2C Other taxa: Herpetofauna
of acetaminophen baits equipped with transmitters, the
actual rate of acetaminophen bait take by brown treesnakes
could vary widely. All three transmitters were regurgitated
prior to death, so no transmitters were recovered in brown
treesnake carcasses. All three transmitters taken by nontargets were later found in faeces; it is unclear whether any
of these animals succumbed to acetaminophen toxicosis.
in unreliable bait cartridge counts as tabulated by the ADM
onboard software. We ensured that EPA label application
rate restrictions were not exceeded by confirming that no
more than 14.66 cases (13,200 bait cartridges) were applied
throughout the treatment area during each application
period.
All vertebrate carcasses encountered during field
activities were collected. This included three brown
treesnakes and one marine toad (Rhinella marina).
Acetaminophen bait fate
DISCUSSION
ADM performance
Future improvements to the ADM will focus on:
baffling of the airstream around the ejector ports to
prevent interference with ejection; improved feeding of
bait cartridges from redesigned magazines and; increased
energetic impact imparted to the bait cartridge at the instant
of firing in order to improve ejection and deployment
reliability. Engineering modifications to the ABMS
will further address the non-deployment issue through
tighter quality control on bait cartridge imperfections and
abatement of tang friction through a redesigned end cap.
Bait cartridge coverage
The accuracy of bait cartridge placement along flight
lines was encouraging. There was very little air movement
during these trials; under windier conditions, bait cartridges
distributed from greater heights will be more likely to drift
further from the intended flight path.
We attribute high variability in bait cartridge spacing
along the flight lines to variability in the times at which
acetaminophen baits deployed after being ejected from the
ADM. When the acetaminophen bait deploys, wind drag
increases greatly and the forward momentum is quickly
attenuated, causing the bait cartridge to drop straight
down. Acetaminophen baits that deploy later maintain
forward momentum longer and will move farther along the
flight path before landing. It is expected that bait cartridge
modifications that improve acetaminophen bait deployment
will also result in less variability in time of deployment,
leading to more consistent spacing along flight paths.
Variability in spacing along the flight path does have
the potential to affect bait cartridge placement accuracy
at the edges of treatment areas where bait cartridge
application begins and ends, potentially leading to a
small number of bait cartridges landing outside of the
desired treatment area. To make up for the inconsistency
of bait cartridge density at these edges, it is advisable that
another application flight should occur along these edges,
perpendicular to the original flight paths, ensuring that the
edges get a full treatment in a more controlled fashion,
similar to coastal aerial rodenticide applications during
island rodent eradications.
Variability in bait cartridge placement along and
perpendicular to the flight path will add apparently random
“noise” to the locations, as opposed to placing bait cartridges
precisely on an idealised 9 x 9 m grid. This variability will
not affect the ability to get acetaminophen baits into the
movement areas of every brown treesnake. The greatest
risk of gaps in coverage might arise from strong changes in
wind direction, which might introduce strong biases in bait
cartridge drift patterns. This will likely factor in with other
considerations leading to recommendations not to apply
baits during high wind conditions.
Wind effects at bait cartridge ejection ports and direct
sunlight on the bait cartridge counter photogates resulted
354
The proportion of transmitter-equipped baits taken by
brown treesnakes was low (5.9%); however, only a very
small portion of the bait cartridges distributed (0.19%)
were equipped with transmitters. If we assume that half of
the acetaminophen baits applied during both applications
properly deployed and were viable, then there were 13,200
acetaminophen baits available for take by brown treesnakes.
If 5.9% of those acetaminophen baits were taken by brown
treesnakes, we would expect approximately 779 brown
treesnakes to have taken an acetaminophen bait. If we
assume a density of 25 brown treesnakes per hectare in this
area (a conservative estimate based on the 25-50/ha range
reported by Rodda, et al. 1999), 2,750 brown treesnakes
would have been exposed to the treatment. If 779 brown
treesnakes took acetaminophen baits and were killed, this
would be a brown treesnake mortality of approximately
28% in what was effectively a single treatment (given
the low deployment rate). The three acetaminophen baits
visually confirmed to have been taken by brown treesnakes
were found on the ground, apparently regurgitated. In
previous NWRC lab efficacy trials of acetaminophen
baits with acetaminophen tablets internally-implanted in
the DNM (rather than glued to the exterior), 26% were
regurgitated, but 100% of the caged brown treesnakes that
regurgitated the acetaminophen bait died within 12 to 36
hours (Savarie, 2002). Based on that result, it is reasonable
to assume that the brown treesnakes that had taken and
regurgitated acetaminophen baits with transmitters in this
study also died.
With respect to deployment and entanglement rates,
caution should be taken in considering transmitterequipped cartridges to be representative of the standard
bait cartridges distributed during this evaluation. Machineassembled bait cartridges were manually unwound and
rewound by hand after the implantation of the radio
transmitter; this may explain why transmitter-equipped
cartridges deployed at a higher rate than those observed
on video. The added mass of the transmitter may also have
an effect on the forces exerted on various parts of the bait
cartridge and acetaminophen bait assembly. However,
it is also possible that unopened bait cartridges without
transmitters actually did deploy lower in the air column
(out of view of the video cameras) or upon impact with
the canopy.
The overall reduction of brown treesnake abundance in
the treatment area – as inferred from a foraging activity
index based on take rates of non-toxic DNM from bait
stations – is currently being monitored as a separate study
for future publication.
CONCLUSION
Upon firing from the ADM, bait cartridge ejection
and acetaminophen bait deployment reliability was
initially low. Performance was improved dramatically
with field-improvised remedial measures. It is estimated
that <50% of acetaminophen baits deployed from the bait
cartridges, resulting in an under-treatment compared to the
target application rate of 120/ha. Canopy entanglement
of acetaminophen baits that properly deployed was
high. Aerial bait cartridge placement and spacing were
satisfactorily accurate. Reliability of bait cartridge ejection
Siers, et al.: Automated bait system for treesnakes on Guam
and acetaminophen bait deployment will be a critical focus
of bait manufacturing and delivery system improvements,
increasing per-cartridge effectiveness. Future advancements
of this technology may include adaptation for payload
management by the pilot alone, incorporation of a longerlasting artificial bait to replace the DNM, and increases in
ejector unit and magazine capacity for greater payloADM.
With this evaluation – and subsequent improvements in
system reliability – we consider the concept of automated
bait production and aerial delivery to be fundamentally
sound. For the first time in the decades-long saga of
the brown treesnake invasion of Guam, the prospect of
landscape-scale suppression hovers on the horizon.
DISCLAIMER
The use of trade or corporation names within this report
is for the convenience of the user in identifying products.
Such use does not constitute an official endorsement
or approval of any product by the U.S. Department of
Agriculture.
ACKNOWLEDGEMENTS
We wish to thank the DOI Office of Insular Affairs
for funding the development of the ABMS and ADM
technologies, and for funding this evaluation. Additional
funding for this evaluation was provided by US Navy, Joint
Region Marianas. The ABMS and ADM systems were
conceived and built by ADC engineers Jonathan Fragoso,
Shane Vogt, Grady Barfoot, and Michael Messaros. Bill
Coon engineered the ADM software and flew as the
navigator and payload manager. Hansen Helicopters
general manager (Rufus Crowe), pilot (Dan O’Brien), and
service shop staff were instrumental in the success of this
evaluation. Methodology for acetaminophen bait fate via
radio transmitters was established by Brian Dorr (Dorr, et
al., 2016). Logistics and field support were provided by
Francine Chlarson, Derek Hendricks, Jerome Larimer,
Joe Rabon, Anthony Thompson, and Rachel Volsteadt.
Additional site support on flight days was provided by
Ray Quichocho, Rico Terrazas, and Marc Hall. Regulatory
and permitting assistance were provided by Shannon
Hebert, James Watkins, Earl Campbell, Diane Vice, Jim
McConnell, and the Guam Environmental Protection
Agency.
REFERENCES
Clark, L., Savarie, P.J., Shivik, J.A., Breck, S.W. and Dorr, B.S. (2012).
‘Efficacy, effort, and cost comparisons of trapping and acetaminophenbaiting for control of brown treesnakes on Guam’. Human–Wildlife
Interactions 6: 222–236.
Clark, L., Clark, C.S. and Siers, S.R. (2018). ‘Brown tree snakes:
methods and approaches for control.’ In: W.C. Pitt, J.C. Beasley and
G.W. Witmer (eds.) Ecology and Management of Terrestrial Vertebrate
Invasive Species in the United States, pp. 107–134. New York: Taylor
& Francis.
Dorr, B.S., Clark, C.S. and Savarie, P. (2016). Aerial application of
acetaminophen-treated baits for control of Brown treesnakes (RC200925; NWRC study number: QA-1828). Environmental Security
Technology Certification Program Final Report, Virginia, USA:
Department of Defense, Alexandria.
Johnston, J.J., Savarie, P.J., Primus, T.M., Eisemann, J.D., Hurley, J.C.
and Kohler, D.J. (2002). ‘Risk assessment of an acetaminophen baiting
program for chemical control of brown tree snakes on Guam: Evaluation
of baits, snake residues, and potential primary and secondary hazards’.
Environmental Science and Technology 36: 3827–3833.
Richmond, J.Q., Wood, D.A., Stanford, J.W. and Fisher, R.N. (2014).
‘Testing for multiple invasion routes and source populations for the
invasive brown treesnake (Boiga irregularis) on Guam: Implications
for pest management’. Biological Invasions 17: 337–349.
Rodda, G.H., McCoid, M.J., Fritts, T.H. and Campbell III, E.W. (1999).
‘Population trends and limiting factors in Boiga irregularis.’ In: G.H.
Rodda, Y. Sawai, D. Chiszar and H. Tanaka (eds.) Problem Snake
Management: the Habu and the Brown Treesnake, pp. 236–254. New
York: Cornell University Press.
Rodda, G.H. and Savidge, J.A. (2007). ‘Biology and impacts of Pacific
island invasive species: 2. Boiga irregularis, the brown tree snake
(Reptilia: Colubridae)’. Pacific Science 61: 307–324.
Savarie, P.J., Shivik, J.A., White, G.C., Hurley, J.C. and Clark, L. (2001).
‘Use of acetaminophen for large-scale control of brown tree snakes’.
Journal of Wildlife Management 65: 356–365.
Savarie, P.J. (2002). Acute oral toxicity of acetaminophen tablets to brown
treesnakes, Unpublished report – QA-636. Fort Collins, Colorado:
National Wildlife Research Center.
Savarie, P.J. and Tope, K.L. (2004). ‘Potential flotation devices for
aerial delivery of baits to brown treesnakes’. In: R.M. Timm and W.P.
Gorenzel (eds.) Proceedings of the 21st Vertebrate Pest Conference, pp.
27–30. Davis, California: University of California.
Savidge, J.A. (1987). ‘Extinction of an island forest avifauna by an
introduced snake’. Ecology 68: 660–668.
Shwiff, S.A., Gebhardt, K., Kirkpatrick, K.N. and Shwiff, S.S. (2010)
‘Potential economic damages from introduction of brown tree snakes,
Boiga irregularis (Reptilia: Colubridae), to the islands of Hawai’i’.
Pacific Science, 64: 1–10.
AUTHOR CONTRIBUTIONS
S. Siers was the principal investigator for the field
evaluation, establishing the study design, coordinating
site access and environmental compliance, curating data,
executing analyses and data visualisation, and writing the
original draft of the manuscript. S. Siers, M. Messaros, C.
Clark, and R. Gosnell supervised field evaluation activities.
W. Pitt and M. Messaros conceptualised the automated
bait cartridge manufacturing and delivery systems. M.
Messaros was the chief engineer and developer of the
intellectual property associated with the bait cartridge and
associated manufacturing and delivery systems. W. Pitt,
L. Clark, A. Shiels, J. Eisemann, and S. Siers contributed
to programme administration, funding acquisition, and
other matters associated with research and development
of the automated systems. J. Eisemann coordinated
pesticide registration and technology transfer matters, and
contributed to on-site coordination of evaluation activities.
All authors contributed to review and editing of the
manuscript.
355
B. Boag and R. Neilson
Boag, B. and R. Neilson. The potential detrimental impact of the New Zealand flatworm to Scottish islands
The potential detrimental impact of the New Zealand flatworm to
Scottish islands
B. Boag and R. Neilson
The James Hutton Institute, Invergowrie, Dundee, Scotland, DD2 5DA, UK. <brian.boag@hutton.ac.uk>.
Abstract: The New Zealand flatworm, Arthurdendyus triangulatus, is an alien invasive species in The British Isles
and the Faroes. It was probably first introduced after WWII and is an obligate predator of our native earthworms. It was
initially considered a curiosity until observations in the 1990s in Northern Ireland found it could significantly reduce
earthworm numbers. In 1992, it was scheduled under the Countryside and Wildlife Act 1981 then transferred to the
Wildlife and Natural Environment (Scotland) Act in 2011 which makes it an offence to knowingly distribute the flatworm.
A retrospective survey in Scotland showed that it was detected in botanic gardens, nurseries and garden centres in the
1960s but then spread to domestic gardens then finally to farms in the 1990s. Although the geographical distribution
of A. triangulatus was initially confined to mainland Scotland it was subsequently found established on 30 Scottish
Islands. Most of the islands are to the north and west of Scotland and have cool damp climates which are favoured by the
New Zealand flatworm. These islands also generally have relatively poor soils that support grassland farming systems.
Evidence from both Northern Ireland and Scotland suggests anecic species of earthworm which occur predominantly in
grassland, which help drainage and are a source of food for both animals and birds are at particular risk from the flatworm.
The detrimental impact of the flatworm on soil processes and wildlife has yet to be quantitatively evaluated but unlike
many other invasive species there is currently no known means of control. The precautionary principle must be therefore
applied wherever possible and every opportunity taken to stop its further spread.
Keywords: Arthurdendyus triangulatus, earthworms, invasive alien species, predator, Scotland
INTRODUCTION
Many of the Scottish islands have impoverished
acidic soils (Boyd, 1957; Glentworth, 1979; Hudson, et
al., 1982) which have been improved in the past by the
addition of seaweed, lime or occasionally imported soil
(Magnusson, 1997; Entwhistle, et al., 2000). Wind-blown
calcareous sandy soils with a high shell content occurring
in the north and west of Scotland, known as “machair”
(Angus, 2001), are characterised with a defined flora and
low input agriculture (Hudson, et al., 1982) and is a fragile
ecosystem listed under the EU Habitats Directive Annex
1. Earthworms prefer soils with a near neutral pH and
are therefore typically found in improved soils compared
with those with a pH levels< 4 (Guild, 1951; Edwards &
Bohlen, 1996). Boyd (1956, 1957) surveyed earthworms
in the Hebrides and recorded a complex of 15 species
from the machair, dominated by Lumbricus rubellus,
Aporrectodea caliginosa and Dendrobaena octaedra. This
contrasted with only six species under acidic heather soils,
the dominant species being L. rubellus and D. octaedra.
The genus of one of the recorded species, Bimastos, has
since been removed from the British list (Sims & Gerard,
1985).
Stop-Bowitz (1968), suggested that in Norway some
species e.g. D. octaedra and L. rubellus could have survived
the Quaternary ice age and this may also have occurred in
Scotland. However, many of the other recorded earthworm
species were probably introduced by man to enhance the
productivity of the land as occurred in New Zealand where
productivity was increased significantly by the addition
of European earthworm species (Stockdill, 1982). The
only other place in the world where the New Zealand
flatworm has become established is the Faroe Islands
where earthworms have been found in closed association
with human settlements possibly due to the inhabited areas
being on fertile land (Enckell & Rundgren, 1988).
The New Zealand flatworm (Arthurdendyus
triangulatus) was probably first accidentally introduced
into the British Isles just after WWII but not officially
recorded in Scotland until 1965 when it was considered
a curiosity (Wakeman & Vickerman, 1979). However,
Blackshaw (1990) reported that the presence of the
New Zealand flatworm was associated with a decline
in earthworm populations to below detectable levels in
Northern Ireland. The results of a survey undertaken in
Scotland during 1991–1992 (Boag, et al., 1994) indicated
that the New Zealand flatworm was initially confined to
botanic gardens, garden centres and nurseries but then
spread to domestic gardens in the 1970s and finally to farms
in the 1980s. It also showed that by 1992 it had spread to
many parts of Scotland including the islands of Skye and
Orkney. In the last 30 years the flatworm has been recorded
from several other islands off the west and north coast of
Scotland.
Further research indicated that the presence of the
New Zealand flatworm reduced the abundance of anecic
earthworm species (Jones, et al., 2001), populations of
which were unlikely to fully recover (Murchie & Gordon,
2013). Anecic earthworm species are those which make
vertical burrows and consume dead organic matter on the
soil surface (Fraser and Boag, 1998), thus play a key role
in soil nutrient processes and are considered ecosystem
engineers (Lavelle, et al., 1997; Blouin, et al., 2013).
Furthermore, they are also a major component of food for
some mammals and birds (Boag & Neilson, 2006).
The New Zealand flatworm prefers cool damp
conditions to survive (Boag, et al., 1998a) and this may
have contributed to it being a problem predominantly
in the north and west of Scotland, Ireland and the Faroe
Islands compared with the east of Scotland and England
(Jones & Boag, 1996). The flatworm is also dependent on
the presence of earthworms which potentially restricts its
distribution in Scotland as earthworms rarely occur in soils
with a pH < 4 (Boag, et al., 1998b).
The aim of the present paper is to document the extent to
which the New Zealand flatworm has become established
in the Scottish islands and to consider the detrimental
impact its presence might, in the future, have on island
agriculture and wildlife.
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
356
up to meet the challenge, pp. 356–359. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Boag & Neilson: New Zealand flatworm on Scottish islands
MATERIAL AND METHODS
The data used for this paper were the records of where
the New Zealand flatworms were found by the general
public in their gardens, parks etc. and these records over
time have also shown how it spread. The New Zealand
flatworm can easily be recognised by the general public as
different from earthworms as it is flat, covered by a sticky
mucus and pointed at both ends. Initially the records were
collated by the National Museum of Scotland until this was
taken over by the senior author after a survey financed by
the Scottish Government (Boag, et al., 1994) and a 1995
BBC TV survey (Jones & Boag, 1996). Subsequent records
have continued to be collected by staff at the James Hutton
Institute and submitted to and curated by the National
Biodiversity Network from where three additional island
records were gleaned for this paper. More recently The
Open Air Laboratory (OPAL) has run a citizen science
survey for the New Zealand flatworm (<https://www.
opalexplorenature.org>). The records are stored by the
National Biodiversity Network and the senior author is the
national expert on this species and verifies all records.
RESULTS
Most flatworm records were from individual
households with a few from farms, garden centres and
schools. Scotland has 790 offshore islands of which 95
are inhabited of which 30 islands recorded New Zealand
flatworm (Table 1). These were distributed from Arran in
the south of Scotland to Shetland in the north. Many of
the infested islands had few inhabitants, but in general the
number of flatworm records reflected the population size
(Table 1).
This was demonstrated across the Orkney archipelago
(Fig. 1) where there were 41 flatworm records, from a
population > 17,000 on mainland Orkney compared with the
outlying islands of Burray, Egilsay, Hoy, North Ronaldsay,
Fig. 1 Distribution of records of Arthurdendyus triangulatus,
the New Zealand flatworm, in Orkney.
Rousay and South Ronaldsay which had a combined
population of > 1,700 and had only seven flatworm records.
The Orkney mainland had the most records of all Scottish
Islands even though it has a smaller land area than Skye,
Shetland, Mull or Lewis. Of these larger islands Orkney
is by far the most fertile with a large proportion covered
with arable crops or permanent pasture (Dry & Robertson,
1982). Most island records only reported the presence of
the flatworm but others reported a reduction or absence of
native earthworms while others reported that large numbers
of flatworms had been collected e.g. a householder from
Baleshare killed 1,445 flatworms over a period between
May 2015 and January 2016. Another householder from
Skye regularly killed 20-40 flatworms daily with a reported
maximum of 150, and an estimated total kill of 15,000 over
a period of one year.
Table 1 Scottish islands infested with Arthurdendyus
triangulatus, the New Zealand flatworm: the number of
records; the human population and; area of the islands.
Island
No of records
Arran
5
Baleshare
1
Barra
3
Bressay
4
Burray
1
Bute
13
Coll
1
Easdale
1
Egilsay
1
Eriskay
1
Fair Isle
1
Gigha
3
Greater
2
Cumbrae
Harris
3
Hoy
1
Iona
1
Islay
2
Isle of Seil
1
Lewis
14
Lismore
4
Mull
6
North
1
Ronaldsay
North Uist
1
Orkney
41
Mainland
Rousay
1
Shetland
13
Skye
12
South
1
Ronaldsay
South Uist
1
Whalsay
1
Population
5,058
58
1,078
360
409
7,228
164
59
26
143
55
110
1,376
Hectares
43,201
910
5,875
2,805
903
12,217
7,685
25
650
703
768
1,305
1,168
1,916
272
120
3,228
21
18,500
146
2,667
72
50,119
14,320
877
61,956
1,329
163,695
2,351
87,535
690
1,271
17,162
30,305
52,325
26
22,000
10,008
909
4,860
96,879
165,625
4,980
1,754
14
32,026
1,970
357
Island invasives: scaling up to meet the challenge. Ch 2D Other taxa: Invertebrates
DISCUSSION
ACKNOWLEDGEMENTS
The records received over the last 25 years show that
the New Zealand flatworm is now widely distributed
in the Scottish islands. Since it is an obligate feeder on
earthworms, the presence of the flatworm also indicates that
these islands must have had an abundance of earthworms.
A possible reason for Orkney having a disproportionately
greater number of New Zealand flatworm records compared
with Lewis, Skye, Mull or Shetland is probably the fact that
Orkney is formed from sedimentary rock while the others
are igneous or metamorphic in origin (Dry & Robertson,
1982) thus more conducive to earthworm establishment
and survival.
The authors thank members of the general public who
have contacted the senior author and, in some cases, sent
photographs and specimens. We also thank Scottish Natural
Heritage and especially Colin MacLeod for submitting the
flatworm records onto the National Biodiversity Network.
The James Hutton Institute receives financial support from
the Scottish Government, Rural and Environment Science
and Analytical Services Division.
The New Zealand flatworm is known to have a
deleterious impact on earthworms in mainland Scotland
and Ireland (Jones, et al., 2001; Murchie & Gordon, 2013)
and it can probably be assumed that is also the case on
these islands. Apart from earthworms playing an important
role in delivering soil function and ecosystem services
(Lavelle, et al., 1997; Blouin, et al., 2013) they are an
important constituent of the diets of some mammals and
birds which live in the islands and are, in some cases,
declining in number e.g. the lapwing (Vanellus vanellus).
To help revive the decrease in lapwing numbers it has
been proposed that lime should be added to increase the
soil pH and hence encourage the build-up of earthworms
upon which lapwing feed (McCallum, et al., 2015). Studies
have also shown that earthworms are a major constituent
of the diet of chough (Pyrrhocorax pyrrhocorax) (Meyer,
1990), a rare breeding corvid found on Islay and Colonsay
with an estimate of c. 50 breeding pairs in 2014 (https://
scotlandsnature.wordpress.com/2014/09/25/goodnews-from-islay-as-population-grows/). It is therefore
concerning that flatworms have been recorded from Islay
as this may confound the conservation of chough on the
island.
Apart from the direct impact of New Zealand flatworm
on wildlife it has been estimated that it could have a
potential detrimental economic effect on agriculture
(Boag & Neilson, 2006). This is particularly relevant to
small holdings with tight farm unit margins such as crofts.
Circumstantial evidence from an area north of Dunoon
infested with flatworm suggested that in addition to an
accumulation of dead organic matter on the soil surface,
undesirable plants such as rushes became established as a
result of frequent flooding after rainfall events. The New
Zealand flatworm may also become a problem where there
are large amounts of arable land and permanent grassland
which occurs in mainland Orkney. Agricultural land in
Scotland can have a wide range of earthworm species
including the anecic species which would be particularly at
risk (Boag, et al., 1997).
No investigations have been undertaken to ascertain
the actual impact of the New Zealand flatworm on either
wildlife or agricultural production in the Scottish islands.
Assumptions on the detrimental impact of the New
Zealand flatworm must therefore be made based on the
knowledge gleaned from the literature on the benefits that
earthworms have on agricultural production and wildlife
(Schmidt & Curry, 1999; Bartlett, et al., 2010). Unlike
many other invasive plants and animals which have been
successfully removed from Scottish islands e.g. mink from
the Uists and rats from Canna (Bell, et al., 2011; Roy, et al.,
2015) there are no prospects of the New Zealand flatworm
being controlled on the Scottish islands once it has become
established. Given the only mechanism known to spread the
New Zealand flatworm is the human mediated movement
of plant material every effort must be made to stop infested
material reaching the islands by informing the general
public of the threat that New Zealand flatworm poses.
358
REFERENCES
Angus, S. (2001). ‘Machair’. In: S. Angus (ed.) The Outer Hebrides,
Moor and Machair, pp. 195–243. Cambridge: The White Horse Press.
Bartlett, M., Briones, M.J.I., Neilson, R., Schmidt, O., Spurgeon, D. and
Creamer R.E. (2010). ‘A critical review of current methods in earthworm
ecology: From individuals to populations’. European Journal of Soil
Biology 46: 67–73.
Bell, E., Boyle, D., Floyd, K., Garner-Richards, P., Swann, B., Luxmoore,
R., Patterson, A. and Thomas, R. (2011). ‘The ground-based eradication
of Norway rats (Rattus norvegicus) from the Isle of Canna, Inner
Hebrides, Scotland’. In: C.R. Veitch; M.N. Clout and D.R. Towns
(eds.) Island invasives: eradication and management, pp. 269–274.
Occassional Paper SSC no. 42. Gland, Switzerland: IUCN and
Auckland, New Zealand: CBB.
Blackshaw, R.P. (1990). ‘Studies on Artioposthia triangulata (Dendy)
(Tricladida: Terricola), a predator of earthworms’. Annals of Applied
Biology 118: 698–694.
Blouin, M., Hodson, M.E., Delgado, E.A., Baker, G., Brussaard, L., Butt,
K.R., Dai, J., Dendooven, L., Peres, G., Tondon, J.E., Cluzeau, D. and
Brun J. (2013). ‘A review of earthworm impact on soil function and
ecosystem services’. European Journal of Soil Science 64: 161–182.
Boag, B., Palmer, L.F., Neilson, R. and Chambers, S.J. (1994).
‘Distribution and prevalence of the predatory planarian Artioposthia
triangulata (Dendy) (Tricladida: Terricola) in Scotland’. Annals of
Applied Biology 124: 165–171.
Boag, B., Palmer, L.F., Neilson, R., Legg, R. and Chambers, S.J. (1997).
‘Distribution, prevalence and intensity of earthworm populations in
arable land and grassland in Scotland’. Annals of Applied Biology 130:
153–165.
Boag, B., Yeates, G.W. and Johns, P.M. (1998a). ‘Limitations to the
distribution and spread of terrestrial flatworms with special reference
to the New Zealand flatworm (Artioposthia triangulata)’. Pedobiologia
42: 495–503.
Boag, B., Jones, H.D., Evans, K.A., Neilson, R., Yeates, G.W. and
Johns, P.M. (1998b). ‘The application of GIS techniques to estimate
the establishment and potential spread of Artioposthia triangulata in
Scotland’. Pedobiologia 42: 504–510.
Boag, B. and Neilson, R. (2006). ‘Impact of the New Zealand flatworm
on agriculture and wildlife in Scotland’. Proceedings Crop Protection
in Northern Britain 2006, pp. 51–56. Dundee, UK: Page Bros, Norwich
Ltd.
Boyd, J.M. (1956). ‘The Lumbricidae in the Hebrides II Geographical
distribution’. Scottish Naturalist 68: 165–172.
Boyd, J.M. (1957). ‘Ecological distribution of the Lumbricidae in the
Hebrides’. Proceedings of the Royal Society Edinburgh B 66: 311–338.
Dry, F.T. and Robertson, J.S. (1982). Soils and Land Capability for
Agriculture, Orkney and Shetland. Aberdeen, UK: University Press.
Edwards, C.A. and Bohlen, P.J. (1996). Biology and Ecology of
Earthworms. London, UK: Chapman and Hall.
Enckell, P.H. and Rundgren, S. (1988). ‘Anthropochorous earthworms
(Lumbricidae) as indicators of abandoned settlements in the Faroe
Islands’. Journal of Archeaological Science 15: 439–351.
Entwhistle, J.A., Dodgshon, R.A. and Abrahams, P.W. (2000). ‘An
investigation of former land-use activity through the physical and
chemical analysis from soils from the Isle of Lewis, Outer Hebrides’.
Archaeological Prospection 7: 171–188.
Fraser, P.M. and Boag, B. (1998). ‘The distribution of lumbricid
earthworm communities in relation to flatworms: a comparison between
New Zealand and Europe’. Pedobiologia 42: 642–553.
Glentworth, R. (1979). ‘Observations on the soils of the Outer Hebrides’.
Proceedings of the Royal Society of Edinburgh B 77: 123–137.
Boag & Neilson: New Zealanf flatworm on Scottish islands
Guild, Mc.L.W.J. (1951). ‘The distribution and population density
of earthworms (Lumbricidae) in Scottish pasture fields’. Journal of
Applied Ecology 20: 88–97.
Hudson, G., Towers, W., Bibby, J.S. and Henderson, D.J. (1982). The
Outer Hebrides: Soil and Land Capability for Agriculture, with a
Contribution from J. S. Robertson. Handbook of the Soil Survey of
Scotland. Aberdeen, UK: University Press.
Jones, H.D. and Boag, B. (1996). ‘The distribution of the New Zealand
and Australian flatworms (Platyhelminthes: Turbelliaria: Tricladida:
Terricola) in the British Isles – the Scottish survey and MEGALAB
WORMS’. Journal of Natural History 30: 955–975.
Jones, H.D., Santoro, G., Boag, B. and Neilson, R. (2001). ‘The diversity
of earthworms in 200 Scottish fields and the possible effect of the
New Zealand flatworms (Arthurdendyus triangulatus) on earthworm
populations’. Annals of Applied Biology 139: 75–92.
Lavelle, P., Bignell, D., Lepage, M., Wolters, V., Roger, P., Ineson, P.,
Heal, O.W. and Dhillion, S. (1997). ‘Soil function in a changing world:
The role of invertebrate ecosystem engineers’. European Journal of Soil
Biology 33: 159–193.
Magnusson, M. (1997). Rum: Nature’s Island. Edinburgh, UK: Luath
Press.
McCallum, H.M., Park, K.J., O’Brien, M.G., Gimona, A., Poggio, L. and
Wilson, J.D. (2015). ‘Soil pH and organic matter content add explanatory
power to the northern lapwing Vanellus vanellus distribution models
and suggest soil amendment as a conservation measure on upland
farmland’. Ibis 167: 677–687.
Meyer, R.M. (1990). ‘Observations on two red-billed choughs
Pyrrhocorax pyrrhocorax in Cornwall: Habitat use and food intake’.
Bird Study 37: 199–209.
Murchie, A.K. and Gordon, A.W. (2013). ‘The impact of the “New
Zealand” flatworm, Arthurdendyus triangulatus, on earthworm
populations in the field’. Biological Invasions 15: 469–586.
Roy, S.S., Chauvenet, A.I.M. and Robertson, P.A. (2015). ‘Removal
of American mink (Neovison vison) from the Uists, Outer Hebrides,
Scotland’. Biological Invasions 17: 2811–2820.
Schmidt, O. and Curry, J.P. (1999). ‘Effects of earthworms on biomass
production, nitrogen allocation and nitrogen transfer in wheat-clover
intercropping model systems’. Plant and Soil 214: 187–198.
Sims, R.G. and Gerard, B.M. (1985). Earthworms: Keys and Notes for
the Identification and Study of Species. Synopss of the British Fauna.
London, UK and Leiden, The Netherlands: E.J. Brill.
Stop-Bowitz, C. (1968). ‘A contribution to our knowledge of the
systematics and zoogeography of Norwegian earthworms (Annelida
Oligochaeta: Lumbricidae)’. Norwegian Journal of Zoology 17: 169–
280.
Stockdill, M.J. (1982). ‘Effects of introduced earthworms on the
productivity of New Zealand pastures’. Pedobiologia 24: 29–35.
Wakeman, D. and Vickerman, K. (1979). ‘The land planarian Artioposthia
triangulata, a new species record for the West of Scotland’. Glasgow
Naturalist 19: 499–501.
359
R.A. Boulton, M. Bulgarella, I.E. Ramirez, C.E. Causton and G.E. Heimpel
Boulton, R.A.; M. Bulgarella, I.E. Ramirez, C.E. Causton and G.E. Heimpel. Management of an
invasive avian parasitic fly in the Galápagos Islands: is biological control a viable option?
Management of an invasive avian parasitic fly in the Galapagos Islands:
is biological control a viable option?
R.A. Boulton1,2, M. Bulgarella3, I.E. Ramirez1, C.E. Causton4 and G.E. Heimpel1
University of Minnesota, Department of Entomology, St Paul, Minnesota, USA. 2University of Exeter, College of Life
and Environmental Sciences, Penryn Campus, Cornwall, UK. <R.Boulton@exeter.ac.uk>. 3Victoria University of
Wellington, School of Biological Sciences, Wellington, New Zealand. 4Charles Darwin Foundation, Puerto Ayora,
Santa Cruz, Galápagos, Ecuador.
1
Abstract The bird-parasitic fly, Philornis downsi, was first recorded in the Galápagos Islands in 1964 where it likely
invaded from mainland Ecuador. This muscid fly is now the leading cause of recent declines in endemic landbird
populations as its larvae feed on the nestlings of at least 19 bird species in the Galápagos, including many species of
Darwin’s finches. As yet, no long-term control method has been implemented for P. downsi, but importation (also known
as classical) biological control may be a viable option. Due to historically high-profile examples of biological control
agents attacking non-target species, some consider biological control to be too risky to be compatible with conservation
aims. However, since biosafety practices were implemented beginning in the 1990s, these risks have been drastically
reduced, and biological control is now an important tool for suppressing invasive species that are difficult to control using
other means. We investigated the safety of a potential biological control agent, the parasitoid wasp, Conura annulifera, that
attacks P. downsi in its native range. Here we summarise the results of a series of field, laboratory and comparative studies
on C. annulifera (methods and results are not reported here) and outline future directions. We used a field experimental
paradigm involving nest boxes baited with non-target hosts, and quarantine laboratory no-choice trials in which nontarget hosts were exposed to C. annulifera. Our work to-date suggests that C. annulifera is restricted to attacking species
within the genus Philornis. Furthermore, a phylogenetically controlled comparative study suggests that C. annulifera is
evolutionarily constrained in its host range. These results lead us to conclude that C. annulifera demonstrates promise as
an ecologically safe agent for the long-term biological control of P. downsi. Studies will now focus on an evaluation of
risks to endemic and native species in the Galápagos.
Keywords: biological control, conservation, Conura annulifera, invasive species, Philornis downsi
THE HISTORY OF BIOLOGICAL CONTROL
When non-native species invade novel ecosystems, the
consequences can be extreme. Anthropogenic change has
increased the frequency and the success of such invasions;
increased traffic to previously isolated sites increases the
chance that non-native species will arrive, and the chances
of colonisation are often increased in areas that are altered
by human activity (Sax, et al., 2002). When colonisation
occurs, non-native species are often far more successful
than in their native range. Although the reasons for this
are likely complex, one often-cited reason is the ‘enemy
release’ hypothesis (ERH) (Liu & Stiling, 2006). Enemy
release occurs when invaders colonise a new area, free from
the natural enemies (predators, parasites and pathogens)
with which they co-evolved, and are released from the
effects that these enemies have on population suppression.
Importation (also known as classical) biological control
involves reconstructing (at least in part) the assemblage
of co-evolved natural enemies present in the native range
of the problematic species in order to control it (Heimpel
& Mills, 2017). Although more commonly known from
agricultural systems, importation biological control
for conservation is a developing sub-discipline and
shows promise as a long-term strategy for dealing with
harmful invasive species (Van Driesche, et al., 2010; Van
Driesche & Reardon, 2017). One noteworthy example
is the introduction of the specialised ladybeetle, Rodolia
cardinalis, which effectively controlled populations of the
cottonycushion scale, Icerya purchasi, in the Galápagos
archipelago. This particular introduction has more than
likely been the saviour of endemic plant species that are
attacked by I. purchasi (Hoddle, et al., 2013).
Unfortunately, historically high-profile cases of
biological control failures that led to non-target effects
on threatened species have received significant media
attention and these examples have hampered progress
in the sub-discipline of conservation-focused biological
control (Van Driesche & Reardon, 2017). In order for
importation biological control to be safe and successful, it
is paramount that we understand the ecology, in particular
the host specificity of the putative natural enemy set for
release. A majority of examples of biological control, for
conservation or for agriculture, have demonstrated both its
success and its safety, particularly since the 1990s when
concerns over biosafety gained momentum (Barratt, et
al., 2010; Van Driesche, et al., 2010; Van Driesche, 2012;
Heimpel & Mills, 2017; Van Driesche & Reardon, 2017;
Heimpel & Cock 2018). However, the negative reputation
of biological control persists due to the memorable nature
of failures that have caused detrimental effects on native
fauna (see Clarke, et al., 1984; Howarth 1991). Sadly, it
is these examples that are more publicly well-known due
to the strong emotions that they elicit (Van Driesche, &
Reardon, 2017). Yet there is still hope for the discipline,
and conservation-focused biological control has initiated
a paradigm shift, demonstrating that biological control
can be more than just compatible with conservation aims,
it can actually promote them (Van Driesche, et al., 2010;
Heimpel & Cock, 2018). In order for these techniques
to be incorporated into the conservation ‘tool box’ it is
imperative that we build trust between practitioners of
biological control, conservationists and the public. To do
so, the biological control community must demonstrate the
pragmatism and caution that go into designing safe and
effective biological control programmes.
Philornis downsi
Philornis downsi (Dodge & Aitken) is a bird-parasitic
muscid fly that is native to mainland South America but is
invasive in the Galápagos Islands where it likely invaded
from mainland Ecuador (Bulgarella, et al., 2015). It was
first reported in the archipelago in 1964 and, in the last
15–20 years has become a major threat to the persistence
of many passerine bird species in the Galápagos, including
In:
360C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 360–363. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Boulton, et al.: Avian parasitic fly in the Galápagos Islands
the majority of species of Darwin’s finches (Fessl, et al.,
2018). This threat occurs because of the way the larvae of
P. downsi feed: the adults are free-living but the larvae are
obligate ectoparasitic blood-feeders on young nestlings,
leading to blood loss and death (Fessl, et al., 2006;
O’Connor, et al., 2010; Kleindorfer, et al., 2014; Koop, et
al., 2016; Heimpel, et al., 2017). P. downsi is considered
the greatest threat to the persistence of many land-bird
species in Galápagos. The critically endangered mangrove
finch (Camarhynchus heliobates) and medium tree finch
(C. pauper) are particularly at risk, with any nestlings of
the former now being protected by ‘head-starting’ (handrearing any eggs collected in the wild; Cunninghame, et
al., 2012). The ramifications of any such extinction would
be extensive. Not only would this represent a tragic loss of
iconic species in a well-protected environment, it would
also be a terrible loss of evolutionary history and the
opportunity to study it. Rosemary and Peter Grant spent the
last forty years studying the evolution of Darwin’s finches
and have commented on the importance and uniqueness of
Darwin’s finches for work of this nature:
‘A final reason that makes them (Darwin’s finches) so
suitable (for studying evolution) is that none of the species
has become extinct as a result of human intervention. This
cannot be said for many other radiations elsewhere in the
world.’ Grant & Grant (2009)
The Grants’ work has demonstrated the power of
evolution and speciation and the underlying mechanisms,
but much more remains to be discovered (e.g. Abzhanov,
2010 and articles therein). Losing even a single species
of Darwin’s finch would represent a terrible loss for
evolutionary biology and could have a profound impact on
future phenomena that the species’ radiation may reveal to
us. Moreover, losing large numbers of individuals of any
of these species will have considerable impacts on the
functioning of Galápagos ecosystems due to the critical
roles that they play in pollination and seed dispersal
(Causton, et al., 2013; Traveset, et al., 2015; Nogales, et
al., 2017).
Need and potential for biological control of Philornis
downsi
Infestation by P. downsi results in extreme nestling
mortality in Galápagos, which has not been observed in
the native range of the fly (Fessl, et al., 2018). The ERH
(Liu & Stiling, 2006) – a paucity of co-evolved natural
enemies in the invaded compared to the native range – is
one likely reason for the increased abundance of P. downsi
in Galápagos compared to the mainland (Bulgarella,
et al., 2015; 2017; Boulton & Heimpel, 2017). The
ERH serves as the theoretical underpinning of modern
importation biological control, whereby one or a suite of
co-evolved natural enemies is liberated into the invasive
range to control the target species (Heimpel & Mills,
2017). The scarcity of natural enemies of Philornis spp.
in the Galápagos compared to the mainland suggests that
importation biological control may be a valuable tool to
control P. downsi (Bulgarella, et al., 2017).
Although several control strategies are currently being
explored and considered, importation biological control
may be critical in protecting Darwin’s finches and other
endemic bird species in Galápagos from P. downsi. Other
possible control methods include short-term strategies,
such as nest treatment with insecticide and mass trapping
using lures (Fessl, et al., 2018). The short-term approaches
are considered mainly as stop-gap measures, whilst longterm measures, such as biological control and sterile male
release, are developed and implemented. Of the longterm measures considered so far, biological control using
natural enemies from the native range is currently the most
promising solution. The release of sterile males is another
potential long-term solution but this is currently hampered
by difficulties in laboratory breeding of P. downsi (Lahuatte,
et al., 2016; Fessl, et al., 2018).
In 2012, a workshop was organised by the Charles
Darwin Foundation and the Galápagos National Park
Directorate in order to form an action plan for conservation
of Darwin’s finches and other small land birds due to
the ever-increasing threat from P. downsi (Causton, et
al., 2013). One priority research goal recognised at this
workshop was to identify natural enemies in the fly’s
native range and investigate the potential for biological
control (Causton, et al., 2013). Over the last four years, we
have discovered several parasitoid wasp species attacking
species of Philornis in mainland Ecuador (Bulgarella, et
al., 2015; 2017). Before any of these parasitoids can be
considered as suitable biocontrol agents, in-depth studies
of their host range need to be conducted. To address this
question, we have been using a holistic approach consisting
of a novel field experimental paradigm, comprehensive
literature review, detailed study of the physiology and
evolutionary ecology of the putative biological control
agents, and traditional laboratory host range tests. In this
manuscript we review and summarise our published work
so far and outline future directions.
FIELD OBSERVATIONS AND EXPERIMENTS
Field work at two field sites in western mainland
Ecuador between 2013 and 2017 has revealed a number of
parasitoid species attacking Philornis spp. pupae collected
from nest boxes (Bulgarella, et al., 2015; 2017). In addition,
we have developed a novel field experimental paradigm
over the last two years that can be used as a preliminary
assay to test whether the parasitoid wasp species that we
have recovered are specific to Philornis spp. in the field.
The experimental set-up was as follows. Nest boxes
that we monitor throughout the bird breeding season for
P. downsi pupae and their parasitoid wasps were paired
with bait boxes. These bait boxes contained a number
of non-target host species that had been reared from the
local area. We also placed pupae of non-target species
inside active bird nests. Any parasitoid wasp species that
attacked Philornis spp. in the nest boxes and nests also had
the opportunity to attack non-target hosts in the adjacent
bait boxes and inside active nests. Using this experimental
paradigm, we were able to determine which (if any) species
of parasitoid wasp did not exclusively attack Philornis
spp. We have concentrated our further efforts on Conura
annulifera (Hymenoptera: Chalcididae), a parasitoid that
has been recorded attacking only Philornis spp. in these
field experiments. We will concentrate on this species
for the remainder of the manuscript but note that we are
also considering other species for biological control of P.
downsi, such as an unidentified species of Trichopria (see
Bulgarella, et al. (2017) and Boulton & Heimpel (2017) for
details). This study is in progress at the time of writing and
the results will be published elsewhere.
Life history and evolutionary ecology of Conura
annulifera
Previous work on the natural host range of C.
annulifera supports our assertion that it is a specialist
on the genus Philornis. It has been recorded in previous
studies throughout South and Central America where is has
been reported as parasitising only Philornis spp. (including
P. downsi and P. deceptivus; Burks, 1960; De Santis, 1979;
Delvare, 1992; Couri, et al., 2006). Moreover, studies
where pupae were reared from other Diptera (Muscidae,
Calliphoridae and Sarcophagidae) in regions where C.
annulifera has been reported never yielded this parasitoid
361
Island invasives: scaling up to meet the challenge. Ch 2D Other taxa: Invertebrates
(Bulgarella, et al., 2017). However, only five studies have
reported finding C. annulifera in the field (Burks, 1960;
De Santis, 1979; Delvare, 1992; Couri, et al., 2006;
Bulgarella, et al., 2017), and so more data were needed in
order to determine whether this species might constitute a
Philornis-specific biological control agent. In the sections
below, we present the evidence that we have accumulated
so far in support of the possibility that C. annulfiera is a
Philornis genus specialist.
Conura annulifera is a solitary pupal ectoparasitoid
(Bulgarella, et al., 2017). It attacks pupae of Philornis spp.,
laying a single egg on the outside of the developing pupa.
More specifically, C. annulifera is a ‘gap-layer’, a parasitoid
that deposits its egg between the hard external puparium
and the soft body of the developing pupa. We hypothesised
that the specificity of this oviposition site is likely to
restrict the range of suitable hosts that C. annulifera can
parasitise to the cyclorrhaphan Diptera, an unranked taxon
that contains families such as the Muscidae, Calliphoridae,
Sarcophagidae and Syrphidae (Griffiths, 1972; Boulton
& Heimpel, 2017). The Cyclorrhapha are the only group
of holometabolous insects that exhibit this gap (Whitten,
1957), and so it is unlikely that species outside this taxon
are physiologically viable hosts for C. annulifera. We
tested this possibility using phylogenetically controlled
comparative studies for all known species of gap-layers
in the superfamily Chalcidoidea and the results support
our hypothesis: gap-laying species exhibit narrower host
ranges than ‘true’ ectoparasitoids (Boulton & Heimpel,
2018). Moreover, these analyses revealed that gap-laying
as a strategy may constitute an evolutionary dead-end.
Compared to endoparasitoids and other ectoparasitoids,
evolutionary transitions towards gap-laying were more
likely than transitions away from it (Boulton & Heimpel,
2018).
This comparative work has implications for biological
control in general and for the specific case of control
outlined here. Our findings suggest that (1) gap-layers
such as C. annulifera are likely to be more host specific,
and so safer putative biological control agents, than ‘true’
ectoparasitoids, and (2) gap-layers including C. annulifera
may represent particularly useful agents for importation
biological control as they are less likely to transition, or
diversify, to attack novel hosts after release outside their
native range. With regards to the specific case of using
C. annulifera to control P. downsi in the Galápagos, this
work improves our understanding of the most at-risk nontarget organisms were a release to be attempted, but it does
not explicitly tell us whether C. annulifera is likely to be
a safe species for importation biological control. To test
this, more traditional host range studies were conducted,
the results of which we outline in the section below.
Laboratory host range studies
Bulgarella, et al. (2017) exposed a range of non-target
host pupae to C. annulifera that were maintained in the
laboratory. This included five cyclorrhaphan Diptera
(Musca domestica, M. autumnalis, Stomoxys calcitrans
(Muscidae), Sarcophaga bullata (Sarcophagidae),
Calliphora vicina (Calliphoridae)), three Lepidoptera
(Epiphyas postvittana (Tortricidae), Manduca sexta
(Sphingidae), Plodia interpunctella (Pyralidae)) and a
hymenopteran (Habrobracon hebetor (Braconidae)).
These species were chosen due to their likely physiological
compatibility with parasitism by C. annulifera (Diptera)
and because other species in the genus Conura have been
shown to attack various Lepidoptera and Hymenoptera
(see Bulgarella, et al., 2017).
In no case did the wasp produce offspring on any of
these non-target species: in these lab studies, C. annulifera
362
only reproduced successfully on P. downsi. This suggests
that, of the species presented so far, only P. downsi
represents a viable host for C. annulifera (Bulgarella, et
al., 2017). However, this experimental design did not allow
us to address the mechanism underlying this apparent
specificity. It could either be that C. annulifera does not
attempt to attack any species other than Philornis (i.e.,
behavioural specificity) or the wasp attempts to parasitise
these species but their offspring fail to develop and emerge
(i.e., only physiological specificity; see Desneux, et al.,
2009). For an importation biological control programme
with C. annulifera to be truly considered safe, it is
important that we rule out the possibility that C. annulifera
would attack non-target hosts, and cause their mortality by
envenomation or oviposition.
To do this, we carried out additional analyses to test
whether exposure to C. annulifera had any influence on
the successful emergence of non-target pupae compared
with controls. We found no evidence that exposure to C.
annulifera resulted in elevated mortality for non-target
hosts (see Bulgarella, et al., 2017). In contrast, when
P. downsi pupae were exposed to the wasp, mortality
increased independently of successful parasitism (i.e. more
unparasitised fly pupae failed to emerge in the exposed
treatment than in the control), perhaps as a result of hostfeeding or envenomation/attempted parasitism. This
finding, plus behavioural observations, suggests that C.
annulifera does not attempt to sting or probe any potential
host other than Philornis spp. pupae (Boulton & Heimpel,
2017; Bulgarella, et al., 2017).
FUTURE DIRECTIONS
Although all the evidence accumulated so far suggests
that C. annulifera is a specialist parasitoid of Philornis spp.
and should be seriously considered as a potential agent
for the biological control of P. downsi in the Galápagos,
one crucial question regarding the host range remains. It
is critical to know whether C. annulifera is able to attack
and develop on native or endemic non-target species
present in the archipelago. As is common for most oceanic
islands, the Galápagos exhibits high rates of endemism in
insects (Peck, 1996). Island endemics may be particularly
vulnerable to the introduction of a non-native parasitoid
due to their lack of shared co-evolutionary history and the
necessary adaptations to evade or resist parasitism. Before
we can consider biological control in the Galápagos using
any natural enemy, we must evaluate the host specificity
of the putative biological control agent in the context
under which it is intended for use. The studies that we
have conducted using C. annulifera thus far represent a
vital first step, suggesting that importation of C. annulifera
into a quarantine facility in the Galápagos for further host
range testing is justifiable. The results of these studies also
allow us to narrow down the list of most at-risk non-target
organisms in the Galápagos, due to the limitations imposed
by its evolutionary and behavioural ecology.
Importation biological control of P. downsi in the
Galápagos constitutes a promising means of population
suppression that may ultimately serve to protect the
extremely vulnerable bird species that the fly attacks
(Boulton & Heimpel, 2017). Establishment of a biological
control agent such as C. annulifera, may, in addition to
ameliorating the current situation, serve as a preventative
measure from future invasions of P. downsi and other bird
parasitic species in the genus Philornis that are found
in Ecuador. Preventative measures such as this may be
deemed particularly judicious given the probability of
further invasions under the high tourism pressure that the
islands currently face (Toral-Granda, et al., 2017).
Boulton, et al.: Avian parasitic fly in the Galápagos Islands
ACKNOWLEDGEMENTS
We would like to thank Paola Lahuatte and Andrea
Cahuana for providing Philornis downsi from Galápagos;
Mauricio Torres for assistance with field work supporting
the reviewed studies; Jonathan Dregni, Carleigh
Windhoerst, Ashleigh Sampedro, Justine Bleick, Leah
Vaugn and Melisa Rodriguez for assistance with bird,
fly and wasp husbandry supporting this work; Stephanie
Dahl and Nik Prenevost for assistance with logistics in
the quarantine facility; Bradley Sinclair for assistance
with insect identifications; and the Ecuadorian Ministry
of Environment and Galapagos National Park Directorate
(Projects PC 10-15 and 18-16) for permission to import
samples and conduct our research. This work is funded by
the Galápagos Conservancy, the International Community
Foundation (with a grant awarded by the Leona M. and
Harry B. Helmsley Charitable Trust). This is contribution
number 2202 of the Charles Darwin Foundation for the
Galapagos Islands.
REFERENCES
Abzhanov, A. (2010). ‘Darwin's Galápagos finches in modern biology’.
Philosophical Transactions of the Royal Society of London B: Biology
365: 1001–1007.
Barratt, B.I., Howarth, F.G., Withers, T.M., Kean, J.M. and Ridley, G.S.
(2010). ‘Progress in risk assessment for classical biological control’.
Biological Control 52: 245–254.
Boulton, R.A. and Heimpel, G.E. (2017). ‘Potential for Biological
Control of a Parasite of Darwin’s Finches’. In: R.G. Van Dreische and
R.C. Reardon (eds.). Supressing Over-abundant Invasive Plants and
Insects in Natural Areas by Use of Their Specialized Natural Enemies,
pp. 23–28, USA: FHTET, USDA.
Boulton, R.A. and Heimpel, G.E. (2018). ‘Mind the gap: The
evolution of oviposition site and specialisation in the parasitoid
superfamily Chalcidoidea’. Biological Journal of the Linnean Society
124: 213–227.
Bulgarella, M., Quiroga, M.A., Dregni, J.S., Cunninghame, F, Muñoz,
D.A.M., Monje, L.D., Causton, C.E. and Heimpel, G.E. (2015).
‘Philornis downsi (Diptera: Muscidae), an avian nest parasite invasive
to the Galápagos islands, in mainland Ecuador’. Annals of the
Entomological Society of America 108: 242–250.
Bulgarella, M., Quiroga, M.A., Boulton, R.A., Ramirez, I., Moon, R.D.,
Causton, C.E. and Heimpel, G.E. (2017). ‘Life cycle and host specificity
of the parasitoid Conura annulifera (Hymenoptera: Chalcididae),
a potential biological control agent of Philornis downsi (Diptera:
Muscidae) in the Galápagos Islands’. Annals of the Entomological
Society of America 110: 317–328.
Burks, B.D. (1960). ‘A Spilochalcis parasitic on flies that infest birds’
nests (Hymenoptera, Chalcididae)’. Acta Hymenopterologica 1: 95–97.
Causton, C.E., Cunninghame, F. and Tapia, W. (2013). ‘Management
of the avian parasite Philornis downsi in the Galápagos Islands: A
collaborative and strategic action plan’. Galápagos Report 2011–2012:
167–173.
Clarke, B., Murray, J. and Johnson, M.S. (1984). ‘The extinction of
endemic species by a program of biological control’. Pacific Science
38: 97–104.
Couri, M.S., Tavares, M.T. and Stenzel, R.R. (2006). ‘Parasitoidism of
chalcidid wasps (Hymenoptera, Chalcididae) on Philornis sp. (Diptera,
Muscidae)’. Brazilian Journal of Biology 66: 553–557.
Cunninghame, F., Switzer, R., Parks, B., Young, G., Carrión, A.,
Medranda, P. and Sevilla, C. (2013). ‘Conserving the critically
endangered mangrove finch: head-starting to increase population size’.
Galápagos Report 2014: 151–157.
De Santis, L. (1979). Catálogo de los Himénopteros Calcidoideos
de América al sur de los Estados Unidos. La Plata, Buenos Aires,
Argentina: Publicación Especial de la Comisión de Investigaciones
Científicas de la Provincia de Buenos Aires.
Delvare, G. (1992). ‘A reclassification of the Chalcidini with a checklist
of the New World species’. Memoirs of the American Entomological
Institute 53: 315.
Desneux, N., Barta, R.J., Hoelmer, K.A., Hopper, K.R. and Heimpel,
G.E. (2009). ‘Multifaceted determinants of host specificity in an aphid
parasitoid’. Oecologia 160: 387–398.
Fessl, B., Heimpel, G.E. and Causton, C.E. (2018). ‘Invasion of an
Avian Nest Parasite, Philornis downsi, to the Galápagos Islands:
Colonization History, Adaptations to Novel Ecosystems, and
Conservation Challenges’. In: P.G. Parker (ed.) Disease Ecology: Social
and Ecological Interactions in the Galapagos Islands, pp. 213–266.,
Dordrecht, The Netherlands: Springer.
Grant, B.R. and Grant, P.R. (2009). ‘Darwin's Finches: Studying Evolution
in Action’. In: T. de Roy (ed.) Galápagos: Preserving Darwin's Legacy.
pp. 122–129. Auckland, New Zealand: David Bateman.
Griffiths, G.C.D. (1972). The Phylogenetic Classification of Diptera
Cyclorrhapha. Dordrecht, The Netherlands: Springer.
Heimpel, G.E., Hillstrom, A., Freund, D., Knutie, S.A. and Clayton,
D.H. (2017). ‘Invasive parasites and the fate of Darwin's finches in the
Galápagos Islands: The case of the vegetarian finch’. Wilson Journal of
Ornithology 129: 345–349.
Heimpel, G.E. and Mills, N.J. (2017). Biological Control: Ecology and
Applications. Cambridge, UK: Cambridge University Press.
Heimpel, G.E. and Cock, M.J.W. (2018). ‘Shifting paradigms in the
history of importation biological control’. BioControl 63: 27–37.
Hoddle, M.S., Ramírez, C.C., Hoddle, C.D., Loayza, J., Lincango, M.P.,
Van Driesche, R.G. and Causton, C.E. (2013). ‘Post release evaluation
of Rodolia cardinalis (Coleoptera: Coccinellidae) for control of Icerya
purchasi (Hemiptera: Monophlebidae) in the Galápagos Islands’.
Biological Control 67: 262–274.
Howarth, F.G. (1991). ‘Environmental impacts of classical biological
control’. Annual Review of Entomology 36: 485–509.
Kleindorfer, S., Peters, K.J., Custance, G., Dudaniec, R.Y. and O’Connor,
J.A. (2014). ‘Changes in Philornis infestation behavior threaten
Darwin’s finch survival’. Current Zoology 60: 542–550.
Koop, J.A., Kim, P.S., Knutie, S.A., Adler, F. and Clayton, D.H. (2016).
‘An introduced parasitic fly may lead to local extinction of Darwin's
finch populations’. Journal of Applied Ecology 53: 511–518.
Lahuatte, P.F., Lincango, M.P., Heimpel, G.E. and Causton, C.E., (2016).
‘Rearing larvae of the avian nest parasite, Philornis downsi (Diptera:
Muscidae), on chicken blood-based diet’. Journal of Insect Science
16(1): 84.
Liu, H. and Stiling, P. (2006). ‘Testing the enemy release hypothesis: a
review and meta-analysis’. Biological Invasions 8: 153–1545.
Nogales, M., Gonzalez-Castro, A., Rumeu, B., Traveset, A., Vargas, P.,
Jaramillo, P., Olesen, J.M. and Heleno, R.H. (2017). ‘Contribution by
vertebrates to seed dispersal effectiveness in the Galapagos Islands: a
community-wide approach’. Ecology 98: 2049–2058.
O’Connor, J.A., Sulloway, F.J., Robertson, J. and Kleindorfer, S. (2010).
‘Philornis downsi parasitism is the primary cause of nestling mortality
in the critically endangered Darwin’s medium tree finch (Camarhynchus
pauper)’. Biodiversity Conservation 19: 853–866.
Peck, S.B. (1996). ‘Origin and Development of an Insect Fauna on a
Remote Archipelago: The Galapagos Islands, Ecuador’. In: A. Keast
and S.E. Miller (eds.) The Origin and Evolution of Pacific Island Biotas,
New Guinea to Eastern Polynesia: Patterns and Processes, pp. 91–122.
Amsterdam, The Netherlands: SPB Academic Publishers.
Sax, D.F., Gaines, S.D. and Brown, J.H. (2002). ‘Species invasions
exceed extinctions on islands worldwide: A comparative study of plants
and birds’. American Naturalist 160: 766–783.
Toral-Granda, M.V., Causton, C.E., Jaeger, H., Trueman, M., Izurieta,
J.C., Araujo, E., Cruz, M., Zander, K.K., Izurieta, A. and Garnett, S.T.
(2017). ‘Alien species pathways to the Galapagos Islands, Ecuador’.
PLOS ONE 12: e0184379.
Traveset, A., Olesen, J.M., Nogales, M., Vargas, P., Jaramillo, P., Antolín,
E., Trigo, M.M. and Heleno, R. (2015). ‘Bird-flower visitation networks
in the Galapagos unveil a widespread interaction release’. Nature
Communications 6: Article Number 6376.
Van Driesche, R.G., Carruthers, R.I., Center, T., Hoddle, M.S., HoughGoldstein, J., Morin, L., Smith, L., Wagner, D.L., Blossey, B.,
Brancatini, V. and Casagrande, R. (2010). ‘Classical biological control
for the protection of natural ecosystems’. Biological Control Supplement
1: S2–S33.
Van Driesche, R.G. (2012). ‘The role of biological control in wildlands’.
BioControl 57: 131–137.
Van Driesche, R.G. and Reardon, R.C. (2017). ‘Supressing Overabundant Invasive Plants and Insects in Natural Areas by Use of Their
Specialized Natural Enemies’. USA: FHTET, USDA. <https://www.
fs.fed.us/foresthealth/technology/pdfs/FHTET-2017-02_Biocontrol_
Natural_Areas.pdf>.
Whitten, M.J. (1957). ‘The supposed pre-pupa in cyclorrhaphous
Diptera’. Quarterly Journal of Microscopical Science 98: 241–250.
Fessl, B., Sinclair, B.J. and Kleindorfer, S. (2006). ‘The life-cycle of
Philornis downsi (Diptera: Muscidae) parasitizing Darwin's finches and
its impacts on nestling survival’. Parasitology 133: 739–747.
363
K. Brown, C.B. Phillips, K. Broome, C. Green, R. Toft and G. Walker
Brown, K.; C.B. Phillips, K. Broome, C. Green, R. Toft and G. Walker. Feasibility of eradicating the large white butterfly
(Pieris brassicae) from New Zealand: Data gathering to inform decisions about the feasibility of eradication
Feasibility of eradicating the large white butterfly (Pieris brassicae)
from New Zealand: data gathering to inform decisions about the
feasibility of eradication
K. Brown1, C.B. Phillips2,5, K. Broome1, C. Green1,5, R. Toft3 and G. Walker4
Department of Conservation, PO Box 10-420, Wellington 6143 <kbrown@doc.govt.nz> 2AgResearch, Private Bag,
4749, Christchurch 8140; 3Entecol Ltd, PO Box 2256, Stoke 7041; 4Plant and Food Research, Private Bag 92-169,
Auckland 1142, New Zealand; 5Better Border Biosecurity, <www.b3nz.org>.
1
Abstract Pieris brassicae, large white butterfly, was first found in New Zealand in Nelson in May 2010. The Ministry
for Primary Industries (MPI) responded with a monitoring programme until November 2012 when the Department of
Conservation (DOC) commenced an eradication programme. DOC was highly motivated to eradicate P. brassicae by
the risk it posed to New Zealand endemic cress species, some of which are already nearly extinct. DOC eliminated the
butterfly from Nelson in less than four years at a cost of ca. NZ$5 million. This is the first time globally that a butterfly has
been purposefully eradicated. Variation in estimates of benefits, costs, the efficacy of detection and control tools, and the
probability of eradication success all contributed to uncertainty about the feasibility. Cost benefit analyses can contribute
to assessing feasibility but are prone to inaccurate assumptions when data are limited, and other feasibility questions are
equally important in considering the best course of action. Uncertainty does not equate to risk and reducing uncertainty
through data gathering can inform feasibility and decision making while increasing the probability of eradication success.
Keywords: biodiversity, cost-benefit analysis, eradication success, extinction risk, invasive species, non-native species
INTRODUCTION
Biological invasions by insects, including Lepidoptera,
are increasing worldwide (Liebhold, et al., 2016; Suckling,
et al., 2017). Insect invaders can cause significant
biodiversity, economic, social and health impacts, which
makes eradication an attractive management strategy
(Liebhold, et al., 2016). Expanding international trade and
travel have increased the numbers of exotic organisms
entering New Zealand (Biosecurity Council, 2003; MPI,
2016).
Pieris brassicae (Lepidoptera: Pieridae), a Northern
Hemisphere species native to Eurasia, was first found in
the wild in New Zealand in Nelson (41°27′S, 173°28′E),
a coastal city at the north of the South Island, in May
2010 (Toft, et al., 2012). An Unwanted Organism under
the Biosecurity Act 1993, and a known pest of cultivated
brassicas, it was referred to locally as ‘great white butterfly’
(GWB), or ‘great white cabbage butterfly’ (GWCB’). In
this paper we use the scientific name Pieris brassicae.
Pieris brassicae can migrate hundreds of kilometres to
new locations within a season (Spieth & Cordes, 2012).
Together with the species’ cold tolerance, its dispersal
ability would put most New Zealand endemic brassica
populations at risk (Kean & Phillips, 2013). However, the
rate of P. brassicae spread in Nelson was uncertain. It was
found at eight sites spread over 10–12 km in urban Nelson
five months after the initial detection, but its distribution
appeared not to have changed significantly after a
further two years (Phillips, et al., 2016). This suggested
unexpectedly slow dispersal for this species, perhaps
impeded by parasitic wasps, predation or other factors.
DOC considered that P. brassicae had potential to
cause extinctions of New Zealand endemic cresses, many
of which occur in isolated, small populations; this makes
them vulnerable to a wide range of threats and expensive
to protect. New Zealand has 79 native cress species within
the Brassicaceae family, most of them endemic and two
already presumed extinct. Fifty-five species are currently
threatened by extinction: 18 listed as nationally critical (the
closest ranking to extinction), four nationally endangered,
five nationally vulnerable, one declining, eight naturally
uncommon, and 19 threatened though not yet ranked
(Townsend, et al., 2008; de Lange, et al., 2013; S. Courtney,
DOC, pers. comm.).
After mating, a P. brassicae female lays a cluster of
50–150 eggs on a host plant, and can lay a total of ca.
500 eggs (Gardiner, 1963; Spieth & Schwarzer, 2001).
After hatching, larvae feed together and can wander up to
350 m in search of food plants. Pieris brassicae develop
through five larval stages, usually defoliating several host
plants in the process. Larvae at the fifth stage crawl away
from their host plants to form pupae. The time required
for P. brassicae to complete its lifecycle depends both on
temperature and day length. It had two to three generations
per year in Nelson (Kean & Phillips, 2013).
The Ministry for Primary Industries (MPI) is New
Zealand’s lead biosecurity agency with responsibilities to
protect New Zealand’s environment, economy, health and
socio-cultural values under the Biosecurity Act 1993. MPI
responded quickly to the 2010 detection of P. brassicae in
Nelson by alerting the public and establishing a monitoring
and surveillance programme. However, they terminated
their response in November 2012 based on the results of
the final of several cost benefit analyses (CBA) (Dustow,
2010; Dustow & van Eyndhoven, 2012; Manning, 2012).
MPI predicted costs would outweigh benefits and that the
probability of success was low (Manning, 2012).
The Department of Conservation (DOC) has a
responsibility to protect native biodiversity under the
Conservation Act 1987. DOC has a strong track record of
pest management and successful eradication of (mostly)
vertebrate pests from islands (Diamond, 1990; Simberloff,
2002; Howald, et al., 2007). On 19 November 2012, DOC
initiated an eradication attempt to eliminate the risk that P.
brassicae posed to New Zealand endemic cresses, primarily
using systematic ground-based searching (Phillips, et al.,
2016). The attempt succeeded: the last P. brassicae was
captured near central Nelson on 16 December 2014, and
the eradication programme closed on 4 June 2016. MPI and
DOC declared P. brassicae eradicated from New Zealand
on 23 November 2016, at a cost of NZ$4.97 million (€3.22
million).
DOC and MPI both had very pressing competing
priorities and were acutely aware that spending money on
an eradication attempt would take resources from other
high priority work. To spend limited taxpayer dollars
wisely, MPI responds to incursions according to carefully
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
364
up to meet the challenge, pp. 364–369. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Brown, et al.: Feasibility of large white butterfly eradication
considered priorities using CBA to prioritise management
responses. When considering eradication, MPI calculates a
Benefit Cost Ratio (BCR) by estimating: the pest’s impact
over 20 years, the predicted cost of the eradication attempt,
and the probability of eradication success. A BCR over 3:1
is required for MPI to initiate an eradication attempt.
Unfortunately, elements of a CBA can be difficult to
quantify. Accurately predicting the impacts of invasive
species can be difficult (Andersen, et al., 2004; Paterson, et
al., 2015; Simberloff, et al., 2013; Simberloff, 2015). There
is no universally accepted way of quantifying the benefit of
conserving biodiversity in dollar terms (Spash, 2008; Parks
& Gowdy, 2013; Barkowski, et al., 2015). Predicting
eradication costs (Donlan & Wilcox, 2007; Holmes, et al.,
2015) and the probability of eradication success also pose
challenges, especially when there are few precedents and
limited field data (Pluess, et al., 2012; Brown & Brown,
2015; Phillips, et al., this publication).
A feasibility study that considers eradication costs
and benefits is a routinely used decision support tool
in DOC (Broome, et al., 2005). Before starting the P.
brassicae eradication attempt, DOC also considered costs,
benefits and probability of success, though in a proposed
eradication strategy rather than a CBA (Toft, et al., 2012).
After commencing it, DOC revised costings, procured an
independent CBA (East, 2013a) and developed additional
feasibility criteria (Phillips, et al., this publication)
In this paper we explore uncertainties in the feasibility
and economics of eradicating P. brassicae and suggest
ways of reducing them to help inform future decisionmaking.
METHODS
We examined the question of when to attempt or abandon
eradication when faced with high uncertainty and discuss
ways to assist future decisions in such circumstances.
Cost Benefit Analysis
Four CBAs were developed, three by MPI and one
by the University of New England for DOC. CBA is a
systematic process for calculating and comparing the
costs and benefits of a decision. Written as a formula it
would read: (discounted benefits × probability of success)/
discounted costs. Costs and benefits were discounted at a
rate of 8% for 20 years based on New Zealand Treasury
advice (NZ Treasury, 2005).
The costs of aerial and ground-based eradication
were predicted using known or estimated costs of service
providers. Predictions also drew on experience with
previous eradication operations regarding the activities
required and their likely timeframes. Costs were included
for active surveillance, passive surveillance (media,
public), organism management (insecticide spraying,
etc.), vegetation (host plant) movement controls, host
plant removal and science support (developing a lure,
augmenting natural enemy populations by releasing
parasitic wasps, developing the sterile insect technique,
data analysis, genetics and modelling).
The benefits of eradication are the avoided impacts. The
impacts on brassica seed production, vegetable growing,
and livestock forage production were calculated based on
the cost of applying additional insecticide to control P.
brassicae. These purely monetary impacts were estimated
using several assumed rates of P. brassicae dispersal that
were based on previous observations of P. brassicae spread
in Chile (400 km in seven years), South Africa (350 km
in two years) and Japan (400 km in five years or less)
(Manning, 2012).
The biodiversity impacts (i.e. the cost of applying
insecticide to endemic cresses to control P. brassicae)
were considered by two CBAs (Dustow & van Eyndhoven,
2012; East, 2013a). In both analyses, ‘willingness to pay’
– a non-market valuation method which is based on a New
Zealand community’s willingness to avoid local extinction
of a native plant – was also used to estimate biodiversity
impacts (Dustow & van Eyndhoven, 2012; East, 2013a:
East, 2013b). However, neither Dustow (2010) nor
Manning (2012) used the cost of applying insecticide
to endemic cresses for controlling P. brassicae in their
‘willingness to pay’ calculations.
Criteria used to evaluate eradication feasibility
Feasibility analysis aims to scope the size of the
project, decide if eradication is possible and identify
issues that require resolution to maximise the chance of
eradication success (Pacific Invasives Initiative, 2013)
and thereby estimate the probability of eradication
success. MPI estimated the probability of success of
ground-based eradication at approximately 30% based on
overseas examples and expert opinion (Manning, 2012).
The feasibility criteria used by MPI when considering
eradication probability are based on Bomford & O’Brien
(1995). They are:
● Rate of removal exceeds rate of increase at all
population densities
● Immigration is zero
● All reproductive pests must be at risk
● Target pest can be detected at low densities
● Cost benefit analysis must favour eradication
● Suitable socio-political environment.
DOC assembled a Technical Advisory Group (TAG)
to support the eradication attempt. The TAG developed a
modified set of nine criteria, which built on the six criteria
above, to evaluate feasibility (including the probability of
success) and guide the eradication attempt (Phillips, et al.,
this publication). The criteria used by MPI are discussed
below.
Technical advice and decision making
Both MPI and DOC used in-house and external
expertise to inform decision making. DOC’s TAG
comprised three senior animal pest technical advisors from
DOC including an entomologist, three senior scientists
from two government research institutes (AgResearch and
Plant and Food Research), and a private insect ecology
consultancy (Entecol Ltd). DOC’s TAG had considerable
experience in ground-based eradication having advised or
been directly involved in multiple animal and weed pest
eradications nationally and worldwide. MPI consulted inhouse technical staff, some of whom had been involved
in previous insect eradication programmes, and also held
a day-long meeting to consult with an external group of
insect ecologists and industry stakeholders about the
feasibility of eradicating P. brassicae. An MPI Governance
Group reviewed the evidence provided by technical staff
and decided not to attempt eradication in September
2012. DOC senior managers decided in November 2012
to attempt eradication based on the technical advice they
received (Toft, et al., 2012).
RESULTS
The greatest variation between the four CBAs is in the
predicted costs of eradication and discounted benefits (Table
1). The former due to differences in method and labour
unit cost, and the latter due to the presence or absence of
biodiversity benefit. Probability of success estimates were
365
Island invasives: scaling up to meet the challenge. Ch 2D Other taxa: Invertebrates
relatively similar, although the Manning CBA, which MPI
ultimately used in their decision to abandon eradication,
was somewhat less optimistic.
Eradication feasibility and cost benefit uncertainty
There was uncertainty about P. brassicae’s New
Zealand distribution, reproductive rate, seasonality, rate of
emigration, and host plant range. Similarly, it was difficult
to predict the response of the public to control measures,
efficacy of control measures, efficacy of detection methods,
ability to monitor progress towards eradication, eradication
costs, eradication benefits and probability of success.
Technical assessment of eradication feasibility criteria
Rate of removal exceeds rate of increase at all population
densities
This was unknown at the outset given the potentially
high reproductive capacity. No pheromones or other
chemical attractants were available for P. brassicae,
therefore trapping could not be used as a control tool, nor
as a surveillance tool to monitor changes in population
density. Aerial insecticide application was considered a
potential method of maximising P. brassicae mortality
at all population densities but was not pursued due to its
likely unacceptability to Nelson residents (see criterion 6)
and some uncertainty over just how vulnerable eggs and
larvae would be to aerial spraying of large-leaved host
plants. The large, conspicuous larvae feeding in groups on
the same host plant did, however, suggest ground-based
searching may be effective if the scale of operation could
match the scale of infestation. Also, most P. brassicae
host plants were likely to be low-growing, which would
facilitate ground-based searching.
Immigration is zero
There was concern that the high dispersal potential
of P. brassicae would make delimiting the population
expensive and unreliable (given the unavailability of
effective lures) and could result in undetected populations
occurring outside the operational area that could reinvade.
However, large commercial brassica crops on arable land
near Nelson city were routinely monitored and by 2012
were still not showing evidence of P. brassicae.
All reproductive pests must be at risk
As described in more detail below, most potential
control methods depended on visually detecting P.
brassicae, but search efficacy was initially unknown. Thus,
the possibility that some individuals would evade detection
and avoid control was a major concern.
If P. brassicae populations occurred outside the
operational area and remained undetected, those individuals
would not be at risk, therefore violating criteria 2 and 3
above. Pieris brassicae adults are highly mobile and can
cover long distances in search of larval food plants and
nectar sources. Individuals are known to fly up to 5 km a
day searching for host plants for egg-laying (Schutte, 1966,
cited in Feltwell, 1982). Given the high dispersal potential
of P. brassicae and the observed rapid spread of the closely
related P. rapae when it appeared in New Zealand (at least
160–190 km within two years of detection) (Muggeridge,
1942), it was assumed P. brassicae would be widespread
in Nelson and that undetected populations existed. It was
considered that P. brassicae was capable of moving outside
Nelson city’s boundaries in the first season post-detection.
There was also the risk that P. brassicae could escape
Nelson in association with human transport, perhaps as
larvae on infested vegetation or as pupae on inanimate
objects including vehicles. However, despite the potential
for rapid dispersal beyond Nelson, by 2012 there was still
no evidence that it had occurred. Possibly dispersal was
density-dependent (Toft, et al., 2012).
There was concern that wild brassicas and other food
hosts in less accessible places would act as refugia if they
could not be found and searched.
There was also concern that some life stages would
not be susceptible to control. For example, eggs can
occur under leaves making them difficult to see and less
vulnerable to insecticide sprays. The cryptically coloured
pupae can attach to man-made structures such as fences and
it seemed they would often be difficult to find. However,
every individual could be put at risk during one or more
stages of its lifecycle through human search effort.
In addition, not all tools depended on people detecting
P. brassicae. There was published evidence that eggs and
larvae were vulnerable to storm events, and eggs, larvae
and pupae would be susceptible to parasitism or predation
by various species of parasitic wasps and paper wasps
that were already present in New Zealand (Muggeridge,
1943; Bonnemaison, 1965; Gould & Jeanne, 1984;
Richards, et al., 2016). Moreover, detection was not an
essential prerequisite for applying control measures such
as insecticides and destroying host plants (e.g. garden
brassicas).
Target pest can be detected at low densities
There was concern that visually searching for
P. brassicae without a lure would be costly, labour
intensive and ineffective at detecting all individuals
at low population densities. All previously successful
eradications of Lepidoptera used pheromone lures (Tobin,
et al., 2014). Pheromones can be used to detect and monitor
populations, and also to disrupt mating, which can be a
particularly effective control method at low pest densities.
However, pheromones and other chemical attractants were
unavailable for P. brassicae. Detection probabilities could
be calculated but only through data gathering and analysis
during an eradication attempt (Phillips, et al., 2014a).
Cost benefit analysis must favour eradication
Four separate CBAs were carried out, three before the
eradication attempt commenced and one a year after the
Table 1 Eradication method, cost, benefit and probability of success.
Reference
Method
Dustow (2010)
Dustow & van
Eyndhoven (2012)
Aerial
Aerial
Ground
Ground
Ground
Manning (2012)
East (2013a)
366
Discounted cost Discounted
(NZ$ m)
benefit (NZ$ m)
25–73
21.7–60.9
25–73
21.7–123.2
13.3
8.9
13.2–26.5
3.9
17.4–70.8
Benefit: cost
ratio
0.3–2.44
0.3–4.93
1.64–9.28
1.5–3
4.8–19.7
Probability of
success (%)
50–75%
50–65%
30–60%
56–76%
Brown, et al.: Feasibility of large white butterfly eradication
eradication attempt started. All used different estimates of
costs, benefits and probability of success and, therefore, all
obtained different BCRs and reached different conclusions
(Table 1).
Dustow (2010) concluded that “the analysis strongly
suggests that it is not economically beneficial to attempt to
eradicate great white butterfly [using the aerial application
of insecticide]”. Dustow & van Eyndhoven (2012)
concluded that “the CBA analysis indicates favourable
benefit cost ratios for all but the most conservative groundbased eradication when biodiversity values are excluded”,
and “relatively low biodiversity values are required to
generate favourable benefit cost ratios for many scenarios”.
Manning (2012) concluded that “given the level of
uncertainty surrounding the development of effective
control tools, low probabilities of successfully eradicating
the GWCB, and the uncertainty surrounding biodiversity
benefits, it is unlikely to be technically or economically
feasible to eradicate the GWCB”. Subsequently, East
(2013a) concluded that “The high expected impacts of the
GWB on New Zealand’s native brassicas, the agricultural
industry and home gardeners result in high net present
values and benefit cost ratios [which suggests] that a GWB
eradication programme in Nelson is warranted”.
Manning (2012) stated that ‘the ground-based
eradication option is considered to have a probability of
success of approximately 30% based on overseas examples
and expert opinion”. MPI used the probability of 30%
when decision making. The probability of success value
(mean 56%; range 50–60%) used in the fourth CBA (East
2013a) a year after eradication commenced was determined
by DOC’s TAG who had the benefit of some hard data on
which to make their estimate.
Cost estimates varied greatly among the four CBAs
(Table 1). Aerial spraying costs were based on previous
experience of using this method against white tussock
moth (Orgyia leucostigma) and painted apple moth
(Orgyia anartoides) in Auckland (Ashcroft, et al., 2010)
and assumed substantial social mitigation costs for
affected residents of Nelson. Ground-based cost estimates
were little more than guesses given uncertainty around
method efficacy, delimitation boundaries and detection
probabilities (which strongly influence the length of time
ground crews must remain operational beyond the last
detection to have confidence in declaring eradication
success). MPI contractor costs were also estimated at three
times higher than DOC staff costs. Again, East (2013a) had
some actual data to work with and consequently her cost
estimate came closer than the others to the final actual cost.
Suitable socio-political environment
An aerial application of the bio-pesticide bacterium
Btk (Bacillus thuringiensis kurstaki) was thought likely to
raise considerable public opposition as it did in Auckland
for white tussock moth and painted apple moth (Ashcroft,
et al., 2010). Ground-based control, on the other hand,
was assumed likely to gain public and political support.
This was evidenced shortly after the initial detection by
positive public responses to official requests for reports of
P. brassicae sightings.
DISCUSSION
When assessing the feasibility of eradication, three
basic questions must be answered (Broome, et al., 2005):
Why do it? Can it be done? What will it take to succeed?
Why attempt eradication?
It was impossible to precisely predict the impact of
P. brassicae on New Zealand endemic brassicas (and
predicting impacts on cultivated brassicas under different
management regimes was also problematic). New
Zealand’s biodiversity has been geographically isolated
for millions of years from Northern Hemisphere plants,
herbivores, predators and parasitoids, which makes it hard
to predict impacts. This is a generic issue for incursion
response management in New Zealand. If the New Zealand
native plants that a non-native herbivore will feed on
cannot be immediately identified, then estimating impacts
can only be achieved either through difficult, expensive
and imperfect laboratory testing, or by watching them
unfold in the wild. Laboratory testing of the suitability
of threatened native cresses for herbivory by P. brassicae
was impractical as most are not cultivated due either to the
difficulty of obtaining seed, or to their complex cultivation
requirements.
The risk of extinction to endemic cresses from
herbivory by P. brassicae was considered significant even
without multiple other threats. Other threats to endemic
cresses include herbivory and disturbance by a range of
pests, viral and fungus attack, weed competition, sea-level
rise and the loss of seabird-driven ecosystem processes
which all impact on different cresses. As Quammen (1996)
pointed out, extinction often results from multiple causes
and “to be rare is to have a lower threshold of collective
catastrophe”.
Preventing extinction of native biodiversity is a core
function of DOC and is fundamental to the Department’s
legislative mandate (Conservation Act 1987). Given the
multiple threats facing endemic cresses in addition to P.
brassicae’s potential to access all endemic brassicas,
dietary preference for brassicas, tendency to deposit
large numbers of eggs on individual plants and voracious
feeding on individual plants by clusters of caterpillars,
DOC’s senior botanists and entomologist concluded there
was a high risk that P. brassicae would drive at least some
New Zealand endemic cresses to extinction. Knowledge of
this risk strongly motivated DOC to attempt eradication,
despite the uncertainties, while using a ‘learn by doing’
approach.
Can it be done?
The value attributed to the probability of success can
significantly influence the benefit value obtained (i.e.
benefit = discounted benefit × probability of success) and
therefore the BCR. Estimating the probability of success
is a subjective process based on evidence from previous
eradication attempts and expert opinion. This becomes
problematic when eradication of the taxon in question has
not been attempted before, and where factors including
the ecology, physiology and behaviour of the non-native
species in the new environment are poorly understood.
Accurately estimating the probability of eradication success
is impossible without knowledge of the effectiveness
of control tools, pest population rate of increase, pest
distribution, and risk of immigration (Bomford & O’Brien,
1995; Tobin, et al., 2014; Phillips, et al., this publication).
The challenge is to gather enough quality data quickly
enough to inform decisions. Choosing a threshold of
certainty – where there is enough information to make
a decision – can be partly based on an assessment of the
consequences of not deciding. As Harvey Cox (1968) puts
it “not to decide is to decide”.
If the pest can be killed at the same time as it is being
surveyed for distribution and abundance, then eradication
may gain a ‘head start’ while critical feasibility information
is being collected. Pre-defined stopping rules can be used to
trigger reassessments of feasibility, thus limiting the risk of
over-investing in eradication attempts that cannot succeed.
For example, the DOC TAG defined the following triggers
367
Island invasives: scaling up to meet the challenge. Ch 2D Other taxa: Invertebrates
for re-evaluating the P. brassicae eradication attempt
(Phillips, et al., 2014b):
● If established P. brassicae populations are detected
outside the residential Nelson operational area
● If the population has expanded outwards after 12
months of being subjected to control
● If P. brassicae has not been eradicated by 30 June
2015
● If no P. brassicae have been detected for two
consecutive years.
Triggers clearly indicate when the objectives in the plan
are or are not being achieved. The initial response gathered
some information about P. brassicae’s distribution prior
to commencing the eradication attempt, but not about its
rate of increase or the efficacy of visual searching. Once
the attempt was underway, however, distribution data
and the effectiveness of control tools was gathered in
a systematic way that was used to inform management
decisions, reduce uncertainty, reassess feasibility through
time, measure progress and eventually provide confidence
that eradication had been achieved (Phillips, et al., 2016).
What will it take to succeed?
The ‘learn by doing’ approach informed the technical
assessment of the probability of success (described above).
It also allowed the level of resourcing and capability that
was needed for the eradication to succeed to be accurately
quantified and adjusted as the programme progressed.
For example, in the early stages of the programme in
2012, the ground control team was limited to a team of
four. However, by April 2013 it had become clear that,
although the methods might be effective, more resources
were needed to achieve success (Table 2). The field
team size was increased to 10 (and up to 30 later in the
programme) and the consequent increased costs were
factored into the final CBA (East, 2013a). By constantly
reassessing resource allocations to different aspects of the
project, efficiencies were gained without jeopardising the
probability of success. Crucial in this decision making
was expert analyses of incoming data by DOC’s TAG that
supported the project.
CONCLUSIONS
Delays in attempting eradication can increase the
programme’s duration, cost and risk of failure.
Table 2 Sites searched, sites infested with Pieris brassicae
and proportion infested by financial year (July to June).
2009–10
2010–11
Sites
searched
3
88
Sites
infested
3
30
2011–12
2012–13
2013–14
2014–15
2015–16
Total
76
23,923
80,263
83,118
76,507
263,978
71
1,121
1,490
170
0
2,885
Year
368
Proportion of
sites infested
1
0.341
0.934
0.047
0.019
0.002
0
0.015
Quick, proactive responses can help to achieve
eradication while simultaneously gathering data to inform
decision making. Stopping rules can be used to assess
if an eradication should cease to minimise the waste of
resources.
In the absence of reliable information about costs,
biodiversity benefits and probability of success, CBAs
should not be relied on as the sole decision making tool.
A TAG can be a powerful tool for providing ongoing
well-structured advice to assess feasibility and assist
eradication decision making.
Close engagement with research agencies facilitates
research support for eradication attempts, which can help
to provide critical analyses, information and management
tools.
ACKNOWLEDGEMENTS
We thank the many dedicated DOC and MPI staff
involved, and the many people from outside these
organisations who generously provided their time, advice,
support and expertise, including researchers in the Better
Border Biosecurity research collaboration, <www.b3nz.
org>, particularly from AgResearch and Plant and Food
Research. Vegetables NZ, the TR Ellet Agricultural
Research Trust, Better Border Biosecurity and Dairy NZ
made financial contributions. We also thank Miriam East
(University of New England, Armidale, Australia) for her
work on P. brassicae cost benefit analysis and Shannel
Courtney (DOC, Nelson, NZ) for providing up to date
information on native cresses at risk and costing the likely
P. brassicae impacts on native cresses. Peer review that
improved the quality of this paper was provided by Shannel
Courtney, Bruce Philip (MPI, Wellington, NZ) and Susan
Timmins (DOC, Wellington, NZ). Thanks also to the two
anonymous referees who challenged our thinking.
REFERENCES
Andersen, M. C., Adams, H., Hope, B., & Powell, M. (2004). ‘Risk
assessment for invasive species’. Risk analysis 24(4), 787–793.
Ashcroft, T., Richardson, E., Voice, D. and Dustow, K. (2010). Strategies
and Feasibility to Undertake Eradication of Pieris brassicae (GWCB)
in New Zealand. Unpublished report. Wellington: Ministry for Primary
Industries.
Barkowski, B., Lienhoop, N. and Hansjurgens, B. (2015). ‘Capturing the
complexity of biodiversity: A critical review of economic valuation
studies of biodiversity’. Ecological Economics 113: 1–14.
Biosecurity Council, (2003). Tiakina Aotearoa: Protect New Zealand:
The Biosecurity Strategy for New Zealand. Wellington: Ministry of
Agriculture and Forestry.
Bomford, M. and O’Brien, P. (1995). ‘Eradication or control for vertebrate
pests?’. Wildlife Society Bulletin 23: 249–255.
Bonnemaison, L. (1965). ‘Insect pests of crucifers and their control’.
Annual Review of Entomology 10: 233–256.
Broome, K.G., Cromarty, P. and Cox, A. (2005). ‘Rat Eradications – How
to Get it Right Without a Recipe’. Proceedings of the 13th Australasian
Vertebrate Pest Conference, Te Papa Wellington, New Zealand, 2–6
May 2005, pp. 152–157. Lincoln, NZ: Manaaki Whenua – Landcare
Research.
Brown, K. and Brown, D. (2015). Control to Eradication of Tradescantia
fluminensis on Stephens Island (Takapourewa): The Importance of
Systematic and Persistent Effort. DOC Research and Development
Series 346. Wellington: Department of Conservation.
Cox, H. (1968). On not Leaving it to the Snake. UK: SCM Press.
Brown, et al.: Feasibility of large white butterfly eradication
de Lange, P.J., Rolfe, J.R., Champion, P.D., Courtney, S.P., Heenan,
P.B., Barkla, J.W., Cameron, E.K., Norton, D.A. and Hitchmough,
R.A. (2013). Conservation Status of New Zealand Indigenous Vascular
Plants, 2012. New Zealand Threat Classification Series 3. Wellington:
Department of Conservation.
Diamond, J.M. (1990). ‘New Zealand as an Archipelago: An International
Perspective’. In: D.R. Towns, C.H. Daugherty and I.A.E. Atkinson,
(eds.) Ecological restoration of New Zealand Islands, pp. 3-8.
Conservation Sciences Publication No. 2. Wellington: Department of
Conservation.
Donlan, C.J. and Wilcox, C. (2007). ‘Complexities of costing
eradications’. Animal Conservation 10: 154–156.
Phillips, C.B., Brown, K., Green, C., Walker, G., Broome, K., Toft, R.,
Vander Lee, B., Shepard, M., Bayly, S. and Rees, J. (2014b). Pieris
brassicae (Great White Butterfly) Eradication Annual Report 2013/14.
Unpublished report. Wellington: Department of Conservation.
Phillips, C.B., Brown, K., Green, C., Broome, K., Toft, R., Walker, G.,
Shepard, M., Murphy, J. and Rees, J. (2016). Pieris brassicae (Great
White Butterfly) Eradication Annual Report 2015/16. Unpublished
report. Wellington: Department of Conservation.
Phillips, C.B., Brown, K., Broome, K., Green, C., and Walker, G. (this
publication). ‘Criteria to Help Evaluate and Guide Attempts to Eradicate
Terrestrial Arthropod Pests’.
Dustow, K. (2010). Great White Cabbage Butterfly Cost Benefit Analysis.
Unpublished report. Wellington: Ministry for Primary Industries.
Pluess, T.; Cannon, R.; Jarošík, V.; Pergl, J.; Pyšek, P.; Bacher, S. (2012).
‘When are eradication campaigns successful? A test of common
assumptions’. Biological Invasions 14: 1365–1378.
Dustow, K. and van Eyndhoven, E. (2012). Great White Cabbage Butterfly
Cost Benefit Analysis. Unpublished report. Wellington: Ministry for
Primary Industries.
Quammen, D. (1996). The Song of the Dodo: Island Biogeography in an
Age of Extinction. UK: Pimlico.
East, M. (2013a). A Benefit Cost Analysis for the Great White Butterfly
Incursion in New Zealand. Final report to Department of Conservation.
Armidale, NSW: Institute of Rural Futures, University of New England.
East, M. (2013b). Impact of the Great White Butterfly (Pieris brassicae)
on New Zealand’s Native Cress Species. Final report to Department of
Conservation. Armidale, NSW: Institute of Rural Futures, University
of New England.
Feltwell, J. (1982). Large White Butterfly the Biology, Biochemistry and
Physiology of Pieris brassicae (Linnaeus). The Hague-Boston-London:
Dr W. Junk Publishers.
Gardiner, B. (1963). ‘Genetic and environmental variation in Pieris
brassicae’. Journal of Research on the Lepidoptera 2: 127–136.
Gould, W.P. and Jeanne, R.L. (1984). ‘Polistes wasps (Hymenoptera:
Vespidae) as control agents for lepidopterous cabbage pests.
Environmental Entomology 13: 150–156.
Holmes, N.D., Campbell, K.J., Keitt, B.S., Griffiths, R., Beek, J., Donlan,
C.J. and Broome, K.G. (2015). ‘Reporting costs for invasive vertebrate
eradications’. Biological Invasions. 17: 2913–2925.
Howald, G.C., Donlan, J., Galvan, J.P., Russell, J.C., Parkes, J.,
Samaniego, A., Wang, Y. Veitch, C.R., Genovesi, P., Pascal, M.,
Saunders, A., Tershey, B. (2007). ‘Invasive rodent eradication on
islands’. Conservation Biology 21: 1258–1268.
Kean, J. and Phillips, C. (2013). Phenology and Diapause Research
for Great White Butterfly (Pieris brassicae). Milestone 4: Detailed
Spreadsheet Model for Phenology (Report for Ministry for Primary
Industries No. RE400/2013/481). MPI 16341. Lincoln: AgResearch.
Richards, N.R., Hardwick, S., Toft, R., and Phillips, C.B. (2016).
‘Enhancing parasitism of Pieris brassicae pupae by Pteromalus
puparum’. New Zealand Plant Protection 69: 126–132.
Simberloff, D. (2002). ‘Today Tiritiri Matangi, tomorrow the world! are
we aiming too low in invasives control? Keynote Address’. In: C.R.
Veitch, and M.N. Clout, (eds.) Turning the tide: the eradication of
invasive species, pp. 4–12. Occasional Paper SSC no. 28. IUCN SSC
Invasive Species Specialist Group. IUCN Gland, Switzerland and
Cambridge, UK.
Simberloff, D. (2015). ‘Non-native invasive species and novel
ecosystems’. F1000Prime Reports 7: 47.
Simberloff, D., Martin, J., Genovesi, P., Maris, V., Wardle, D.A., Aronson,
J., Courchamp, F., Galil, B., Garcia-Berthou, E., Pascal, M., Pysek, P.,
Sousa, R., Tabacchi, E. and Vila, M. (2013). ‘Impacts of biological
invasions: what’s what and the way forward’. Trends in Ecology and
Evolution. 28: 58–66.
Spash, C. L. (2008). ‘How much is that ecosystem in the window? The
one with the bio-diverse tail’. Environmental Values 17: 259–284.
Spieth, H.R. and Schwarzer, E. (2001). ‘Aestivation in Pieris brassicae
(Lepidoptera: Pieridae): Implications for parasitism’. European Journal
of Entomology 98: 171–176.
Spieth, H.R. and Cordes, R. (2012). ‘Geographic comparison of seasonal
migration events of the large white butterfly, Pieris brassicae'.
Ecological Entomology 37: 439–445.
Suckling, D. M., Conlong, D. E., Carpenter, J. E., Bloem, K. A.,
Rendon, P., and Vreysen, M. J. B. (2017). ‘Global range expansion
of pest Lepidoptera requires socially acceptable solutions’. Biological
Invasions 19(4), 1107–1119.
Liebhold, A. M., Berec, L., Brockerhoff, E. G., Epanchin-Niell, R. S.,
Hastings, A., Herms, D. A., & Yamanaka, T. (2016). ‘Eradication of
invading insect populations: from concepts to applications’. Annual
Review of Entomology 61: 335–352.
Tobin, P.C., Kean, J.M., Suckling, D.M., McCullough, D.G., Herms,
D.A. and Stringer, L.D. (2014). ‘Determinants of successful arthropod
eradication programs’. Biological Invasions 16: 401–414.
Manning, R. (2012). Biosecurity Response Cost Benefit Analysis: Great
White Cabbage Butterfly. Unpublished report. Wellington: Ministry for
Primary Industries.
Toft, R., Brown, K., Courtney, S. and Green, C. (2012). Proposed
Eradication Strategy for Great White Cabbage Butterfly (Pieris
brassicae). Unpublished report Wellington: Department of Conservation.
MPI. (2016). Biosecurity 2025 Direction Statement for New Zealand’s
biosecurity system. Wellington.
Townsend, A.J., de Lange, P.J., Duffy, C.A.J., Miskelly, C.M., Molloy,
J. and Norton, D.A. (2008). New Zealand Threat Classification System
Manual. Wellington: Department of Conservation. <http://www.doc.
govt.nz/Documents/science-and-technical/sap244.pdf>.
Muggeridge, J. (1942). ‘The white butterfly (Pieris rapae L.). Its
establishment, spread and control in New Zealand’. New Zealand
Journal of Science and Technology 24 (A): 107–129.
NZ Treasury. (2005). Cost Benefit Analysis Primer. Wellington: New
Zealand Government.
Pacific Invasives Initiative. (2013). Resource kit for rodent and cat
eradication. <http://www.pacificinvasivesinitiative.org/rk/intro/index.
html>.
Parks, S. and Gowdy, J. (2013). ‘What have economists learned about
valuing nature? A review essay’. Ecosystem Services 3: e1–e10.
Paterson, R.A., Dick, J.T.A., Pritchard, D.W., Ennis, M., Hatcher,
M.J., and Dunn, A.M. (2015). ‘Predicting invasive species impacts:
a community module function response approach reveals context
dependencies’. Journal of Animal Ecology. 84: 453–463.
Phillips, C., Kean, J., and van Koten, C. (2014a). Detection Rates of Pieris
brassicae Eggs and Larvae (No. RE400/2014/589). Unpublished report.
Lincoln: AgResearch.
369
C. Green
Green, C. Effort required to confirm eradication of an Argentine ant invasion: Tiritiri Matangi Island, New Zealand
Effort required to confirm eradication of an Argentine ant invasion:
Tiritiri Matangi Island, New Zealand
C. Green
Department of Conservation, P.B. 68-908, Newton, Auckland, New Zealand. <cgreen@doc.govt.nz>.
Abstract Tiritiri Matangi Island in the Hauraki Gulf, Auckland, New Zealand is a 220 ha restoration island managed by
the Department of Conservation as an open sanctuary. Following eradication of the only mammalian predator, the Pacific
rat (Rattus exulans) in 1993, a variety of threatened birds, lizards and a giant invertebrate have been transferred to the
island. In March 2000, Argentine ant (Linepithema humile) (Hymenoptera: Formicidae) was discovered and delimiting
surveys revealed a 10 ha infestation. Managers were concerned that the ant could have significant negative impacts on
invertebrates, birds and lizards. Early surveys confirmed a dramatic decline in all other ant species within the infested
area. In February 2001, an eradication programme commenced with paste baits (a.i. 0.01% fipronil) spread manually
in a 2 m × 3 m grid over the entire area. The second year employed a 1 m × 3 m spacing. A second incursion part way
through the programme extended the area to 11 ha. The same toxic bait was used throughout the programme to kill
residual colonies and a non-toxic version was used as a lure to intensively monitor progress. Eradication was declared in
2016. Critical parts of the programme included detection of post treatment survivors and the level of effort required to
confirm successful eradication. New treatment techniques were developed to kill the last small nests by placing toxic baits
inside vials on the ground to prolong bait life. Such nests exhibited non-invasive behaviour, short foraging distances, and
were prone to disturbance leading to foraging cessation. Bait densities and field placement were critical to success. Sites
with residual nests were deemed free of Argentine ant once there had been no detections over three consecutive years of
ongoing monitoring. With few successful Argentine ant eradications in the world the techniques used here can inform and
improve success rates for other ant eradication attempts.
Keywords: Linepithema humile, monitoring vials, paste bait, surveillance, toxic baiting
INTRODUCTION
Argentine ant (Linepithema humile) is one of the world’s
worst invasive ant species and an important conservation
concern (Holway, et al., 2002) with considerable negative
impacts to native biodiversity (Rowles & O’Dowd, 2007;
Stringer, et al., 2009). Argentine ant infestations have
proven difficult to eradicate with few reports of successful
programmes (Silverman & Brightwell, 2008; Hoffmann, et
al., 2011; Hoffman, et al., 2016). To date, only around 10%
of ant eradications have been greater than 10 ha (Hoffmann,
et al., 2016). Detectability of ants in low densities is one
of the most critical factors to increase the likelihood of
successful eradication (Hoffmann, 2011). Despite a long
history of invasive ant management, utilising widely
varying approaches, eradication failures are common
(Hoffmann, et al., 2016).
for invertebrate food sources. Modelling has predicted that
sites near Auckland, including Tiritiri Matangi, are hot
spots for potential Argentine ant occupancy (Pitt, et al.,
2009) with consequent implications for island biosecurity
programmes. Two infestations were found on the island,
one large area covering ca. 10 ha centred around the wharf
and a second, smaller (<0.5 ha) area at Northeast Bay at
the northern end of the island (Fig. 1). The latter arose
A variety of techniques are used to sample Argentine
ant such as visual searching, baits placed on the substrate,
in vials or in pitfall traps (Stanley, et al., 2008; Casellas,
et al., 2009). However, visual detection is less effective in
more complex vegetated environments (Ward & Stanley
2013), such as on offshore islands that act as conservation
sanctuaries, compared to urban areas. A study by Ward
& Stanley (2103) of the detection probability of an
Argentine ant population, using vials with honey and
sausage meat, found that a site should be surveyed three
times to be confident about the presence or absence of
ants. Pest eradication programmes on islands are generally
considered successful if no detections are found during two
years of post-treatment monitoring (Howald, et al., 2007).
Argentine ant was first detected in New Zealand in 1990
(Green, 1990), and subsequently on Tiritiri Matangi Island
in March 2000 (Harris, 2002). Following eradication of the
only mammalian predator, the Pacific rat (Rattus exulans)
from Tiritiri Matangi in 1993, a variety of threatened
birds, lizards and a giant invertebrate were transferred to
the island (Galbraith & Cooper, 2013). Managers were
concerned that the ant could have considerable negative
impacts on invertebrates (Sanders, et al., 2003), birds
(Sockman, 1997; Suarez, et al., 2005) and lizards (Suarez
& Case, 2002) through direct predation and competition
Fig. 1 Tiritiri Matangi Island Argentine ant infestations,
tracks and roads.
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
370
up to meet the challenge, pp. 370–374. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Green: Effort needed to confirm Argentine ant eradication
from the movement of an infested dingy from the wharf
area (Harris, 2002). During the eradication programme a
second incursion occurred at Hobbs Beach in 2008 at the
northern end of the Tiritiri wharf infestation (Fig. 1). The
shape of the newly infested ca. 0.5 ha area indicated its
likely origin as the south end of Hobbs Beach. This new
incursion extended the total infested area on Tiritiri to
11 ha.
Here I outline the programme against Argentine ant on
Tiritiri Matangi Island and describe the effort required to
confirm eradication.
METHODS
Study area
Tiritiri Matangi Island (Fig. 1) is a 220 ha Scientific
Reserve 28 km north of Auckland City in the Hauraki Gulf,
New Zealand, managed by the Department of Conservation
as an open sanctuary. There are over 38,000 visitors to
the island annually (DOC, 2014) with most arriving via
commercial ferry to the wharf but the public are free to
land on the beaches via private craft. The Supporters of
Tiritiri Matangi and the ferry operator facilitate biosecurity
measures relating to clean footwear and public awareness
messaging on pests including Argentine ant. All freight and
goods for island management arrive at the wharf and pest
detection operates via inspections for all species, including
Argentine ants, plus traps or indicator baits for rodents.
The nearest land is the Whangaparaoa Peninsula 3.5 km to
the west. The island is low lying and has been the subject
of an extensive restoration programme involving the
planting of over 280,000 trees over ten years from 1984–
1994 (Galbraith & Cooper, 2013). The two areas infested
with Argentine ants, Northeast Bay and around the Tiritiri
wharf, host a range of plants characteristic of coastal
habitats in the region including flax (Phormium tenax)
(Fig. 2), karamu (Coprosma robusta), taupata (Coprosma
repens), mahoe (Melicytus ramiflorus) and the coastal vine
pohuehue (Meulenbeckia complexa). Typically, the canopy
height was up to 6 m with the occasional pohutukawa tree
(Metrosideros excelsa) exceeding 10 m.
Toxic baiting
Following the discovery of Argentine ant on the
island in 2000, the infestation was delimited using visual
assessment. A ca. 20 m buffer was added to the boundary
of the entire infested area. Bait treatment during the first
two years consisted of a single application of Xstinguish™
Argentine ant bait (a.i. 0.01% fipronil) over the 11 ha
infestation. The paste baits (2–3 g) were hand laid using a
caulking gun to extrude baits on the ground in a grid over
the entire area with 2 m × 3 m and 1 m × 3 m spacing
in February 2001 and the following season in December
2001, respectively. Where possible, baits were placed
under vegetation to avoid exposure to the sun and reduce
desiccation. From 2003, all remnant infestations were
treated twice a year with toxic bait, four to eight weeks
apart. The 2008 incursion (Fig. 1) was double-treated
in 2009 with 1 m × 3 m spacing. From 2010, treatment
consisted of toxic baits placed inside vials (25 mm × 50
mm) on the ground for five days, repeated two weeks later.
Vials were spaced 1 m apart out to 5 m from the remnant
colony then extruded baits on the ground out a further 5
m. Vials had netting covers to prevent lizards and larger
invertebrates entering. Vials were placed in shade beneath
vegetation to reduce desiccation.
All baiting operations were carried out when the
ground was dry and weather conditions were warm (air
temperature 20–25 ºC) dry, and no rain forecast for at least
24 hours. These conditions were optimal for Argentine ant
activity on Tiritiri and thus maximised the chances of bait
detection.
Post-treatment monitoring
Intensive post-treatment monitoring commenced from
2003 using ca. 2 g of non-toxic Xstinguish™ Argentine
ant monitoring paste lure in vials placed every 2–5 m in a
grid over the target area. Although some visual detection
was possible for larger remnant infestations during 2003,
this was largely ineffective at detecting small infestations.
Thus, from 2004 all monitoring used the lure in vials as
above. During 2003, baits were left out for approximately
four hours. From 2004, this was extended to 24 hours.
During collection, the open vials were sealed with a lid and
all trapped ants were identified and later verified using a
microscope.
During the 16-year programme, the entire treated
area was only intensively monitored on two occasions,
in 2006 and 2008. Following the 2006 monitoring the
whole previously infested 10 ha wharf area was assessed
for sites that appeared to be preferred by Argentine ant.
Due to limited resources, during years other than 2006
and 2008 varying levels of monitoring focused on these
preferred Argentine ant sites. In addition, all detection sites
from 2003 onwards were intensively monitored. Due to the
initial very high densities of ants and the ongoing survival
of a few nests, sites close to the wharf were monitored every
year for all 16 years of the programme. Sites where nests
remained undetected in alternate years were monitored two
or three occasions per year to increase the likelihood of
detection.
All lure operations were carried out under the same
environmental conditions as described above for baiting,
i.e. warm and dry.
RESULTS
Fig. 2 Flax (Phormium tenax) (foreground) on beach edge
as a typical preferred habitat for Argentine ants on Tiritiri
Matangi Island.
Toxic baiting was extremely effective at reducing
Argentine ant numbers to very low levels (Fig. 3). No
Argentine ants were seen at Northeast Bay after 2001.
However, remnant populations persisted at the larger wharf
infestation after the initial single treatment per year. Thus,
from 2003 toxic treatments were applied on two occasions
each year, with a period between treatments sufficient to
allow surviving ants to regroup into functioning nests, with
foraging ants susceptible to being attracted to baits. This
371
Island invasives: scaling up to meet the challenge. Ch 2D Other taxa: Invertebrates
Repeated intensive use of lured vials detected surviving
nests. On some occasions, ants were detected in consecutive
vials on adjacent lines indicating a larger population,
likely to be more than one nest. However, detections were
predominantly made in a single vial reflecting the presence
of a single nest. Some of these remnant nests appeared to
be very small as trails featured few ants and vials contained
less than 10 Argentine ants when collected. Much of the
lure was still present indicating a lack of substantial feeding
activity over the 24 hours. In contrast, lure monitoring
early in the programme when large colonies were detected
yielded hundreds of ants with little lure remaining after
four hours.
DISCUSSION
Fig. 3 Percentage of vials with Argentine ants during the
16-year programme. There were zero detections in
2007, 2012, 2014–16.
strategy effectively reduced the infestation to very small
colonies, each consisting of a few nests, and sometimes
just a single nest.
Argentine ants were detected visually for the first three
years of the programme. However, for the remainder of the
programme surviving nests could only be detected by using
lured vials. This was largely due to the complex nature of
the vegetated habitat and the small size of the remnant
colonies. As the programme continued, fewer nests were
detected (Fig. 3).
The entire treated area was intensively monitored in
2006 and revealed six infestations. Following treatment,
no Argentine ants were detected in 2007. Intensive
monitoring of the entire treated area again in 2008 detected
two residual colonies. However, during these years in the
middle of the programme, some small colonies remained
undetected in some years. There were at least two sites
within the wharf infestation where ants appeared to
survive after the 2006 and 2008 treatments as they were
re-detected one or two years later in very similar locations.
One colony was not detected until 2011, despite the site
being monitored annually since 2003. This site was within
5 m of the vehicle and trailer used to transport all arriving
baggage and freight to buildings at the top of the island. All
these surviving colonies vanished after toxic baited vials
were introduced in 2010. Bait inside the vials generally
remained moist and palatable to ants for the full five days.
No Argentine ants were detected after 2014.
The 2008 incursion (Fig. 1) was discovered and doubletreated in 2009. Three surviving colonies were detected in
2010 and two in 2011, with at least one of these being a
survivor from 2010. No Argentine ants were detected
following treatment in 2011.
The 2006 assessment for preferred Argentine ant sites
revealed sites typically characterised by a warm northerly
aspect where sun could reach the ground during much of
the day. Vegetation was less than 3 m tall and usually had
open areas within or adjoining, such as roads, tracks, the
coast or exposed banks with just ground cover vegetation.
Flax plants (Fig. 2) were often a feature of preferred sites
although not a prerequisite.
When Argentine ant was first discovered on the island,
the population density was very high close to the wharf,
which was assumed to be the entry point. Some of the most
problematic nests to destroy were located at sites near
the wharf. Thus, these were monitored on two or three
occasions per year from 2014. However, this repeated
monitoring did not yield any additional detections.
372
A single application of toxic bait was not sufficient
to eradicate Argentine ant from Tiritiri Matangi Island
(Harris, 2002). Although the bait was successful at quickly
eradicating the small, recent infestation at Northeast Bay the
larger wharf infestation required many years of intensive
baiting of small remnant nests to achieve eradication.
Increased levels of effort were required throughout the
programme to improve both ant detection and treatment
techniques to eliminate nests.
Early in the programme when Argentine ant first
established on Tiritiri the species’ behaviour fitted the
usual pattern of being extremely competitive with other
ant species for food sources (Human & Gordon, 1999).
Foraging Argentine ants recruited to any new food source,
including the toxic bait and non-toxic lure, in very large
numbers, often within minutes. This behaviour contrasts
with that of foraging ants from small, post-treatment,
remnant nests that were not necessarily attracted to bait or
the lure given the availability of other natural food sources.
Detectability of ants in low densities is one of the most
critical factors to increase the likelihood of successful
eradication (Hoffmann, 2011).
In the latter stages of the programme there were
occasions when small nests were detected but not seen
again despite there being no toxic treatment in that area
during that season. All single nests appeared to be lacking
the “invasive” element in their behaviour and were
observed foraging over short distances. It is possible that
these ants lacked competitiveness to survive with other ant
species (Rice & Silverman, 2013). All ants in monitoring
vials were identified during the programme and some ant
species were in high numbers. Several of those recorded,
including Monomorium antarcticum, a New Zealand
endemic, and Ochetellus glaber, a naturalised Australian
species, have been shown to be competitive with Argentine
ant (Westermann, et al., 2014).
The lack of competitiveness and aggressive behaviour
normally seen in invasive species made detection of
remnant Argentine ant nests more difficult. It is often
true that the last remaining few in an eradication attempt
require the greatest effort (Morrison, et al., 2007). As the
programme continued, it was necessary to prolong the time
that lured vials were available to foraging ants. While a
four-hour monitoring period was adequate to measure the
level of Argentine ant activity when ants exhibited invasive
behaviour, it became clear later that even 24 hours was
not adequate so needed to be repeated, as recommended
by Ward & Stanley (2013). For the most preferred sites,
particularly on coastal banks exposed to the sun most of
the day, 24-hour monitoring was repeated three times per
season for three seasons to verify eradication.
Ants on trails from small remnant nests often appeared
uninterested in lures or baits even when placed next to the
trail, despite their known palatability as seen early in the
Green: Effort needed to confirm Argentine ant eradication
programme. Argentine ants prefer liquid or paste baits/lures
(Nyamukondiwa & Addison, 2014), but the disadvantage
of such baits/lures is that they have a very short field-life
once applied in the environment. This is especially the case
when used in warm/hot conditions which are optimal for
ant activity. The life of the paste baits when placed on the
ground was short (<12 hrs: Harris, 2002), which gave a
limited time for ants to be attracted to them and commence
feeding. Baits had to compete with other natural food
sources for the attention of ants. In addition, on several
occasions foraging ceased if the trail was disturbed while
placing baits on the ground. This may have contributed to
nest survival at some baited sites.
To increase the time of interaction between ants and
toxin, baits can be delivered repeatedly through the season,
as on Santa Cruz Island (Boser, et al., 2014, Boser, et
al., 2017), or the life of each bait can be extended after
application by slowing desiccation. Baits placed in vials
and shaded under vegetation retained moisture and
remained palatable for at least five days. Hoffman, et al.
(2001; 2016) highlighted the need for new techniques to
eradicate invasive ants. The innovation of placing toxic
baits in vials reported here allowed the potential interaction
between ants and toxic baits to occur over five days rather
than 12 hours. Once toxic baited vials were deployed at
detection sites, no further Argentine ants were seen at these
sites and eradication was achieved.
During the programme there were at least two sites
within the wharf infestation where ants apparently survived
the 2006 and 2008 treatments as they were re-detected one
or two years later in very similar locations. It is possible that
the toxic baiting had sub-lethal effects on either Argentine
ants and/or other ant species, such as M. antarcticum
(Barbieri, et al., 2013), leading to changed interspecific
dynamics and subsequent survival of Argentine ant nests.
It is also possible that a surviving Argentine ant nest moved
away from the monitored area and was not detected until it
moved back in a subsequent season. All ant species readily
move their nests if disturbed and this was observed with
Argentine ants. Trails from surviving small nests were
particularly prone to disturbance. The two sites in the
wharf infestation were on the edge of Wharf Road in highly
preferred locations. They could have moved away from
the edge into less preferred locations beyond monitoring
lines due to disturbance but returned to the edge and were
detected in subsequent seasons.
Toxic vials were used only around the immediate
vicinity of remnant nests to restrict the non-target impacts
on other invertebrates. Relatively few vials had all the
bait removed over the five days. In contrast, non-toxic
lured vials used for monitoring often had much of the
bait removed by non-target species over just 24 hours.
Therefore, it was not worthwhile to leave the monitoring
vials out longer than 24 hours.
Since the eradication of Argentine ants from Tiritiri
Matangi, the island’s biosecurity procedures have altered
to include annual surveillance for any new incursions.
This study has confirmed that ants from new, expanding
populations are readily attracted to baits (Ward & Stanley,
2013), and the level of surveillance monitoring can be
less intensive compared to that required to confirm post
treatment eradication. Early detection of new incursions
through surveillance programmes gives a greater chance
of successful eradication (Clout & Williams, 2009; Ward,
et al., 2010).
There are very few reported, successful Argentine ant
eradications (Silverman & Brightwell, 2008, Hoffmann,
et al., 2011, Hoffmann, et al., 2016). The successful
Argentine ant eradication programme reported here
required considerable effort and improved techniques to
achieve eradication. It took 13 years to extirpate the last
ants from the main infested area near the wharf, which had
areas of very high population density. Problematic remnant
nests were mostly found in these high-density areas.
With the new monitoring and surveillance techniques
developed here, there is confidence that if a new incursion
is detected that eradication will be possible within a
much shorter timeframe, as demonstrated by the 2008
Hobbs Beach incursion site which took only three years.
These techniques would be readily applicable to discrete
Argentine ant populations infesting 10 ha or less elsewhere
in the world, thus achieving an increased success rate of
eradication attempts.
ACKNOWLEDGEMENTS
Thanks to Richard Harris and Jo Rees (Landcare
Research) and the island rangers Ray and Barbara
Walters, Shaun Dunning, and Ian McLeod for assistance
over the early years. Special thanks to Helen Lindsay for
supervision of monitoring teams and relentless assistance
with all aspects of the field programme during the final 10
years. The advice and support of Viv van Dyk on the bait
and lure used is much appreciated. Thanks also to the many
volunteers who helped with the toxic baiting during the
first two years of the programme – success would not have
been possible without your assistance. Suggestions from
two anonymous referees have improved the manuscript.
REFERENCES
Barbieri, R.F., Lester, P.J., Miller, A.S. and Ryan, K.G. (2013). ‘A
neurotoxic pesticide changes the outcome of aggressive interactions
between native and invasive ants’. Proceedings of the Royal Society B:
Biological Sciences, 280(1772): 20132157.
Boser, C.L., Hanna, C., Faulkner, K.R., Cory, C., Randall, J.M. and
Morrison S.A. (2014). ‘Argentine ant management in conservation
areas: results of a pilot study’. Monographs of the Western North
American Naturalist 7: 518–530.
Boser, C.L., Hanna, C., Holway, D.A., Faulkner, K.R., Naughton, I.,
Merrill, K., Randall, J.M., Cory, C., Choe, D.-H., and Morrison S.A.
(2017). ‘Protocols for Argentine ant eradication in conservation areas’.
Journal of Applied Entomology 141: 540–550.
Casellas, D., Gomez, C. and Clavero, M. (2009). ‘Comparing methods of
evaluating the spread of Argentine ants in natural habitats: Pitfall traps
vs. baiting’. Sociobiology 53(3): 927–938.
Clout, M.N. and Williams, P.A. (2009). Invasive Species Management:
A Handbook of Principles and Techniques. Oxford: Oxford University
Press.
DOC. (2014). CMS: Conservation Management Strategy: Auckland
2014–2024. Volume 1. Wellington, New Zealand: Department of
Conservation.
Galbraith, M. and Cooper, H. (2013). ‘Tiritiri Matangi – an overview of
25 years of ecological restoration’. New Zealand Journal of Ecology
37(3): 258–260.
Green, O.R. (1990). ‘Entomologist sets new record at Mt. Smart, or
Iridomyrmex humilis established in New Zealand’. Weta 13:14–16.
Harris, R.J. (2002). ‘Potential Impact of the Argentine Ant (Linepithema
humile) in New Zealand and Options for its Control’. Science
for Conservation 196. Wellington, New Zealand: Department of
Conservation.
Hoffmann, B., Davis, P., Gott, K., Jennings, C., Joe, S., Krushelnycky,
P., Miller, R., Webb, G. and Widmer, M. (2011). ‘Improving ant
eradications: details of more successes, a global synthesis and
recommendations’. Aliens: The Invasive Species Bulletin 31: 16–23.
Hoffmann, B.D. (2011). ‘Eradication of populations of an invasive ant
in northern Australia: successes, failures and lessons for management’.
Biodiversity and Conservation 20(13): 3267–3278.
Hoffmann, B.D., Luque, G.M., Bellard, C., Holmes, N.D. and Donlan,
C.J. (2016). ‘Improving invasive ant eradication as a conservation tool:
A review’. Biological Conservation 198: 37–49.
373
Island invasives: scaling up to meet the challenge. Ch 2D Other taxa: Invertebrates
Holway, D.A., Lach, L., Suarez, A.V., Tsutsui, N.D. and Case, T.J. (2002).
‘The causes and consequences of ant invasions’. Annual Review of
Ecology and Systematics 33(1): 181–233.
Human, K.G. and Gordon, D.M. (1999). ‘Behavioural interactions of the
invasive Argentine ant with native ant species’. Insectes Sociaux 46:
159–163.
Howald, G., Donlan, C.J., Galvan, J.P., Russell, J.C., Parks, J.,
Samaniego, A., Wang, Y., Veitch, D. Genovesi, P., Pascal, M., Saunders,
A. and Tershy, B. (2007). ‘Invasive rodent eradication on islands’.
Conservation Biology 21(5): 1258–1268.
Morrison, S.A., Macdonald, N., Walker, K., Lozier, L. and Shaw, M.R.
(2007). ‘Facing the dilemma at eradication’s end: uncertainty of absence
and the Lazarus effect. Frontiers in Ecology and the Environment 5:
271–276.
Nyamukondiwa, C. and Addison, P. (2014). ‘Food preference and
foraging activity of ants : Recommendations for field applications of
low-toxicity baits’. Journal of Insect Science 14: 14–48.
Pitt, J.P.W., Worner, S.P. and Suarez, A.V. (2009). ‘Predicting Argentine
ant spread over the heterogeneous landscape using a spatially explicit
stochastic model’. Ecological Applications 19(5): 1176–1186.
Rice, E.S. and Silverman, J. (2013). ‘Propagule pressure and climate
contribute to the displacement of Linepithema humile by Pachycondyla
chinensis’. PLOS ONE 8(2): e56281.
Rowles, A.D. and O’Dowd, D.J. (2007).’Interference competition by
Argentine ants displace native ants: implications for biotic resistance to
invasion’. Biological Invasions 9: 73–85.
Sanders, N.J., Gotelli, N.J., Heller, N.E. and Gordon, D.M. (2003).
‘Community disassembly by an invasive ant species’. Proceedings of
the National Academy of Sciences 100: 2474–2477.
Silverman, J. and Brightwell, R.J. (2008). ‘The Argentine ant: Challenges
in managing an invasive unicolonial pest’. Annual Review of Entomology
53: 231–252.
Sockman, K.W. (1997). ‘Variation in life-history traits and nest-site
selection affects risk of nest predation in the California gnatcatcher’.
The Auk 114(3): 324–332.
Stanley, M., Ward, D., Harris, R., Arnold, G., Toft, R. and Rees, J.
(2008). ‘Optimizing pitfall sampling for the detection of Argentine ants,
Linepithema humile (Hymenoptera: Formicidae)’. Sociobiology 51(2):
461–472.
Stringer, L.D., Stephens, A.E.A., Suckling, D.M. and Charles, J.G. (2009).
‘Ant dominance in urban areas’. Urban Ecosystems 12(4): 503–514.
Suarez, A.V. and Case, T.J. (2002). ‘Bottom-up effects on persistence
of a specialist predator: ant invasions and horned lizards’. Ecological
Applications 12(1): 291–298.
Suarez, A.V., Yeh, P. and Case, T.J. (2005). ‘Impacts of Argentine ants on
avian nesting success’. Insectes Sociaux 52: 378–382.
Ward, D.F., Green, C., Harris, R.J. Hartley, S., Lester, P.J., Stanley,
M.C., Suckling, D.M. and Toft R.J. (2010). ‘Twenty years of Argentine
ants in New Zealand : past research and future priorities for applied
management’. New Zealand Entomologist 33: 68–78.
Ward, D.F. and Stanley, M.C. (2013). ‘Site occupancy and detection
probability of Argentine ant populations’. Journal of Applied
Entomology 137(3): 197–203.
Westermann, F.L., Suckling, D.M. and Lester, P.J. (2014). ‘Disruption
of foraging by a dominant invasive species to decrease its competitive
ability’. PLOS ONE 9(3): e90173.
374
C. Herrera, A. Marqués, V. Colomar and M.M. Leza
Herrera, C.; A. Marqués, V. Colomar and M.M. Leza. Analysis of the secondary nest of the yellow-legged
hornet found in the Balearic Islands reveals its high adaptability to Mediterranean isolated ecosystems
Analysis of the secondary nest of the yellow-legged hornet found
in the Balearic Islands reveals its high adaptability to
Mediterranean isolated ecosystems
C. Herrera1, A. Marqués1, V. Colomar2 and M.M. Leza1
Laboratory of Zoology, Department of Biology, University of the Balearic Islands, Cra. Valldemossa km 7.5, CP:
07122 Palma, Illes Balears, Spain. <mar.leza@uib.es>. 2Consortium for the Recovery of the Fauna of the Balearic
Islands (COFIB), Crta. Sineu km 15, CP: 07142 Santa Eugènia, Illes Balears, Spain.
1
Abstract The yellow-legged hornet (Vespa velutina) was detected for the first time in the north of Spain in 2010, but
was not detected in Majorca, Balearic Islands until 2015 and only one secondary nest, with 10 combs, was found in the
northwest of the island. During 2016, nine more nests were found in the same region. To better understand the biology
of V. velutina in isolated conditions, the following objectives were proposed: (I) describe the architecture and structure
of nests; (II) analyse the shape of combs and develop a new method to confirm the circular pattern of breeding; (III)
determine the colony size and (IV) determine the succession of workers and sexual individuals throughout the season.
For these reasons, nests that were removed were frozen for at least 48 days until analysis. Our results show that this
species has a high reproductive potential under isolated conditions. Results reveal that parameters such as weight, height,
diameter, number of cells and total individual production are directly related. Moreover, each mature nest can produce up
to 9,000 individuals and several hundred potential founder queens. All results inform formulation of an efficient control
or eradication programme in the Balearic Islands, as we are in the early stages of invasion and intervention is essential to
eradicate V. velutina on Majorca Island.
Keywords: architecture, breeding provision, caste differentiation, individual production, Latter’s formula, Majorca,
Vespa velutina nigrithorax
INTRODUCTION
The yellow-legged hornet (Vespa velutina Lepeletier
1836) is a social Hymenopteran of the family Vespidae. It
is native to tropical and subtropical areas of Southeast Asia
(Archer, 1994; Martin, 1995; Carpenter & Kojima, 1997).
It was reported for the first time in south-west France in
2004 (Haxaire, et al., 2006; Rome, et al., 2009; Villemant,
et al., 2011) and rapidly spread to nearby European
countries: Spain (Castro & Pagola-Carte, 2010; López,
et al., 2011), Portugal (Grosso-Silva & Maia, 2012), Italy
(Demichelis, et al., 2014), Belgium (Bruneau, 2011; Rome,
et al., 2013) and Germany (Witt, 2015). This species is also
established in South Korea (Choi, et al., 2012; Choi, et al.,
2013) and Japan (Ueno, 2014). The most recent incursion
was in Great Britain in 2016, and Switzerland in 2017 (UK
National Bee Unit, 2016; Budge, et al., 2017).
The introduction of V. velutina to Europe could lead
to important economic and ecological impacts. The main
impact of the yellow-legged hornet is the likely decrease
in honeybee (Apis mellifera) populations (Tan, et al.,
2007; Monceau, et al., 2013a; Monceau, et al., 2013b), as
wasp larvae feed on the proteins of honeybees. Honeybees
are considered one of the most important pollinators for
agriculture, so the decrease of A. mellifera populations
is anticipated to decrease the production of their crops
resulting in economic losses for the farmers (Villemant, et
al., 2011; Arca, et al., 2014). In addition, it is possible that
the yellow-legged hornets attack humans when colonial
nests are established in urban areas (Villemant, et al.,
2006). In the particular case of Majorca, a yellow-legged
hornet invasion could be devastating for the populations
of honeybees, the fragility of the ecosystem (typical of the
island ecosystems) and the impact on endemic insects.
The life cycle of Vespa velutina is annual. In optimal
ambient conditions, when the temperature is high and the
food resources are abundant, one founder queen will build
an embryo nest (Edwards, 1980; Archer, 2010), after that
the workers begin to emerge. In spring the workers build
combs around the embryo nest; this is called the primary
nest. The primary nest has an irregular structure with
the embryo nest in the centre (Spradbery, 1973). During
summer, the colony increases and the primary nest is
left and another nest is built in the same location, if the
conditions are favourable (food resources, temperature,
humidity, etc.). If the ambient conditions are unfavourable
(cold conditions and limited food resources), they build
the secondary nest in a different location, normally in large
trees. This new nest is named the secondary nest and is
larger than the primary nest, with the objective that the
colony increases. The nests of this invasive species are
classified as a calyptodomus type (concealed nest) (Fig.
1), having an external spherical structure, but the combs
are of a conical structure. The upper combs have large
diameters and the lower ones smaller diameters, with a
slight narrowing in the last comb (Jeanne, 1975). When
the reproductive caste emerges in autumn, the nest is
Fig. 1 Calyptodomus nest of V. velutina.
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 375–380. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
375
Island invasives: scaling up to meet the challenge. Ch 2D Other taxa: Invertebrates
called mature, because it is possible to separate males from
females by their morphology (Choi, et al., 2012; Rome, et
al., 2015). The male hornets fertilise new founder queens,
after which the colony dies during the winter. Only new
founder queens survive the winter and build new nests the
following season and start the annual life cycle (Edwards,
1980; Matsuura & Yamane, 1990).
The yellow-legged hornet is established in the northern
regions of Spain (Navarra, Basque Country, Galicia and
Cantabria) (Castro & Pagola-Carte, 2010; López, et al.,
2011), and in Catalonia (Pujade-Villar, et al., 2012). In
2015 it was reported in Soller (Majorca, Balearic Islands).
The hornet was detected by a beekeeper and was identified
by the laboratory of Zoology of the University of Balearic
Islands. Together with the local authorities, an intensive
survey was implemented to detect nests, as is described
in Leza, et al. (2017). In 2015 only one nest of V. velutina
was found in the north-west of the island. However, during
2016 nine more nests were found in the same region. At
this moment, the invasion is in its early stages (Leza, et
al., 2017), and is the first incursion on an island where
eradication through locating and destroying nests can be
used to control the spread; a scenario very different to
mainland Europe. This immediate intervention plays an
important role in the invasion or eradicating the species
on the island.
Although the general structure and production of the
nests of this species has been previously described in Asia
(Spradbery & Kirk, 1978; Matsuura, 1991) and Europe
(Rome, et al., 2011), it is important to study the nests in
local conditions in order to find out if the adaptation of
V. velutina is similar to other regions or if they would
be unable to breed on the island. For this purpose, the
detailed study of the nests found on Majorca (3,667 km2,
situated 176 km from the mainland) can be a useful tool
to understand if this invasive species has the same biotic
fitness or if they have some problems adapting in an
island context. The results could help plan future surveys
and possible dedicated control or eradication measures.
Therefore, the study’s goal was to better understand the
biology of V. velutina in isolated conditions. For this
reason, the following objectives were proposed: (I)
describe the architecture and structure of nests; (II) analyse
Fig. 2 Diagram of a nest comb. The grey cells represent
operculated cells and the lines indicate the three
diameters, which pass through the breeding centre and
separated by 60º, where the stages of the individuals are
determined.
376
the shape of combs and develop a new method to confirm
the circular pattern of breeding; (III) determine the colony
size and (IV) determine the succession of workers and
sexual individuals throughout the season.
MATERIAL AND METHODS
Nest collection
Nine nests were located from August to November
2016, after an active search for nests using the triangulation
method (Leza, et al., 2017). All nests were entirely removed
and frozenfor a minimum of 48 hours. Nests were kept
frozen at -25°C until dissection.
All nests were located in the ‘‘Serra de Tramuntana’’,
in the north-west of Majorca (the exact location is shown
in Table 1). This region has a meso-Mediterranean climate
(Emberger classification), where there is more precipitation
than in other parts of the island (mean of 1,400–1,600 mm
per annum) and cooler temperatures.
Architecture and structure of nests
External morphology of nests was analysed and
described. Weight, height and maximum diameter of each
nest was measured and the number of combs was recorded.
Total weight was the result of the weight of the structure
and its individuals and the total height corresponds to the
height of the whole nest with the external envelope.
Shape of combs
In order to check the circular organisation of the combs
described in other species of wasps (Spradbery & Kirk,
1978; Matsuura, 1991), a new method was proposed. It
follows a similar methodology of comparison between two
sequences of DNA (Brudno, et al., 2004). In our study the
sequences were the diameters of combs and the nucleotides
were the different brood stages, as follows: every
developmental stage (empty cell, egg, larvae, prepupae,
pupae and teneral adult) in cells across three diameters in
each comb was analysed and compared with each other
(the first with the second, the second with the third and the
first with the third) (Figs. 2 & 3).
Each diameter comparison received an arbitrary
categorisation: "2" was assigned if the stage was the same
in cells at the same distance from the breeding centre,
"1" if one of the stages was before or after the other stage
(for example: in one diameter it is a larva and the other a
Fig. 3 Example of the three diameters, aligned in the
breeding centre, and comparing the development stages
at the same distance from the centre. EC = Empty cell,
E = Egg, L = Larva, PP = Prepupa, P = Pupa and TA =
Teneral adult.
Date removed
No. of combs
Weight (g)
Height (cm)
Max diameter (cm)
Workers
Error interval
No. of
females Founders
Total
Males
No. of adults
L. cells
L. eggs
L. larvae
L. meconium
GP
GP/cell
UTM
ID2
39°46'56.30"N
2°45›40.72»E
31/08/2016
5
2,884.12
30.6
25.5
789
0
0
810
0
810
2,712.0
196.5
774.25
796.25
2,577.0
0.95
ID1
39°45'26.33"N
2°42›37.04»E
23/08/2016
6
5,090.0
49.3
33.0
1338
4
0
1350
0
1,350
6,197.75
131.25
1,055.0
2,074.0
4,610.25
0.74
39°46'0.80"N
2°41’4.30”E
8/09/2016
5
3,882.6
41.5
24.6
285
1
1
293
0
293
3,077.25
16.5
947.25
1,202.0
2,458.75
0.80
ID3
39°46'5.80"N
2°44’25.70”E
13/09/2016
5
2,073.7
14.4
24.6
615
0
0
624
0
624
2,778.75
203.0
1,066.75
1,186.75
3,080.5
1.11
ID4
39°44'17.29"N
2°38’7.82”E
14/09/2016
2
136.3
7.0
12.0
127
0
0
127
0
127
200.0
1.25
176.25
0.0
304.5
1.52
ID5
Table 1 Characteristics of the nine nests. L = number obtained by Latter’s formula. GP = individual production.
ID7
ID8
ID9
39°45'36.7"N 39°47'36.5"N 39°47'28.0"N 39°44'14.3"N
2°39›14.3»E 2°46›02.6»E 2°46’03.5”E 2°41’29.5”E
21/09/2016
17/10/2016
2/11/2016
24/11/2016
6
7
5
9
3,313.0
2,893.4
1,448.9
3,583.3
18.3
32.6
18.6
47.6
36.0
23.0
21.2
38.0
684
610
351
151
2
40
15
316
0
0
0
180
701
653
367
656
436
0
134
429
1,137
653
501
1,085
7,008.25
2,916.25
2,412.75
9,355.0
155.5
1.25
74.25
40.25
563.75
809.0
36.2
440.5
3,370.75
1,239.25
1,629.75
7,750.25
5,227.0
2,702.5
2,241.2
9,316.0
0.75
0.93
0.93
1.0
ID6
Herrera, et al.: Analysis of yellow-legged hornet secondary nest
377
Island invasives: scaling up to meet the challenge. Ch 2D Other taxa: Invertebrates
prepupa) and "0" if it does not coincide with any of the
previous cases, as long as both cells have some stage or
are empty (for example: in one diameter it is an egg and
the other a pupa) If one diameter had more cells than the
other, those cells in a diameter that did not have their
partner in the other would not receive any value. For each
diameter comparison the sum of each arbitrary punctuation
was divided by the number of cells multiplied by “2”, the
maximum arbitrary punctuation, obtaining a coincidence
percentage with the circular organisation.
Colony size
The number of cells was estimated with Latter’s
formula (Latter, 1935): N = (3n/2 + 1) ·n/2, where N is
the total number of cells in one comb and n is the number
of cells counted across its maximum diameter. This
formula was extrapolated to estimate the number of eggs,
immature stages (larvae and pupae) and meconium pellets
(meconium is the gut content eliminated immediately by
an individual when moulting from larval to pupal instars
and was recorded only as presence or absence, indicating
that at least one individual had bred). The number of adults
was counted manually.
The estimated total individual production of a nest
was defined as the sum of the estimated number of eggs,
immature stages and meconium pellets, estimated with
Latter’s formula, and adults.
Pearson’s and Spearman’s rank correlations were
made between the estimated individual production and
the following variables: number of combs, weight, height,
diameter, cells, eggs, immature stages or meconium pellets.
Sexual and caste differentiation
Females and males were distinguished by morphological
differences (apex of last sternite bilobate in male but
sharp in female). For females, founders and workers were
distinguished based on their wet weight. Below 593.09 mg
individuals can be considered as workers and individuals
weighing over 593.09 mg can be considered as potential
future queens. The 5% level of uncertainty was reached
beyond 525.44 mg for workers and below 664.84 mg for
founders. Dry and wet weights are strongly correlated (rho
= 0.88, p < 2.2·10-16) with the following linear regression
formula: y (Wwet) = 2.05 · (Wdry) + 80.59 and dry and wet
weights and proved to be useful to discriminate workers
and queens (Rome, et al., 2015). Every female was
weighed with a precision balance (ADAM NBL 423i: 420
g capacity and precision of 0.001 g).
Statistical analysis
RStudio 3.3.2 software (R core team 2016) was used
for analysis. It evaluated the correlation factor and its
significance differences between the nest characteristics
(Kruskal - Wallis and its Dunn post-hoc).
RESULTS
Architecture and structure of nests
Table 1 presents the characteristics of the nine V.
velutina nests collected from August to November 2016 in
Majorca. All the nests found in Majorca analysed in this
work were secondary nests (no embryo nests were found
inside), presented a calyptodomus typology and had ovoid
morphology.
The number of combs within nests analysed ranged
from five to nine (the last one, ID9, found in November),
except for one nest (ID5, found in September) that had
only 2 combs. The weight varied from 136 g (ID5) to 5,090
g (2811,7±482,4), the height from 7 cm (ID5) to 49.3 cm
(28.9±5.1) and the maximum diameter from 12 cm (ID5)
to 38 cm (26.4±2.7) (Table 1).
Shape of combs
The lower combs had a high coincidence percentage
(88.6%) with a circular organisation. However, the
coincidence percentage drops in the upper combs to 62.1%.
A t-test was applied to observe if the mean of coincidence
percentage with a circular organisation of the two upper
combs was different from the two lower combs, which
produced a p = 0.0065, so the circular organisation is lost
ascending in the combs because the percentage in upper
combs was lower than lower combs (Table 2).
Also, in upper combs there is reduced individual
production, in the number of eggs and larvae; lower than
in the first lower comb. In lower combs there are more
immature stages so the individual production moves to the
lower combs. There were significant differences relating to
the number of cells between combs (p = 0.0006), and also
in individual production (p < 0.0001). In both cases, cells
and individual production, the significant differences were
for the first and second lower combs.
Colony size
The evaluation of the total number of cells in the 50
combs of the nine mature nests, using Latter’s formula,
revealed that the number of cells ranged from 200 (ID5) to
9,355 (4,073.1±947.5) and the general production (which is
the sum of the estimated number of eggs, immature stages,
meconium pellets (estimated with Latter’s formula) and
adults) varied between 304.5 (ID5) and 9,316 individuals
(3,613.08±853.67) (Table 1).
Spearman’s rank correlation test showed that diameter
and the general production are highly correlated (rho = 0.895,
p = 0.001) indicating that diameter is a good parameter for
estimating the colony size. The general production follows
the exponential function: y = 0.1778x2.8995, where y is the
Table 2 Values obtained after applying the arbitrary categorisation. Total = sum of coincidence values between two
diameters. Max = sum of the number of paired cells multiplied by the arbitrary categorisation “2”, it corresponds to a
100% coincidence percentage with a circular organisation. Total / Max = index of coincidence to a circular arrangement
of breeding.
Combs
1
Total
15
19
12
Maximum 16
22
14
Total/
0.938 0.864 0.857
Maximum
Mean
0.886
378
26
40
2
27
40
32
44
32
52
3
30
46
22
46
27
50
4
36
58
25
52
27
40
5
24
42
32
52
0.650 0.675 0.727
0.615 0.652 0.478
0.540 0.621 0.481
0.675 0.571 0.615
0.684
0.582
0.547
0.621
Herrera, et al.: Analysis of yellow-legged hornet secondary nest
general production of the nest and x is the largest diameter
of the nest.
Sexual and caste differentiation
Males started to appear during autumn (ID6, ID8 and
ID9), except for ID7, which was removed in October and
had no males (Table 1). Season starts at 21 of September in
north hemisphere and males of these three nests represented
57.9% of the adults found (999 males; 1,724 females in
nests ID6, ID8 and ID9). All other nests (ID1-ID5) were
found and removed before the first fortnight in September,
and had not produced males.
Caste differentiation was determined by weighing all
individuals and a weight increase was observed over time.
Individuals in the last nests collected, weighed 63 mg more
than those in the first nests removed in summer. During
the dissection of the nine nests, a total of 5,581 females
were weighed. The 97.4% of females found in the first
eight nests (ID1-ID8) were workers and only the 1.3% of
the females were in the uncertainty interval. However, in
the last nest found (ID9) the percentage of females in the
uncertainty interval was 48.2%.
DISCUSSION
All the nests found in Majorca during 2016 had a
calyptodomus typology, and the number of combs within
nests analysed ranged from 5 to 9. In comparison, the
general production in mature nests found in France
(4,797.75±606.40) revealed that the analysed nest, under
isolated conditions, had similar production. So, the nests
found in Majorca presented the same morphology as those
nests found in other regions of Europe (Rome, et al., 2015).
Here, we suggest that the diameter of the nest is
a good parameter to estimate the colony size, and the
general production follows the exponential function: y =
0.1778x2.8995, where y is the general production of the nest
and x is the largest diameter of the nest. This is interesting
in order to analyse the fitness of the species, and provide an
easy way to analyse it as, by taking only one measurement
(the largest diameter), the potential of each nest can be
estimated.
Regarding the shape of combs, we provide a new method
to check the circular organisation. The lower combs had a
high coincidence percentage with a circular organisation
and the coincidence percentage drops in the upper combs.
The loss of the circular organisation when ascending in
the combs and the higher number of immature individuals
(such as eggs and larvae), in the lower combs, and pupae
and meconium pellets in upper combs, is due to the fact
that that this species of genus Vespa does not clean the cells
after adult emergence (Janet, 1903), limiting each cell to
produce between one and four individuals (Archer, 2008).
This pattern is similar to the nest structure of V. crabro
(Nadolski, 2012). Other species of Vespa have a higher
number of meconium pellets per cell (Yamane & Makino,
1977; Yamane, 1992; Archer, 1993; Makino & Yamane,
1997), with four as the maximum (Archer, 2011) before the
queen stops laying eggs inside the combs. Moreover, some
authors suggest that un-cleaned combs are the reason the
offspring are found in lower combs, which are cleaner than
upper combs (Janet 1895). Moreover, the nest analysed
presented the lower combs with more immature stages,
so the individual production moves to the lower combs,
which corresponded with Martin (1991, 1992).
Regarding the sexual and caste differentiation, it is
important to note that the method of caste differentiation
used in this work, based on the wet weight, was not very
useful in our analysis as many individuals were in the
error interval, and the increase in weight of 63 mg in
workers was observed in autumn. Other authors proposed
alternative methods for caste differentiation, such as
Perrard, et al. (2012), who used size and nerve structure
of the wings to distinguish individuals. Other possible
methods are by genitalia differentiation or molecular
methods used for other species of Hymenoptera (Barchuk,
et al., 2007). So, for future research we will use these other
methodologies. The presence of males during autumn is
important information for a management plan on the island
as it indicates the possibility that new founder queens can
mate and create new nests the following season, signifying
the beginning of the expansion of the invasive species.
In conclusion, the analysis of the secondary nests of the
yellow-legged hornet found in the Balearic Islands reveals
the high adaptability of this species to Mediterranean
isolated ecosystems, which has important implications for
the development of an effective eradication plan.
ACKNOWLEDGEMENTS
The authors would like to thank members of the
Conselleria de Medi Ambient de les Illes Balears, the
Agents de Medi Ambient and the field technicians of
COFIB for providing the nests used in this study. Thanks
to three anonymous reviewers for constructive feedback on
an earlier draft of this manuscript.
REFERENCES
Arca, M., Papachristoforou, A., Mougel, F., Rortais, A., Monceau, K.,
Bonnard, O. and Arnold, G. (2014). ‘Defensive behaviour of Apis
mellifera against Vespa velutina in France: Testing whether European
honeybees can develop an effective collective defence against a new
predator’. Behavioural Processes 106: 122–129.
Archer, M. (1993). ‘The life history and colonial characteristics of the
hornet, Vespa crabro L. (Hym., Vespinae)’. Entomologist's Monthly
Magazine 129: 151–163.
Archer, M. (1994). ‘Taxonomy, distribution and nesting biology of
the Vespa bicolor group (Hym., Vespinae)’. Entomologist's Monthly
Magazine 130: 149–158.
Archer, M. (2008). ‘Taxonomy, distribution and nesting biology of species
of the genera Provespa Ashmead and Vespa Linneaus (Hymenoptera,
Vespidae)’. Entomologist’s Monthly Magazine 144: 69–101.
Archer, M. (2010). ‘The queen colony phase of vespine wasps
(Hymenoptera, Vespidae)’. Insectes Sociaux 57: 133–145.
Archer, M. (2011). ‘Analysis of a nest of Vespa multimaculata Pérez,
1910’. Malayan Nature Journal 63: 561–567.
Barchuk, A.R., Cristino, A.S., Kucharski, R., Costa, L.F., Simões, Z.L. and
Maleszka, R. (2007). ‘Molecular determinants of caste differentiation
in the highly eusocial honeybee Apis mellifera’. BMC developmental
biology 7: 70.
Brudno, M., Poliakov, A., Salamov, A., Cooper, G.M., Sidow, A., Rubin,
E.M., Solovyev, V., Batzoglou, S. and Dubchak, I. (2004). ‘Automated
whole-genome multiple alignment of rat, mouse, and human’. Genome
Research 14: 685–692.
Bruneau, E. (2011). ‘Le frelon asiatique, déjà là’. ActuApi 55: 1–6.
Budge, G.E., Hodgetts, J., Jones, E.P., Ostoja-Starzewski, J.C., Hall, J.,
Tomkies, V., Semmence, N., Brown, M., Wakefield, M. and Stainton,
K. (2017). ‘The invasion, provenance and diversity of Vespa velutina
Lepeletier (Hymenoptera: Vespidae) in Great Britain’. PLOS ONE 12:
12.
Carpenter, J.M. and Kojima, J. (1997). ‘Checklist of the species in the
subfamily Vespinae (Insectea: Hymenoptera: Vespida)’. Natural History
Bullety Ibaraki University 1: 51–92.
Castro, L. and Pagola-Carte, S. (2010). ‘Vespa velutina Lepeletier,
1836 (Hymenoptera: Vespidae), recolectada en la Península Ibérica’.
Heteropterus Revista de Entomología 10: 193–196.
379
Island invasives: scaling up to meet the challenge. Ch 2D Other taxa: Invertebrates
Choi, M.B., Martin, S.J. and Lee, J.W. (2012). ‘Distribution, spread, and
impact of the invasive hornet Vespa velutina in South Korea’. Journal
of Asia-Pacific Entomology 15(3): 473–477.
Choi, M.B., Lee, S-A., Suk, H.Y. and Lee, J.W. (2013). ‘Microsatellite
variation in colonizing populations of yellow-legged Asian hornet,
Vespa velutina nigrithorax, in South Korea’. Entomological Research
43: 208–214.
Demichelis, S., Manino, A., Minuto, G., Mariotti, M. and Porporato,
M. (2014). ‘Social wasp trapping in north west Italy: comparison of
different bait-traps and first detection of Vespa velutina’. Bulletin of
Insectology 67: 307–317.
Edwards, R. (1980). Social Wasps. Their Biology and Control. Rentokil
Ltd.
of Evolutionary Biology 25: 1389–1398.
Pujade-Villar, J., Torrell, A. and Rojo, M. (2012). ‘Confirmada la presència
a Catalunya d’una vespa originària d’Àsia molt perillosa per als ruscs’.
Butlletí de la Institució Catalana d'Història Natural 77: 173–176.
Rome, Q., Muller, F., Gargominy, O. and Villemant, C. (2009). ‘Bilan 2008
de l’invasion de Vespa velutina Lepeletier en France (Hymenoptera:
Vespidae)’. Bulletin de la Société entomologique de France 114: 297–
302.
Rome, Q., Perrard, A., Muller, F., and Villemant, C. (2011). ‘Monitoring
and control modalities of a honeybee predator, the yellow-legged hornet
Vespa velutina nigrithorax (Hymenoptera: Vespidae).’ Aliens 31: 7–15.
Grosso-Silva, J.M. and Maia, M. (2012). ‘Vespa velutina Lepeletier,
1836 (Hymenoptera, Vespidae), a new species for Portugal’. Arquivos
entomolóxicos 6: 53–54.
Rome, Q., Dambrine, L., Onate, C., Muller, F., Villemant, C., GarcíaPerez, L., Maia, M., Carvalho Esteves, P. and Bruneu, E. (2013). ‘Spread
of the invasive hornet Vespa velutina Lepeletier, 1836, in Europe in
2012 (Hym., Vespidae)’. Bulletin de La Société Entomologique de
France 118: 15–21.
Haxaire, J., Bouguet, J-P. and Tamisier, J-P. (2006) ‘Vespa velutina
Lepeletier, 1836, une redoutable nouveauté pour la faune de France
(Hym., Vespidae)’. Bulletin de la Société entomologique de France 111:
194.
Rome, Q., Muller, F.J., Touret-Alby, A., Darrouzet, E., Perrard, A. and
Villemant, C. (2015). ‘Caste differentiation and seasonal changes in
Vespa velutina (Hym.: Vespidae) colonies in its introduced range’.
Journal of Applied Entomology 139: 771–782.
Janet, C. (1895). ‘Etudes sur les fourmis, les guêpes et les abeilles. 9th
note. Sur Vespa crabro L. Histoire d’un nid depuis son origine (Studies
on ants, wasps and bees. 9th note. About Vespa crabro L. History of a
nest since its origin)’. Mémoires de la Société Zoologique de France
8: 1–140.
Spradbery, J.P. (1973). Wasps: An Account of the Biology and Natural
History of Solitary and Social Wasps with Particular Reference to Those
of the British Isles. Sidgwick and Jackson, London.
Janet, C. (1903). Observations sur les Guêpes. C. Naud, Paris.
Jeanne, R.L. (1975). ‘The adaptiveness of social wasp nest architecture’.
The Quarterly Review of Biology 50(3): 267–287.
Latter, O.H. (1935). ‘A reason for the order in which the queen wasp
constructs the cells of the comb, a method for computing the number
of cells in a comb, and an instance of the limitation of the instincts of
wasps’. Proceedings of the Royal Entomological Society of London 10:
74–78.
Leza, M., Miranda, M.Á. and Colomar, V. (2017). ‘First detection of
Vespa velutina nigrithorax (Hymenoptera: Vespidae) in the Balearic
Islands (Western Mediterranean): a challenging study case’. Biological
Invasions: 1-7.
López, S., González, M. and Goldarazena, A. (2011). ‘Vespa velutina
Lepeletier, 1836 (Hymenoptera: Vespidae): first records in Iberian
Peninsula’. EPPO Bulletin 41: 439–441.
Makino, S. and Yamane, S. (1997). ‘Nest contents and colonial adult
productivity in a common hornet, Vespa simillima simillima Smith,
in northern Japan (Hymenoptera, Vespidae)’. Japanese Journal of
Entomology 65: 47–54.
Martin, S.J. (1991) ‘A simulation model for colony development of the
hornet Vespa simillima (Hymenoptera, Vespidae).’ Japan Journal of
Entomology 59: 105–124.
Martin, S.J. (1992) ‘Development of the embryo nest of Vespa affinis
(Hymenoptera: Vespidae) in Southern Japan.’ Insectes Sociaux 39:
45–57.
Martin, S.J. (1995). ‘Hornets (Hymenoptera: Vespinae) of Malaysia’.
Malayan Nature Journal (Malaysia) 49: 71–82.
Matsuura, M. (1991) ‘Vespa and Provespa.’ In: K.G. Ross and R.W.
Matthews (eds.) The social biology of wasps, pp. 232–262. Cornell
University Press, New York,
Matsuura, M. and Yamane, S. (1990). Biology of the Vespine Wasps.
Springer Verlag, Berlin.
Monceau, K., Maher, N., Bonnard, O. and Thiéry, D. (2013a). ‘Predation
dynamics study of the recently introduced honeybee killer Vespa
velutina: learning from the enemy.’ Apidologie 44: 209–221.
Monceau, K., Arca, M., Leprêtre, L., Mougel, F., Bonnard, O., Silvain,
J.F., Maher, N., Arnold, G. and Thiéry, D. (2013b). ‘Native prey and
invasive predator patterns of foraging activity: the case of the yellowlegged hornet predation at European honeybee hives.’ PLOS One 8:
e66492.
Nadolski, J. (2012). ‘Structure of nests and colony sizes of the European
hornet (Vespa crabro) and Saxon wasp (Dolichovespula saxonica)
(Hymenoptera: Vespinae) in urban conditions.’ Sociobiology 59: 1075–
1120.
Perrard, A., Villemant, C., Carpenter, J.M. and Baylac, M. (2012).
‘Differences in caste dimorphism among three hornet species
(Hymenoptera: Vespidae): forewing size, shape and allometry’. Journal
380
Spradbery, J.P., and Kirk, A.A. (1978). ‘Aspects of the ecology of siricid
woodwasps (Hymenoptera: Siricidae) in Europe, North Africa and
Turkey with special reference to the biological control of Sirex noctilio
F. in Australia.’ Bulletin of Entomological Research, 68(3): 341–359.
Tan, K., Radloff, S.E., Li, J.J., Hepburn, H.R., Yang, M.X., Zhang, L.J.
and Neumann, P. (2007). ‘Beehawking by the wasp, Vespa velutina, on
the honeybees Apis cerana and A. mellifera.’ Naturwissenschaften 94:
469–472.
Ueno, T. (2014). ‘Establishment of the invasive hornet Vespa velutina
(Hymenoptera: Vespidae) in Japan’. International Journal of Chemical,
Environmental and Biological Sciences 2: 220–222.
UK National Bee Unit (2016) <http://www.nationalbeeunit.com/index.
cfm?pageId=208>. Accessed 29 May 2017.
Villemant, C., Haxaire, Je. and Streito, J.C. (2006). ‘Premier bilan de
l’invasion de Vespa velutina Lepeletier en France (Hymenoptera,
Vespidae) (First assessment of the invasion of Vespa velutina
Lepeletier in France (Hymenoptera, Vespidae)’. Bulletin de La Société
Entomologique de France 111: 535–538.
Villemant, C., Barbet-Massin, M., Perrard, A., Muller, F., Gargominy,
O., Jiguet, F. and Rome, Q. (2011). ‘Predicting the invasion risk by the
alien bee-hawking yellow-legged hornet Vespa velutina nigrithorax
across Europe and other continents with niche models’. Biological
Conservation 144: 2142–2150.
Witt, R. (2015). ‘Erstfund eines nestes der Asiatischen nornisse Vespa
velutina Lepeletier, 1838 in Deutschland und details zum nestbau
(Hymenoptera, Vespinae)’. Ampulex 7: 42–53.
Yamane, S. (1992). ‘A huge nest of Vespa basalis collected in Taiwan
(Hymenoptera: Vespidae)’. Chinese Journal of Entomology 12: 1–11.
Yamane, S. and Makino, S.I. (1977). ‘Bionomics of Vespa analis
insularis and V. mandarinia latilineata in Hokkaido, northern Japan,
with notes on vespine embryo nests (Hymenoptera: Vespidae)’. Insecta
Matsumurana. Series Entomology. New Series 12: 1–33.
M. Houghton, A. Terauds and J. Shaw
Houghton, M.; A. Terauds and J. Shaw. Methods for monitoring invertebrate response to vertebrate eradication
Methods for monitoring invertebrate response to vertebrate eradication
M. Houghton1, A. Terauds2 and J. Shaw1
Centre for Biodiversity and Conservation Science, School of Biological Sciences, The University of Queensland,
Brisbane, QLD 4072, Australia. <m.houghton@uq.net.au>. 2Antarctic Conservation and Management, Australian
Antarctic Division, Department of the Environment, 203 Channel Highway, Kingston, TAS 7050, Australia.
1
Abstract Once an island vertebrate eradication is deemed successful, it is typically assumed that ecosystem recovery
will follow. To date, most post-eradication monitoring focuses on the recovery of key threatened or charismatic species,
such as seabirds. Little attention has been given to monitoring and quantifying the response of invertebrate communities.
Rabbits (Oryctolagus cuniculus), house mice (Mus musculus), and ship rats (Rattus rattus) impacted sub-Antarctic
Macquarie Island for over 140 years, with wide ranging ecosystem impacts. In 2014, the eradication of rabbits and rodents
was officially declared successful. To determine whether management objectives are being met, we are investigating the
response of invertebrate communities to pest eradication, using both historic data and contemporary surveys to track
changes over space and time. To achieve this, we have developed a survey strategy that is effective and efficient. Here we
report on the merits of utilising a variety of invertebrate trapping methodologies to establish current baselines for future
invertebrate monitoring. We identify sampling techniques that are most effective for specific groups of taxa, particularly
those of interest to post-eradication monitoring, and how the implementation of such methods can improve and facilitate
effective post-eradication monitoring of invertebrates.
Keywords: baselines, conservation, insects, island, rabbits, restoration, rodents, sub-Antarctic
INTRODUCTION
Island invertebrates are impacted by invasive species,
particularly on remote, unpopulated islands in the Southern
Ocean (Chown, et al., 2008; Angel, et al., 2009; St Clair,
2011: Russell, 2012; Thoresen, et al., 2017). Non-native
plants and invertebrates have been unintentionally
introduced to Southern Ocean Islands (SOI) (Frenot,
et al., 2005; Chown, et al., 2008; Convey & Lebouvier,
2009). Non-native vertebrates have also been introduced,
both intentionally and inadvertently. For example, rabbits
(Oryctolagus cuniculus), cats (Felis catus), dogs (Canis
lupus familiaris), sheep (Ovis aries), goats (Capra
aegagrus hircus), weka (Gallirallus australis), pigs (Sus
scrofa domesticus), brown trout (Salmo trutta) and reindeer
(Rangifer tarandus) were all intentionally introduced
to SOI, whereas ship rats (Rattus rattus), brown rats
(Rattus norvegicus) and house mice (Mus musculus), were
unintentional introductions (Copson & Whinam, 2001;
Courchamp, et al., 2003; Convey and Lebouvier 2009;
McGeoch, et al., 2015). Grazing by non-native vertebrates
on SOI has led to invertebrate habitat modification (Vogel,
et al., 1984; Chapuis, et al., 2004), and direct predation by
rodents has severely modified and depleted invertebrate
populations (Copson, 1986; Chown & Smith, 1993; Angel,
et al., 2009; St Clair, et al., 2011; Russell, 2012; Treasure,
et al., 2014).
Macquarie Island (54.6208° S, 158.8556° E) lies
1,500 km south-east of Tasmania, Australia. The island
is a World Heritage area managed as a Nature Reserve
by the Tasmanian Parks and Wildlife Service (Copson &
Whinam, 2001). Discovered in 1810, the island’s early
human history involved seal harvesting (elephant seals,
Mirounga leonina; fur seals Arctocephalus pusillus,
A. forsteri, A. tropicalis) and penguin harvesting (king
penguins, Aptenodytes patagonicus and royal penguins,
Eudyptes schlegeli). Many non-native species of flora
and fauna were introduced during this time, both
intentionally and inadvertently. Ongoing control of cats
and rabbits by various methods and at varying levels of
effort led to fluctuating populations (Robinson & Copson,
2014; Terauds, et al. 2014). Consequently, native fauna
and vegetation were impacted by varying levels of
predation and grazing (Scott & Kirkpatrick, 2008; Scott
& Kirkpatrick, 2013; Bergstrom, et al., 2009; Whinam,
et al., 2014). Over time, island managers have removed
almost all invasive vertebrate species from Macquarie
Island (Copson & Whinam, 2001), the most recent being
cats in 2000 (Robinson & Copson, 2014) and rabbits and
rodents in 2014 (Springer, 2016). The latter were the target
of the Macquarie Island Pest Eradication programme,
which was the largest multi-species project of its kind at
the time, costing AU$24.5 million. The project’s overall
objective was to ‘…restore Macquarie Island biodiversity
and geodiversity to a natural balance – free of the impacts
of introduced pest species… [with] …vegetation, seabird
and invertebrate populations recovered to levels naturally
supported by the environment’ (Parks and Wildlife, 2009).
We developed a study to assess the success of this project
for invertebrates; specifically, to better understand if
they have ‘recovered’ following removal of mammalian
herbivores and predators, using both historic data and
contemporary surveys.
Invertebrates play a key role in ecosystem function
(Kremen, et al., 1993; Hutcheson, et al., 1999; Gerlach, et
al. 2013). They drive nutrient-cycling and decomposition
processes on SOI (Smith & Steenkamp, 1990; 1992; Smith,
2007; 2008). Thus, establishing a baseline and measuring
their response to ecosystem change informs the state of
the island ecosystem. Many types of invertebrates are
useful proxies for assessing ecosystem change, reflected
in their species richness, species turnover and community
composition (Kremen, et al., 1993; Hutcheson, et al., 1999;
Towns, et al., 2009). Indicator taxa are particularly useful
in monitoring the effects of habitat management at the
ground layer (e.g. ants, millipedes, snails, ground beetles,
some spiders), in foliage (e.g. ants, leaf beetles, some
spiders and moths), and in open habitats (e.g. ants, crickets,
grasshoppers, and butterflies) (Gerlach, et al., 2013).
Moreover, their high density, short life span, ubiquity and
rapid response to changing environmental conditions, make
invertebrates ideal for long-term monitoring (Samways, et
al., 2010; McGeoch, et al., 2011).
Despite their suitability as indicators monitoring of
invertebrates post-eradication is rarely undertaken and
their response to eradication is infrequently determined.
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 381–388. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
381
Island invasives: scaling up to meet the challenge. Ch 2D Other taxa: Invertebrates
Developing appropriate survey methods and sampling
strategies is crucial for a monitoring programme. Here
we test a variety of invertebrate survey techniques
and report on the merits of using specific invertebrate
trapping methodologies to establish baselines for future
invertebrate monitoring and to facilitate comparisons with
previous surveys. Our recent surveys included most of the
invertebrate trapping techniques previously employed on
the island by historical surveys. Our survey design aimed
to measure invertebrate response to vertebrate eradication
and vegetation rehabilitation, track change in invertebrate
community composition and numbers, and establish
baselines for future monitoring. Specifically, our objectives
in this paper, are to 1) compare the efficacy of using
different invertebrate trap types in achieving monitoring
objectives, 2) assess the effectiveness of historical trapping
methods in informing contemporary survey design, and 3)
investigate the benefits and limitations of using historical
data for tracking changes over time. We also discuss how
choosing appropriate methods is a key process for effective
and efficient post-eradication monitoring of invertebrates.
METHODS
Survey design
Determining invertebrate community changes over
time requires definitive and repeatable methods and
detailed site information. Our experimental design
(i.e. our choice of survey/trapping techniques and site
selection) was informed by analysing invertebrate trapping
experimental designs, methods and results from historical
surveys on Macquarie Island. Following a thorough review
of the scientific literature five key resources were selected
to inform our experimental trapping design and methods:
Watson (1967), Greenslade (1987), Anonymous (1993–94,
reported in Stevens, et al., 2010), Davies & Melbourne
(1999), and Stevens, et al., (2010). Each of these historical
surveys utilised different combinations of methods (Table
1). Based on this information, the following survey methods
were tested in our study: pitfall traps, sweep netting,
litter extraction, and timed hand collecting (referred to as
‘20-minute counts’).
Site selection
For this study, sampling was carried out at ten historic
and ten new sites (Fig. 1). This provided 20 sites in total
for the 2015/16 post-eradication survey. The ten new
sites were selected to ensure broader island coverage and
survey additional vegetation communities across the five
dominant vegetation structures on Macquarie Island (based
on Selkirk, et al., 1990) – feldmark (plateau), lower coastal
slopes dominated by Stilbocarpa polaris (Macquarie
Island cabbage), tall grassland (tussock) dominated by Poa
foliosa, short grasslands (including Deschampsia spp.,
Festuca contracta, Agrostis magellanica, Luzula crinita,
Uncinia spp.), and herbfield dominated by Pleurophyllum
hookerii. Most sites were heavily impacted by rabbits in
the past (Bergstrom, et al., 2009; Whinam, et al., 2014).
In total, four Stilbocarpa polaris sites were surveyed in
2015/16, three short grassland sites, five tall grass sites,
four herbfield sites, and four feldmark sites.
Sampling techniques
Five pitfall traps were established at each of the
20 sites, in a line transect along a recorded bearing,
five metres apart. Expert advice from the Tasmanian
Department of Primary Industries, Parks, Water and the
Environment (M. Driessen, pers. comm.), informed the
pitfall trap preparation, spacing, pattern of positioning, and
preservative used. Pitfall traps were constructed of straight
sided, plastic jars 7 cm in diameter, approximately 7 cm
deep, with ca. 1 cm of 100% propylene glycol preservative
added. Pitfall diameter was selected based on other studies
Table 1 Trapping methodology employed during invertebrate sampling studies on Macquarie Island – Watson in 1961
(reported in Watson, 1967), Greenslade in 1986–87 (reported in Greenslade, 1987), Anonymous in 1993–94 (reported in
Stevens, et al., 2010), Davies and Melbourne in 1996 (reported in Davies & Melbourne, 1999), Stevens, et al., in 2009–10
(reported in Stevens, et al., 2010).
Length of sampling
Year-long
Extent of sampling
Island-wide
Greenslade
1986–87
December –
January
Northern sites
# Sites
Not specified
8
4
67
41
# Pitfalls/site
0
5
Not specified
3
3
# Pitfall trap days
-
5–20
30
ca. 42
7–21
Pitfall diameter (cm)
-
1.8
'Large' & 'small'
'Large' & 'small'
0
3
Ethylene glycol/
detergent
1
No
No
Yes, # not specified
No
No
No
0
0
1 over 1 m²
n/a
n/a
Berlese funnels
5
0
0
No
Yes
No
Watson 1961
Pitfall medium
Ethanol
Yes, # not
# Yellow pan trap/site
0
specified
Yes, # not
Vegetation Beating
No
specified
Yes,
#
not
Yes,
# not
Vegetation Sweeping
specified
specified
Yes, # not
Litter volume (Lt)
2–4
specified
Litter extraction method Berlese funnels Berlese funnels
Yes, # not
# Soil Cores
11–16
specified
20-minute counts (hand Yes (not timed) Yes (not timed)
collecting)
382
Anonymous
1993–94
Davies &
Melbourne 1996
Stevens, et al.,
2009–10
Year-long
February –March
October –January
Northern sites
Island-wide
Northern sites
Not specified
Ethanol
0
Houghton, et al.: Monitoring invertebrate response to vertebrate eradication
that have proven the effectiveness of larger trap sizes
(Brennan, et al., 1999; Ward, et al., 2001; Work, et al.,
2002; Woodcock, 2005). Propylene glycol was chosen a
preservative due to it being environmentally benign, highly
viscous and slow to evaporate. Pitfall trap holes were dug
with a soil corer to ensure a snug fit and the pitfall rim was
flush with the ground surface. Where necessary, a small
amount of vegetation was cleared from the trap rim. At the
20 sites, pitfall traps were collected approximately every
10 days between October and December and reset upon
collection for a total period of up to ca. 42 days. Further
pitfall sampling was repeated monthly in January, February
and March for approximately 5–10 days at eight sites.
For litter sampling, at each site, a 1m² quadrat was
used to define the collection area, and three collections
were made of one litre of litter at each site. Litter was
transported back to the station laboratory for sorting and
invertebrate extraction within a maximum of three days
from collection. In feldmark sites where litter was scarce,
litter collection was over 4 m² and up to 1 litre of material
– the exact area and volume was recorded.
Timed counts of 20 minutes were conducted at least
twice over the study period at each site, involving focussed
searching with an aspirator tube and tweezers, collecting
all invertebrates encountered, particularly at the base of
vegetation and under stones.
Sweeping of vegetation tops with nets required
dry conditions with light winds. Hence, sweeping was
conducted opportunistically, at a minimum of twice at each
site over the study period, in all vegetation types regardless
of the canopy (i.e. also in feldmark), by walking slowly
and dragging the net across the vegetation tops 30 times,
on three different trajectories in the site area per sampling
event.
One temperature logger i-button was installed at each
of the 20 sites to monitor microclimate, at the first pitfall
trap of each transect. They were attached approximately 10
cm above the ground surface on a stake with a protective
housing. At each site, site-specific features such as aspect,
altitude, landscape features, vertebrate fauna presence and
vegetation were noted.
Processing and identification
All samples were transferred promptly to 100%
ethanol and transported back to the Australian Antarctic
Division for identification and storage. Using a dissecting
microscope, specimens were counted and identified to
species where possible, except for Acarina, Annelida, and
Nematoda, which were identified to Class or Phylum level.
Data analysis
We undertook preliminary comparisons of the 2015/16
survey data on Order richness and diversity with data
from historical sites established in 1986/87 (Greenslade,
1987). We calculated taxonomic richness and diversity of
invertebrates in different trap types, vegetation groups and
in historical data. For these purposes we pooled data from
different sites to obtain the total number of taxa trapped
in different trap types and vegetation groups. Invertebrate
richness was calculated by summing the total number of
invertebrate Orders recorded in the trap type or vegetation
group of interest. Simpson’s Index of Diversity (SID) was
selected to compare diversity, as it takes into account both
abundance and richness in each habitat. We compared the
Order richness and diversity of our pitfall traps to seven
historical sites and also quantified changes in abundance
for three target groups (beetles, spiders and moths), using a
subset of the historic and contemporary pitfall data.
We analysed data at the level of Order/Class/Phylum
(hereafter referred to as ‘Order’) to facilitate preliminary
comparison with historic data sorted to Order resolution.
For analysis, larval stages and adults were grouped together
for Lepidoptera, Thysanoptera, Coleoptera and Diptera.
RESULTS
Contemporary survey
Fig. 1 Map of 20 invertebrate trapping sites surveyed at
Macquarie Island in 2015/16. All historic sites sampled in
1986/87 (indicated by grey diamonds) were resampled
in 2015/16.
Our preliminary results from the 2015/16 survey
demonstrated that pitfall traps collect the largest number
of individuals – in particular, Collembola (Table 2). Even
when Collembola were removed from the analysis, pitfalls
still collected more individuals than other trapping methods.
Despite the abundance of invertebrates in pitfalls, there
was considerable variance in the nature and abundance of
taxa caught by the different trapping methods, with some
methods proving more effective for specific taxa than
others (Table 2).
Richness (number of different orders caught) and
diversity (SID – richness combined with the relative
383
Island invasives: scaling up to meet the challenge. Ch 2D Other taxa: Invertebrates
Fig. 2 Order richness (summed across 20 sites) and
Simpson’s diversity of four different trapping methods
on Macquarie Island in 2015/16 following mammal
eradication.
Fig. 3 Order richness (summed across 20 sites) of
four trapping types in five vegetation communities
on Macquarie Island in 2015/16 following mammal
eradication.
abundance of the different orders caught) varied between
trapping methods (Fig. 2). The SID demonstrated that
although pitfalls traps yielded the greatest richness, they
also had the lowest diversity, attributable largely to the
dominance of Collembola. Conversely, sweeping had
relatively low species richness but high SID, an indication
of the greater relative abundance of different taxa trapped.
communities. Twenty-minute counts were effective in
feldmark, where richness was proportional to effort. The
low number of taxa in this habitat were found more readily
through this method of focused searching (disturbing
stones and turf), than via passive pitfall trapping or surface
litter collection.
Pitfalls collected the most species regardless of habitat
type (Table 2, Fig. 3). Effectiveness of the other trap
types varied by vegetation community (Fig. 3). Sweeping
vegetation was far more effective in tall grassland and S.
polaris, which are often characterised by dense protective
foliage, than for herbfield and feldmark vegetation, which
typically have more prostrate, sparsely distributed plants.
Litter collection also yielded high relative Order richness,
particularly in tall grassland and short grassland vegetation
The SID of pitfall trap samples across vegetation types
was almost the inverse of their richness (Fig. 4). Across
all vegetation types (except for feldmark), pitfall trapping
diversity was much lower than for other trap types; a
likely reflection of the dominance of the Collembola in
pitfall traps except in feldmark. For short grassland, litter
sampling proved to be exceptionally diverse. Interestingly,
although taxonomic richness of sweeping in herbfield was
relatively low, SID was high. Across all vegetation types,
20-minute counts were almost equal in diversity.
Table 2 The number of individuals from each Order of invertebrates collected via four different trapping methods on
Macquarie Island following mammal eradication: pitfall traps, sweeping, 20-minute counts, and litter collection in the
2015/16 season following mammal eradication.
Order
Gastropoda
Psocoptera
Hemiptera
Thysanoptera
Coleoptera
Diptera
Lepidoptera
Hymenoptera
Isopoda
Araneae
Platyhelminthes
Annelida
Copepoda
Tardigrada
Acarina
Siphonaptera
Nematoda
Collembola
384
Common Name
Snails/slugs
Booklouse
Aphids/Bugs
Thrips
Beetles
Flies
Moths
Wasps
Crustacea
Spiders
Flatworms
Worms
Copepods
Tardigrades
Mites
Fleas
Nematodes
Springtails
TOTAL
Pitfalls
935
44
3
144
2,512
945
3
1
209
2,467
20
284
3,615
69
5,219
1
19
43,641
60,131
Sweep
2
2
0
21
2
61
0
0
1
40
0
3
0
0
40
0
0
277
449
20 minute count
1,294
0
0
4
12
51
4
0
207
169
1
493
2
0
108
0
0
3,609
5,954
Litter
1,019
129
0
4
240
61
8
0
636
380
1
1,489
8
0
340
1
0
5,040
9,356
Houghton, et al.: Monitoring invertebrate response to vertebrate eradication
Fig. 4 Simpson’s Index of Diversity (Order) of four trap
types in five vegetation communities on Macquarie
Island in 2015/16 following mammal eradication.
Comparisons with historical surveys
Preliminary comparisons of our data on Order richness
and diversity data from the 1986/87 sites (Fig. 5) indicate
considerable changes in invertebrate communities since
the earlier surveys. Both Order richness and SID were
generally lower during the earlier sampling period
compared to 2015/16, with the exception of the feldmark
F2 site, where 1986/87 samples were more speciose and
more diverse. Diversity in the tall grassland site P2 and
herbfield H1 were also lower in 2015/16 sampling, though
richness was much higher.
Mouse prey species that were predicted to respond
favourably to mouse removal, such as Coleoptera (beetles),
Lepidoptera (moths) and spiders (Araneae), were trapped
via pitfalls in 1986/87 and again in 2015/16 at seven
sites across five vegetation types (Table 3). Coleoptera
abundance was inconsistently higher in 1986/87, whereas
Araneae were trapped in much higher numbers during the
2015/16 sampling. Lepidoptera were rarely trapped in both
sampling events, present only in the feldmark F2 site.
DISCUSSION
When monitoring ecosystem responses following an
eradication, it is critical to first identify the objectives of
the management intervention. In this case, the facilitation
of the “recovery” of macro-invertebrates on Macquarie
Island was explicit. However, no mechanisms were put in
place to assess the success (or otherwise) of this objective.
Here, our preliminary study tackles the issue of how
to effectively survey a suite of invertebrate species on a
Southern Ocean island to detect recovery and response
of invertebrates after an eradication event, and informs
Fig. 5 (a) Order richness (summed across 20 sites) and,
(b) Simpson’s Index of Diversity of pitfall trapping
(Order level) at seven invertebrate monitoring sites at
Macquarie Island that were first sampled in 1986/87
(prior to mammal eradication) and repeat sampled in
2015/16 (post mammal eradication).
the selection of appropriate survey methods for specific
species.
One of the clearest findings of our study was that
pitfall traps collect the greatest abundance and richness of
invertebrates, particularly Collembola, although they were
the least diverse. Despite the difficulties in comparing
abundance and sampling effort across different techniques
(for example, the longer deployment time of pitfall traps
compared to other trapping techniques), it is apparent that
different trapping methods are more effective at capturing
certain taxonomic groups. This is based on the functional
traits, behaviours and preferred habitats of different taxa.
For example, Psocoptera were primarily collected from
litter samples, as they are detritivores with a preference
for damp conditions under vegetation (Greenslade, 2006).
However, some were also collected during vegetation
sweeping, where they occur in smaller numbers under
leaves (Greenslade, 2006). Tardigrades and Copepods
Table 3 Abundance of Coleoptera, Lepidoptera and Araneae in pitfall traps sampled at
seven sites at Macquarie Island in 1986/87 (prior to mammal eradication) and 2015/16
(post- mammal eradication).
Coleoptera
1986/87 2015/16
P1
P2
S1
H1
H2
F1
F2
Tall grass
Tall grass
Stilbocarpa
Herbfield
Herbfield
Feldmark
Feldmark
25
19
2,426
8
2
4
4
1,909
7
277
6
1
3
0
Lepidoptera
Araneae
1986/87 2015/16 1986/87 2015/16
0
0
0
0
0
0
1
0
0
0
0
0
0
4
151
105
84
27
42
8
28
124
280
175
123
191
120
127
385
Island invasives: scaling up to meet the challenge. Ch 2D Other taxa: Invertebrates
were collected principally via pitfall traps, most likely due
their existence in soil or at the soil surface, particularly
in moist sites. Their small size makes them unlikely to
be detected through other trapping means. Cosmopolitan
groups like Coleoptera, Collembola and Acarina were
detected by all trapping methods. For the Collembola, their
ubiquity in many samples exemplified their abundance and
diversity on the island, with 35 species recorded (Phillips,
et al., 2017). They also occur in a variety of habitats,
from soil-dwellers to canopy species (Greenslade, 2006;
Terauds, et al., 2011). Similarly, the collection of predatory
Staphylinidae coleopterans across all trapping methods
suggests this group are wide-ranging across vegetation,
possibly to maximise opportunities to encounter prey.
Isopoda, Annelidae and Platyhelminthes were collected by
all means except sweeping (with a few exceptions), in line
with their cryptic habits under vegetation, close to the soil
surface and in litter (Greenslade, 2006).
Knowledge of the target group is critical to inform the
experimental design of trapping. For example, and perhaps
counter-intuitively, sweeping did not yield high numbers of
moths or flies. One reason may be that many resident flies
on Macquarie Island are flightless and largely stay close to
the ground (Greenslade 2006). Furthermore, the endemic
moth Eudonia mawsoni, which is not nocturnal, is known
to drop to the ground when dislodged from vegetation (i.e.
by sweep nets) (Jackson, 1995), and often stays close to the
ground, taking shelter in winds over 10 km/hr (Greenslade,
2006). Sweeping can only occur during low wind
conditions, however winds are typically high on the island
(Pendlebury & Barnes-Keoghan, 2007), dispersing many
taxa (both flightless and flying) (Hawes & Greenslade,
2013). The moth’s flight is stimulated by rain, however
sweeping is not possible during rains as wet vegetation
renders the sweep net ineffective. Such background
understanding of target taxa and the environment informs
the design and interpretation of trapping surveys.
If the monitoring or management objectives focus on
a particular group or species it is important to consider
the varying effectiveness of trapping methods (Zou, et
al., 2012). For example, mice on SOI prey mainly on
invertebrates, especially spiders, moths, beetles, aphids,
Orthoptera (e.g. crickets), snails and earthworms (Copson,
1986; Crafford & Sholtz, 1987; Rowe-Rowe, et al., 1989;
Le Roux, et al., 2002; Jones, et al., 2003; Angel, et al.,
2009; St Clair 2011; Russell, 2012). Copson (1986)
identified that mice on Macquarie Island had a particular
preference for spiders (67% of 108 mouse stomach
contents), caterpillars of the endemic moth E. mawsoni
and, to a lesser extent, other invertebrates such as beetles
and dipteran larvae. Therefore, increases in these taxa
following mouse eradication and the removal of predation
pressure could be anticipated. Our preliminary comparisons
provided some support for this hypothesis (see below).
To measure the response of invertebrates preyed upon by
mice on Macquarie Island, our results indicate that pitfall
trapping is effective for spiders and beetles and is therefore
the most efficient mechanism for assessing their recovery.
Monitoring of moths will require greater consideration and
ongoing effort, as they were not detected in high numbers
by any trapping method during the 2015/16 season.
Comparisons to historic datasets are vital to detect
responses to eradication. It is important to consider there
may be different responses and recovery times in different
species. Again, although our comparative analyses are only
preliminary, they do show a higher abundance of spiders in
pitfalls in 2015/16 compared with 1986/87 pitfalls, across
all sites. There is a high likelihood that this is related to
the eradication of mice, given spiders were a major prey
item (Copson, 1986). However, beetle abundance did not
change consistently between the two trapping events, with
386
numbers trapped varying across sites and between years
(Table 3). One possible reason is that Staphylinidae beetles
(which comprised all of the beetles caught) can occur in
dense numbers where rich detritus or rotting material is
present on coastal terraces in vegetation (Greenslade 2006).
As a result, they can be very abundant in an individual
sample from one event, and then absent in others at the
same site. Vegetation recovery is slow, and therefore, if
beetle abundance is driven by vegetation, there will be a
delay in beetle response to rabbit eradication.
For the moth E. mawsoni, despite anecdotal reports
of increased abundance across the island, our preliminary
data do not show this. The moth pitfall counts were similar
in 1986/87 and 2015/16, with adults rarely trapped, which
is consistent with other studies on Macquarie Island
(Jackson, 1995; Potts, 1997; Stevens, et al., 2010; Hawes
& Greenslade, 2013). The low number of adults in our
data could be due to the timing of our sampling regime,
i.e. our trapping effort was low in late December and
early January – the time when adults are most abundant
and active (Watson, 1967). Davies & Melbourne (1999)
captured many adults using yellow pan traps. With this
knowledge, we have added this method to our trapping
regime for future seasons of the invertebrate monitoring
project to identify change. We also extended our future
sampling to occur later in the summer, between January–
March, to identify seasonal changes in species, improve
likelihood of encountering different species, and improve
chances of detecting different life stages in species (such
as the moths and moth larvae). Species life history must be
considered when designing trapping to inform responses to
management.
Terrestrial invertebrate communities that are dependent
on or restricted to specific vegetation or habitat types
are hypothesised to be most likely to be impacted by
rabbit grazing on Macquarie Island (Parks and Wildlife,
2009). Vegetation has undergone considerable changes
between 1986/87 and 2015/16 (Copson & Whinam, 1998;
Bergstrom, et al., 2009; Shaw, et al., 2011). Our preliminary
results highlight the potential utility of historical data,
when combined with targeted and appropriate sampling
techniques, to explore the relationship between vegetation
and invertebrates. Overall, there appears to be an increase
in richness and diversity from 1986/87 to 2015/16. During
the period of the initial sampling in 1986/87, rabbit
numbers and the commensurate vegetation damage were
relatively high (Terauds, et al., 2014), which may explain
the low numbers of vegetation-dependant invertebrates.
Herbfields were favoured by rabbits and heavily impacted
by grazing (Scott, 1988; Selkirk, et al., 1990; Copson &
Whinam, 1998). Herbfield invertebrate communities were
particularly low in richness and diversity in 1986/87.
Feldmark communities vary little between 1986/87 and
2015/16, most likely as rabbit impacts were low in this
high-altitude vegetation group (Copson & Whinam, 1998).
Another important consideration of these comparisons
is that the historical survey data we accessed were
generally based on higher taxonomic groupings, which
may impact our ability to detect subtle changes in
invertebrate communities over time (e.g. Grimbacher, et
al., 2008). Higher taxonomic groupings, unsurprisingly,
do not necessarily reflect the finer details of invertebrate
community assemblages, nor nuanced changes in their
structure and responses (Grimbacher, et al., 2008; Driessen
& Kirkpatrick, 2017) and may aggregate species that have
different ecological or functional traits and responses to
disturbance (Lenat & Resh, 2001; Heino & Soininen, 2007;
Schipper, et al., 2010). A limitation of using historical
datasets is that often clarification and further detail is
simply not available.
Houghton, et al.: Monitoring invertebrate response to vertebrate eradication
Identification of specimens to fine taxonomic resolution,
such as to species level, takes considerable time. Although
we focussed on higher-level taxa here, our forthcoming
analyses at finer taxonomic resolution, (such as to family,
genus and for some groups, species) will provide further
insights on trapping efficacy, survey design and most
importantly, invertebrate community changes over time.
However, there is also good evidence to suggest that
higher-level identification can be appropriate surrogates
for species and effective for detecting major disturbance
events on invertebrate community structure, particularly
where there are significant environmental perturbations
or gradients (Driessen & Włodarska-Kowalczuk & Kędra,
2007; Bevilacqua, et al., 2012; Kirkpatrick, 2017). Sorting
samples to higher taxonomic resolution is often more
practical and cost-effective as it requires less training
and maximises the potential for swift sample processing
(Driessen & Kirkpatrick, 2017). The selected taxonomic
resolution must be balanced between available resources
and the value of results – the decision ultimately depending
on the study objectives (Driessen & Kirkpatrick, 2017). It
is fundamentally important to decide at the outset which
invertebrate taxa, (if not all of them), are the focus of the
monitoring, and what taxonomic resolution will best meet
the objectives of the monitoring. For example, the varying
ranges and habitat associations of the five Staphylinidae
beetle species on Macquarie Island are not represented
when grouped to the Coleoptera order in our data. A second
example is the ubiquitous and numerous Collembola
(springtails), that when aggregated to Order, fail to highlight
the very different species trapped by each medium, such
as those that were trapped via sweeping which inhabit
the canopy, foliage and flowers in vegetation, and those
edaphic groups trapped via pitfalls that either inhabit the
soil or live close to the surface. Such details are important
to our Macquarie Island monitoring objectives as we assess
invertebrate communities in recovering vegetation.
The availability of historical data greatly enhances the
power of long-term effective monitoring. In this instance,
considerable time was invested in tracking down historical
datasets and their metadata. Our future work will include
in-depth analysis of contemporary survey results in relation
to a broader suite of historical data. Whether historical
data are available at the outset or not, establishing a
baseline from which to measure changes into the future
is critical for long-term monitoring, for making informed
management decisions, and assessing management success.
Our preliminary results demonstrate that invertebrate
monitoring in a post-vertebrate eradication ecosystem can
yield important and promising results. Effective monitoring
for invertebrates also leads to improved surveillance
for non-native species arrivals and potential non-native
species impacts. Our future work includes the collection
of two additional years of invertebrate surveys (2016/17
and 2018) across Macquarie Island and the establishment
of four additional invertebrate monitoring sites to improve
island and vegetation community coverage. We will also
employ additional trapping methods (vegetation beating
and yellow-pan trapping), and use Berlese funnels in
the 2016/17 and 2018 surveys for more efficient litter
processing. These improvements combined, will further
develop baseline knowledge of invertebrate communities
on Macquarie Island and inform future monitoring. This
work will provide a comprehensive snapshot of ecosystem
function and recovery following vertebrate eradication. We
will use these results to develop and propose an efficient
means of invertebrate monitoring using specific taxa or
groups as biological indicators of broader ecosystem
changes, to enable robust and efficient monitoring into the
future.
ACKNOWLEDGEMENTS
We thank Michael Driessen for input to research design
and analyses. We thank Jasmine Lee for helpful comments
on the draft manuscript. We are grateful for the helpful
advice and comments given by two anonymous reviewers.
We thank Kimberley Mitchell, Marcus Salton, George
Brettingham-Moore, Rowena Hannaford, Jacqueline
Comery and Penelope Pascoe for their assistance in the
field with invertebrate trapping. We thank the Australian
Antarctic Division for logistic support and the Tasmanian
Parks and Wildlife Service for granting access to the
island and permits to collect invertebrates (this work
under Scientific Permit FA15234). This study was
supported by funding from the Australian Government’s
National Environmental Science Programme through the
Threatened Species Recovery Hub and the Australian
Antarctic Science program (AAS 4305).
REFERENCES
Angel, A., Wanless, R.M. and Cooper, J. (2009). ‘Review of impacts of
the introduced house mouse on islands in the Southern Ocean: Are mice
equivalent to rats?’ Biological Invasions 11(7): 1743–1754.
Bergstrom, D.M., Lucieer, A., Kiefer, K., Wasley, J., Belbin, L., Pedersen,
T.K. and Chown, S.L. (2009). ‘Indirect effects of invasive species
removal devastate world heritage island’. Journal of Applied Ecology
46(1): 73–81.
Bevilacqua, S., Terlizzi, A., Claudet, J., Fraschetti, S. and Boero, F.
(2012). ‘Taxonomic relatedness does not matter for species surrogacy
in the assessment of community responses to environmental drivers’.
Journal of Applied Ecology 49(2): 357–366.
Brennan, K.E., Majer, J.D. and Reygaert, N. (1999). ‘Determination of
an optimal pitfall trap size for sampling spiders in a Western Australian
jarrah forest’. Journal of Insect Conservation 3(4): 297–307.
Chapuis, J.-L., Frenot, Y. and Lebouvier, M. (2004). ‘Recovery of native
plant communities after eradication of rabbits from the subantarctic
Kerguelen Islands, and influence of climate change’. Biological
Conservation 117(2): 167–179.
Chown, S.L. and Smith, V.R. (1993). ‘Climate change and the short-term
impact of feral house mice at the sub-Antarctic Prince Edward Islands’.
Oecologia 96(4): 508–516.
Chown, S.L., Lee, J.E. and Shaw, J.D. (2008). ‘Conservation of
Southern Ocean islands: Invertebrates as exemplars’. Journal of Insect
Conservation 12(3): 277–291.
Copson, G. R. (1986). ‘The diet of the introduced rodents Mus musculus
L. and Rattus rattus L on sub-Antarctic Macquarie Island’. Wildlife
Research 13(3): 441–445.
Copson, G. and Whinam, J. (1998). ‘Response of vegetation on
subantarctic Macquarie Island to reduced rabbit grazing’. Australian
Journal of Botany 46(1): 15–24.
Copson, G. and Whinam, J. (2001). ‘Review of ecological restoration
programme on subantarctic Macquarie Island: Pest management
progress and future directions’. Ecological Management and
Restoration 2(2): 129–138.
Courchamp, F., Chapuis, J.-L. and Pascal, M. (2003). ‘Mammal invaders
on islands: Impact, control and control impact’. Biological Reviews
78(03): 347–383.
Crafford, J.E. and Scholtz, C.H. (1987). ‘Quantitative differences between
the insect faunas of sub-antarctic Marion and Prince Edward islands:
A result of human intervention?’. Biological Conservation 40(4): 255–
262.
Davies, K.F. and Melbourne, B.A. (1999). ‘Statistical models of
invertebrate distribution on Macquarie Island: A tool to assess climate
change and local human impacts’. Polar Biology 21: 240–250.
Driessen, M.M. and Kirkpatrick, J.B. (2017). ‘Higher taxa can be effective
surrogates for species‐level data in detecting changes in invertebrate
assemblage structure due to disturbance: a case study using a broad
range of orders’. Austral Entomology: doi.org/10.1111/aen.12315.
Frenot, Y., Chown, S.L., Whinam, J., Selkirk, P.M., Convey, P., Skotnicki,
M. and Bergstrom, D.M. (2005). ‘Biological invasions in the Antarctic:
Extent, impacts and implications’. Biological Reviews 80(1): 45–72.
Gerlach, J., Samways, M. and Pryke, J. (2013). ‘Terrestrial invertebrates
as bioindicators: An overview of available taxonomic groups’. Journal
of Insect Conservation 17: 831–850.
387
Island invasives: scaling up to meet the challenge. Ch 2D Other taxa: Invertebrates
Greenslade, P. (1987). Report on invertebrate fieldwork Macquarie
Island, December 1986 – January 1987. Unpublished report. Tasmania,
Australia: Department of Primary Industry, Parks, Water and the
Environment.
Scott, J.J. and Kirkpatrick, J.B. (2013). ‘Changes in the cover of plant
species associated with climate change and grazing pressure on the
Macquarie Island coastal slopes, 1980–2009’. Polar Biology 36(1):
127–136.
Greenslade, P. (2006). The Invertebrates of Macquarie Island. Kingston:
Australian Antarctic Division.
Selkirk, P., Seppelt, R. and Selkirk, D. (1990). Subantarctic Macquarie
Island: Environment and Biology. Cambridge: Cambridge University
Press.
Grimbacher, P.S., Catterall, C.P. and Kitching, R.L. (2008). ‘Detecting
the effects of environmental change above the species level with beetles
in a fragmented tropical rainforest landscape’. Ecological Entomology
33(1): 66–79.
Hawes, T.C. and Greenslade, P. (2013). ‘The aerial invertebrate fauna of
subantarctic Macquarie Island’. Journal of Biogeography 40(8): 1501–
1511.
Heino, J. and Soininen, J. (2007). ‘Are higher taxa adequate surrogates
for species-level assemblage patterns and species richness in stream
organisms?’. Biological Conservation 137(1): 78–89.
Hutcheson, J., Walsh, P. and Given, D.R. (1999). ‘Potential Value of
Indicator Species for Conservation and Management of New Zealand
Terrestrial Communities’. Science for Conservation 109. Wellington,
New Zealand: Department of Conservation.
Jackson, J. (1995). ‘Notes on the biology of Eudonia mawsoni
(Lepidoptera: Pyralidae) on subantarctic Macquarie Island’. Polar
Biology 15(4): 289–294.
Jones, A.G., Chown, S.L., Webb, T.J. and Gaston, K.J. (2003). ‘The
free-living pterygote insects of Gough Island, South Atlantic Ocean’.
Systematics and Biodiversity 1(2): 213–273.
Kremen, C., Colwell, R.K., Erwin, T.L., Murphy, D.D., Noss, R.F. and
Sanjayan, M.A. (1993). ‘Terrestrial arthropod assemblages: Their use in
conservation planning’. Conservation biology 7(4): 796–808.
Le Roux, V., Chapuis, J.-L., Frenot, Y. and Vernon, P. (2002). ‘Diet of the
house mouse (Mus musculus) on Guillou Island, Kerguelen archipelago,
Subantarctic’. Polar Biology 25(1):49–57.
Lenat, D.R. and Resh, V.H. (2001). ‘Taxonomy and stream ecology—the
benefits of genus-and species-level identifications’. Journal of the North
American Benthological Society 20(2): 287–298.
McGeoch, M.A., Sithole, H., Samways, M.J., Simaika, J.P., Pryke, J.S.,
Picker, M., Uys, C., Armstrong, A.J., Dippenaar-Schoeman, A.S. and
Engelbrecht, I.A. (2011). ‘Conservation and monitoring of invertebrates
in terrestrial protected areas’. Koedoe 53(2): 131–143.
McGeoch, M.A., Shaw, J.D., Terauds, A., Lee, J.E. and Chown, S.L.
(2015). ‘Monitoring biological invasion across the broader Antarctic: A
baseline and indicator framework’. Global Environmental Change 32:
108–125.
Parks and Wildlife. (2009). Macquarie Island Pest Eradication Plan. Part
C: Environmental Impact Statement. Hobart, Tasmania: Department of
Primary Industries, Parks, Water and the Environment.
Pendlebury, S. and Barnes-Keoghan, I. (2007) ‘Climate and climate
change in the sub-Antarctic’. Papers and Proceedings of the Royal
Society of Tasmania 141(1): 67–81.
Phillips, L., Janion-Scheepers, C., Houghton, M., Terauds, A., Potapov, M.
and Chown, S.L. (2017). ‘Range expansion of two invasive springtails
on sub-Antarctic Macquarie Island’. Polar Biology: 1–6.
Potts, T. (1997). ‘Monitoring Invertebrate Migration and Dispersal
on Subantarctic Macquarie Island’. BSc Honours Thesis. Hobart:
University of Tasmania, Institute of Antarctic and Southern Ocean
Studies (IASOS).
Robinson, S.A. and Copson, G.R. (2014). ‘Eradication of cats (Felis
catus) from subantarctic Macquarie Island’. Ecological Management
and Restoration 15(1): 34–40.
Rowe-Rowe, D., Green, B. and Crafford, J. (1989). ‘Estimated impact of
feral house mice on sub-Antarctic invertebrates at Marion Island’. Polar
Biology 9(7): 457–460.
Russell, J.C. (2012). ‘Spatio-temporal patterns of introduced mice and
invertebrates on Antipodes Island’. Polar Biology 35(8):1187–1195.
Samways, M.J., McGeoch, M.A. and New, T.R. (2010). Insect
Conservation: A Handbook of Approaches and Methods. Oxford:
Oxford University Press.
Schipper, A.M., Lotterman, K., Geertsma, M., Leuven, R.S. and
Hendriks, A.J. (2010). ‘Using datasets of different taxonomic detail to
assess the influence of floodplain characteristics on terrestrial arthropod
assemblages’. Biodiversity and Conservation 19(7): 2087–2110.
Scott, J. (1988) ‘Rabbit distribution history and related land disturbance,
Macquarie Island’. Papers and Proceedings of the Royal Society of
Tasmania 122(1): 255–266.
Scott, J. and Kirkpatrick, J. (2008). ‘Rabbits, landslips and vegetation
change on the coastal slopes of subantarctic Macquarie Island, 1980–
2007: implications for management’. Polar Biology 31(4): 409–419.
388
Shaw, J., Terauds, A. and Bergstrom, D. (2011). ‘Rapid commencement
of ecosystem recovery following aerial baiting on sub-Antarctic
Macquarie Island’. Ecological Management and Restoration 12(3):
241–244.
Smith, V.R. and Steenkamp, M. (1990). ‘Climatic change and its
ecological implications at a subantarctic island’. Oecologia 85: 14–24.
Smith, V.R. and Steenkamp, M. (1992). ‘Soil macrofauna and nitrogen on
a sub-Antarctic island’. Oecologia 92(2): 201–206.
Smith, V.R. (2007). ‘Introduced slugs and indigenous caterpillars as
facilitators of carbon and nutrient mineralisation on a sub-Antarctic
island’. Soil Biology and Biochemistry 39: 709–713.
Smith, V.R. (2008). ‘Energy flow and nutrient cycling in the Marion
Island terrestrial ecosystem: 30 years on’. Polar Record 44(3): 211–226.
Springer, K. (2016). ‘Methodology and challenges of a complex multispecies eradication in the sub-Antarctic and immediate effects of
invasive species removal’. New Zealand Journal of Ecology 40(2):
273–278.
St Clair, J.J.H. (2011). ‘The impacts of invasive rodents on island
invertebrates’. Biological Conservation 144(1): 68–81.
St Clair, J.J.H., Poncet, S., Sheehan, D.K., Székely, T. and Hilton, G.M.
(2011). ‘Responses of an island endemic invertebrate to rodent invasion
and eradication’. Animal Conservation 14(1): 66–73.
Stevens, M., Hudson, P., Greenslade, P. and Potter, M. (2010). Report on
invertebrate monitoring of long term field sites on Macquarie Island
2009/2010. Unpublished report. South Australia: South Australia
Museum.
Terauds, A., Chown, S.L. and Bergstrom, D.M. (2011). ‘Spatial scale and
species identity influence the indigenous–alien diversity relationship in
springtails’. Ecology 92(7): 1436–1447.
Terauds, A., Doube, J., McKinlay, J. and Springer, K. (2014). ‘Using
long-term population trends of an invasive herbivore to quantify the
impact of management actions in the sub-Antarctic’. Polar Biology
37(6): 833–843.
Thoresen, J.J., Towns, D., Leuzinger, S., Durrett, M., Mulder, C.P. and
Wardle, D.A. (2017). ‘Invasive rodents have multiple indirect effects on
seabird island invertebrate food web structure’. Ecological Applications
27(4): 1190–1198.
Towns, D.R., Wardle, D.A., Mulder, C.P., Yeates, G.W., Fitzgerald, B.M.,
Parrish, R.G., Bellingham, P.J. and Bonner, K.I. (2009). ‘Predation
of seabirds by invasive rats: Multiple indirect consequences for
invertebrate communities’. Oikos 118(3): 420–430.
Treasure, A.M., Chown, S.L. and Diez, J. (2014). ‘Antagonistic effects
of biological invasion and temperature change on body size of island
ectotherms’. Diversity and Distributions 20(2): 202–213.
Vogel, M., Remmert, H. and Lewis-Smith, R.I. (1984). ‘Introduced
reindeer and their effects on the vegetation and the epigeic invertebrate
fauna of South Georgia (subantarctic)’. Oecologia 62(1): 102–109.
Ward, D.F., New, T.R. and Yen, A.L. (2001). ‘Effects of pitfall trap
spacing on the abundance, richness and composition of invertebrate
catches’. Journal of Insect Conservation 5(1): 47–53.
Watson, K.C. (1967). ‘The terrestrial arthropoda of Macquarie Island’.
ANARE Scientific Reports, Series B (1) Zoology 99: 1–90.
Whinam, J., Fitzgerald, N., Visoiu, M. and Copson, G. (2014). ‘Thirty
years of vegetation dynamics in response to a fluctuating rabbit
population on sub‐Antarctic Macquarie Island’. Ecological Management
and Restoration 15(1): 41–51.
Włodarska-Kowalczuk, M. and Kędra, M. (2007). ‘Surrogacy in natural
patterns of benthic distribution and diversity: selected taxa versus lower
taxonomic resolution’. Marine Ecology Progress Series 351: 53–63.
Woodcock, B.A. (2005). ‘Pitfall Trapping in Ecological Studies’. In: S.
Leather (ed.) Insect Sampling in Forest Ecosystems, pp. 37–57. Oxford:
Blackwell.
Work, T.T., Buddle, C.M., Korinus, L.M. and Spence, J.R. (2002).
‘Pitfall trap size and capture of three taxa of litter-dwelling arthropods:
Implications for biodiversity studies’. Environmental Entomology
31(3): 438–448.
Zou, Y., Jinchao, F., Dayuan, X., Weiguo, S. and Axmacher, J.C. (2012).
‘A comparison of terrestrial arthropod sampling methods’. Journal of
Resources and Ecology 3(2): 174–182.
N. Maczey, D. Moore, P. González-Moreno and N. Rendell
Maczey, N.; D. Moore, P. González-Moreno and N. Rendell. Introduction of biological control
agents against the European earwig (Forficula auricularia) on the Falkland Islands
Introduction of biological control agents against the European earwig
(Forficula auricularia) on the Falkland Islands
N. Maczey1, D. Moore1, P. González-Moreno1 and N. Rendell2
CABI, Bakeham Lane, Egham, Surrey, TW20 9TY, UK. <n.maczey@cabi.org>. 2Falkland Islands Government,
Environmental Planning Department, P.O. Box 611, Stanley, Falkland Islands.
1
Abstract The Falkland Islands (FI), as with many island ecosystems, is vulnerable to invasive species, which can cause
wide ranging social and environmental consequences. Control of invasive species is widely recognised as a priority, but
there have never been attempts to use classical biological control (CBC) for this purpose in FI. The European earwig
was recently introduced to the FI and has since become abundant in the Stanley area and some other settlements on the
islands. Earwigs now cause considerable damage to garden crops and also pose a number of health hazards. There are
also concerns that earwigs have started to spread into grasslands and irreversibly alter this important native ecosystem.
After extensive stakeholder consultations it was decided to use the invasive earwigs as a case study for the introduction
of CBC to the FI. Based on previous work on earwig control, supplemented by additional host range testing, two tachinid
flies, Triarthria setipennis and Ocytata pallipes, were selected as the most suitable control agents for the Falkland Islands.
Extensive awareness raising activities, focusing on the threat of invasive species, benefits and risks of CBC, secured the
support of the wider public to go ahead with the release of both control agents during 2015 and 2016. Major challenges
encountered during the release process were the need to install makeshift quarantine facilities and the switchover of the
life-cycle of both control agents to southern hemisphere seasons.
Keywords: awareness raising, invasive species, Ocytata pallipes, tachinid flies, Triarthria setipennis, UK Overseas
Territories
INTRODUCTION
The European earwig, (Forficula auricularia, Order
Dermaptera) is widely regarded as a beneficial predator of
insect pests in fruit orchards within its native range of Europe
and West Asia (Nicholas, et al., 2005; Dib, et al., 2010),
however outside this range there are reports that this species
can cause significant agricultural problems (Kuhlmann, et
al., 2001). In 1997/1998 the European earwig was reliably
recorded for the first time in the Falkland Islands (FI),
an archipelago in the South Atlantic Ocean, around 500
km off the southern Patagonian coast of South America.
Since then the earwig has become a significant pest on
the islands causing damage to garden and greenhouse
plants and leading to a halt in the production of a number
of commercial crops (Maczey, et al., 2012). The earwigs
are also posing a number of health hazards, particularly
in autumn (March/April) when they invade buildings in
large numbers. They have been found in asthma inhalers
and beneath the seals of oxygen masks causing the local
hospital to introduce additional safety procedures checking
equipment for the presence of earwigs prior to use. Many
households currently spend substantial amounts of money
to control earwigs primarily by having the foundations of
their houses sprayed with pesticides once or twice a year.
Since its introduction the earwig has become common in
the Stanley area and a number of settlements in the wider
countryside. There is also concern they may spread into
native grasslands, with a risk of irreversibly altering the
indigenous ecosystem posing a particular threat to a high
number of endemic arthropods (Maczey, et al., 2012).
Classical biological control (CBC) has the potential
to offer effective, economic and sustainable control of
this invasive species. This method involves the deliberate
release of specialist natural enemies – mainly insects and
fungi – from the invasive’s native range. The aim is to
reduce the abundance of problem species in its introduced
range below an ecological or economic threshold. The
European earwig is a promising target species for CBC
on the FI, particularly as chemical sprays are ineffective,
and because of its great mobility (Santini & Caroli, 1992).
Off the shelf solutions using parasitoid tachinid flies from
Europe, including Great Britain, are feasible. One such
species Triarthria setipennis has established successfully
in British Columbia and Newfoundland where studies have
indicated a considerable reduction in earwig numbers, most
probably due to high levels of parasitism in the mid-1970s
(Morris, 1984). However, since 1978, no further evaluation
of parasitoid impact has been undertaken. A second species
of parasitoid, Ocytata pallipes, was introduced into Canada
to control the European earwig during the 1990s but no
monitoring took place and establishment is unknown
(Kuhlman, et al., 2001). Ocytata pallipes and T. setipennis
have also been introduced into the USA as early as the
1920s (Oregon) and also into New Zealand (Kuhlman, et
al., 2001). Again, little is known about the success of these
releases.
During a workshop in Stanley in March 2012 there
was consensus on the feasibility of biological control of
invasive non-native species on the South Atlantic UK
Overseas Territories and that the European earwig would
be a target well suited for CBC in the FI. Experts working
on invasive species on the FI and also members of the
general public saw an urgent need for sustainable control
of this species. Equally, the Government of South Georgia
saw this as an opportunity to reduce the risk of future
introductions of earwigs to South Georgia. The Falkland
Island Government (FIG) therefore decided to commission
a host range testing programme to assess the safety and
suitability of two parasitoid flies, believed to be host
specific to the European earwig, for introduction into the
FI (Maczey, et al., 2016).
No native earwig species inhabit the FI, therefore host
range tests were conducted on insect species (crickets and
cockroaches) representing insect orders which are closely
related to earwigs. The Falklands have one native species
of cricket, the camel cricket (Parudenus falklandicus). The
results showed that there was no indication that either of
the two assessed fly species (O. pallipes and T. setipennis)
can develop or otherwise impact on the viability of any
of the test species, even when artificially forced to ingest
parasitoid eggs or inoculated with fly larvae, which would
rarely happen under natural conditions (Maczey, et al.,
2016). The tests confirmed our opinion that there would be
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 389–393. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
389
Island invasives: scaling up to meet the challenge. Ch 2D Other taxa: Invertebrates
no risks to non-target species if one or both of these highly
earwig-specific tachinid fly species were released on the FI
(Maczey, et al., 2016).
Based on the results from the host range testing, the
Environmental Committee and Executive Council of the
FIG decided to go ahead with the release of both agents,
provided there was sufficient support from the wider
public. Up until this point stakeholder acceptance for the
introduction of a new species into the FI had not been
assessed and the Environmental Committee decided to
conduct a range of awareness raising activities to encourage
residents to voice their concerns and engage in open
discussion on the safety and scope of CBC of earwigs. This
paper covers both the outcomes of the awareness raising
activities and the results of the subsequent release of the
control agents conducted between 2015 and 2017.
section, was made available on the FIG website. In addition,
a two-page flyer providing information on our work was
distributed throughout Stanley prior to any public events.
Advertising opportunities to get more detailed information
and voice any concerns
Website and flyer announcements were made of
the dates for presentations and opportunities for open
discussions. The documents also gave contact e-mails to
arrange meetings or discussions outside these dates or to
voice any concerns via e-mail. Times and locations for
all events were also broadcasted by radio and announced
in the local newspaper. An invitation to add to the FAQ
was also given on the website. Presentations given – one
broadcasted by local TV – also included invitations to
forward any questions or concerns to the project team.
To present CABI’s work on earwig control and engage
with the public
MATERIALS AND METHODS
Stakeholder consultation
We wanted to engage with the residents of the FI to
understand whether biological control, in general, and
the release of two parasitoid fly species, in particular,
was of any concern or would be largely welcomed by the
general public and/or experts and scientists working in
conservation or agriculture on the FI.
At the core of all consultations with stakeholders we
communicated four major premises:
● The release of the control agents is safe and does not
pose any risks for native species, human health or
food production, in contrast to the current use of large
quantities of a highly toxic pesticide (Demand® CS).
● Costs for release would largely be covered through
secured funding from Defra (Darwin Initiative).
● Although we saw no major hurdles for a successful
establishment of both fly species, establishment can
never be guaranteed, and this could be a reason for
failure.
● Equally, if successful establishment has taken place,
the amount of control exerted by the released agents
is difficult to predict. Although in the absence of
hyperparasitoids (in this case parasitic wasps known to
develop inside the pupal stage of the tachinid flies) the
likelihood for a good control is high, this is something
which cannot be predicted with absolute certainty.
Stakeholder consultation regarding the biological
control of earwigs focused on three main steps:
Providing initial background information
Information explaining the principles of biological
control and the safety testing of the proposed control
agents, including a ‘frequently answered questions’ (FAQ)
Aside from the widely advertised events, discussions
with residents took place on many other occasions. These
included meetings with pest controllers, members of the
legislative assembly (MLAs), scientists from government
departments and NGOs, teachers, commercial growers and
farmers. Discussions continued after FIG was confident
enough that it would have the backing of the public for
the release of the control agents throughout the length of
the project and also included direct demonstrations of the
activities at the release facilities.
Release programme
In the native range, rates of parasitism by T. setipennis
and O. pallipes vary considerably between metapopulations of earwigs, and large numbers of the host need
to be collected to obtain sufficient parasitoids for release
and establishment. There are no estimates of how many
individuals need to be released to achieve the formation
of a parasitoid population in a new environment, but as a
general rule the more individuals are released the better the
chances are for establishment.
Trapping of earwigs took place in orchards in England
during 2015 and 2016. Sites selected for collecting focused
on locations combining ease of access with high numbers
of specimens, both of earwigs and parasitoids likely to be
obtained. Trapping involved installing flowerpots, 10 x 10
x 17 cm, filled with egg cartons, into trees 1 to 2 m above
ground. Distribution of traps and the collecting regime
are given in Table 1. Earwigs were collected at roughly
monthly intervals three times per year.
Collected earwigs were kept in 40l plastic containers,
housing no more than an estimated 2,500 earwigs per
container. Egg cartons were used to provide hiding places
and lids were fitted with netted openings to give sufficient
aeration. The edges of containers were covered with
Table 1 Earwig/parasitoid collecting regime.
Site
Silwood Park,
Berkshire
South Darenth,
Kent
East Malling
Research, Kent
(EMR)
Target Farm,
Marden, Kent
390
No. of traps
2015
43
No. of traps
2016
31
Setup date
2015
29/6/15
Setup date
2016
27/5/16
Collecting period
2015
29/7/15 – 29/9/15
Collecting
period 2016
26/7/16 – 21/9/16
230
200
03/7/15
25/5/16
28/7/15 – 10/9/15
15/7/16 – 20/9/16
300
325
13/7/15
01/6/16
5/8/15 – 25/9/15
11/7/16 – 16/9/16
-
160
-
23/5/16
-
13/7/16 – 20/9/16
Maczey, et al.: Biological control of earwig, Falkland Is
Fluon® to prevent earwigs escaping. Food consisted of
a mixture of vegetables (lettuce and carrots) and dry dog
food applied three times a week. The earwigs were kept
for a period of six to eight weeks and, afterwards, when
the majority of parasitoid larvae had left their hosts, were
released back at the trapping sites.
Earwig cultures in the lab were checked for parasitoid
pupae three times a week. Pupae were separated to species
and stored in glass tubes sealed with a mesh cover to allow
aeration whilst preventing any potential hyperparasitoids
from escaping (Fig. 1). The tubes were then placed inside
a larger plastic container with a meshed opening to allow
for aeration. Inside this plastic container moistened tissue
facilitated high humidity to prevent desiccation of the
pupae.
In 2015 all pupae were stored at 16˚C until midSeptember, afterwards O. pallipes at 12˚C and T. setipennis
at 8–10˚C until their shipment to Stanley in November
2015. In 2016, pupae of O. pallipes were stored at 10–12˚C.
At this temperature hatching was delayed long enough to
allow two separate shipments to Stanley in August and
September. Pupae of T. setipennis, which hibernates in this
stage, were stored at room temperature (18–20˚C) to mimic
natural conditions. From October onwards, the pupae of
T. setipennis were kept at 10˚C to simulate more natural
overwintering conditions.
On arrival at Stanley, sealed storage boxes containing
vials with pupae were stored in a specifically developed
quarantine shed (details provided at: <http://www.
darwininitiative.org.uk/project/DPLUS033/>) and kept
at 20˚C to trigger hatching. Quarantine facilities were
used as a safety precaution in case hyperparasitoids had
contaminated the fly cultures. Hatched flies were transferred
into rearing tents located in a polytunnel on a daily basis.
O. pallipes were kept there for mating and depositing of
micro-eggs on carrot pieces previously exposed to earwigs,
so that they had obtained the scent of the host species. The
carrot pieces contaminated with fly eggs were then fed
to locally collected earwigs. After inspection confirmed
that most fly eggs had been ingested by earwigs, these
were released at sheltered locations in Stanley with high
densities of earwigs. Adult flies of T. setipennis were kept
only 4–5 days in rearing tents, to allow mating, after which
they were released at sheltered locations with high host
densities.
RESULTS
Stakeholder consultation
Attendance of public events varied from only three
visitors on one occasion to up to 30 visitors during the
demonstration of the release facilities. Feedback after
presentations centred on the safety of CBC. Most frequent
questions were: whether the release control agents could
replace one nuisance species with a second one; or what
the flies would feed on once earwigs went down in
numbers. Our impression was that within the attending
audience it was relatively straightforward to dispel such
concerns by explaining in more detail the host specificity
and dependence of the control agents on host density levels
and that CBC will not lead to complete eradication of the
target species.
People were relieved when seeing the small size of
pinned specimens of the agents passed around, having
expected something much larger. Worries about flies
invading buildings could be dispelled by pointing out that
these species, in contrast to house flies and other species,
will not actively be attracted to houses. Some gardeners
worried that eggs or larvae of the biological control flies
would end up on vegetables; although not being a health
hazard in any way this was seen as unpleasant. The answer
to this was that the flies will deposit eggs and larvae only
on items already smelling strongly of earwigs and in the
case of food items these would be already heavily damaged
crops beyond consideration for human consumption.
Repeatedly, residents raised general concerns about the
continued use of pesticides. Worries about the build-up of
resistance, has already led to changed usage of different
products. There were also concerns that spraying may
temporarily reduce earwig densities to a satisfactory level
which in turn could result in diminished support for CBC.
However, most residents seemed to be aware of natural
fluctuations and also that earwig numbers would be likely
to increase when the use of pesticides is reduced. Several
times the decline in native ‘black beetles’ (a species of
rove beetle, Staphylinidae) was pointed out, which was
also attributed to the use of pesticides. The loss of native
‘black beetles’ was mostly regretted but on occasion the
intrusion of insects of any kind into buildings was seen as
undesirable. On occasion it was suspected that the decline
of native species was caused by the earwigs themselves
and related to a scarceness of such species in areas with
high earwig densities.
Fig. 1 Pupae of T. setipennis inside their storage containers.
Frequently, people had questions about possible
obstacles to the establishment of the control agents.
Comments were made on the possible impact the current
use of pesticides may have on the establishment and
efficacy of the control agents. Pesticides are mostly applied
in autumn when the T. setipennis will only be present as
dormant pupae. However, spraying may still have some
impact on the O. pallipes, which overwinters as larvae
inside living earwigs. Given the climatic conditions on
the FI the majority of earwigs will still overwinter outside
and therefore escape pesticides. We expect that as the
flies begin to establish and gradually start to control the
population of earwigs in Stanley the need for spraying
391
Island invasives: scaling up to meet the challenge. Ch 2D Other taxa: Invertebrates
will reduce so impacting less on both earwigs and flies.
There was also concern about the availability of flowers
providing pollen and nectar for adult flies, something we
believe is not a problem during the time period when adult
flies occur during late spring and summer. Generally, the
audience was also keen to reconstruct the history of earwig
introduction with various speculations on time and entry
points being discussed. There was general agreement that
the biological control will support a reduction in demand
for chemical treatment both reducing costs and risks for
human health and the environment.
Release programme
During 2015, an estimated 50,000 earwigs were
collected in the UK. In 2016 numbers dropped to 18,500
earwigs despite an increase of traps from 573 to 716. Earwig
densities peaked in mid-August with the majority collected
up to this time still being larvae. In each year, numbers
dropped considerably until the end of the collecting period
at the end of September.
A total of 147 pupae of T. setipennis and 237 of O.
pallipes were obtained from the earwigs up to 28 October
2015. Discounting prematurely hatched flies, altogether 145
pupae of T. setipennis and 212 of O. pallipes were shipped
to Stanley for a first release trial on the Falkland Islands
in November 2015. In 2016, 358 pupae of T. setipennis
and 284 of O. pallipes were collected until 21 December.
Discounting prematurely hatched flies, a total of 256 pupae
of T. setipennis were shipped to Stanley in January 2017,
and 225 of O. pallipes in August and September 2016. A
breakdown of collected parasitoids per site and estimated
parasitism rates is given in Table 2.
In November 2015, hatching rates of O. pallipes at
quarantine facilities in the FI were poor, with all flies dying
shortly after emergence. The most likely cause for this was
the prolonged storage of fly pupae under cold conditions
prior to the release, which aimed to synchronise hatching
with the onset of summer in the southern hemisphere.
At the same time T. setipennis did not hatch at all and
emergence only started in January/February 2016. Only a
few flies hatched over several weeks, which were kept in
the mating tents (Fig. 2) and, after six days, altogether only
15 flies were released into an open polytunnel containing
high densities of earwigs.
Shortened storage periods for O. pallipes and prolonged
hibernation of T. setipennis allowed a significantly
improved hatching rate in 2016. More than 200 O. pallipes
hatched in August and September 2016. They mated and
subsequently deposited a large number of micro-eggs
on carrot pieces which had previously been exposed to
earwigs. 1,800 earwigs collected locally were then fed with
pieces of carrots contaminated with fly eggs and released
in Stanley in October. From 256 pupae of T. setipennis
transported to Stanley in January 2017, 185 flies hatched
during February. Some flies died within a short period
Fig. 2 Dave Moore demonstrating the fly rearing tents
during open day at Government House gardens, Stanley
in Nov. 2015 (photo: Sharon Jaffray, Penguin News).
after hatching, but a large proportion were released into
sheltered places in Stanley.
DISCUSSION
Stakeholder consultation
This was the first introduction of a non-native species
for the control of an invasive species on the FI, and a
certain level of concern from expert stakeholders and
the general public was anticipated. Therefore, we tried to
encourage residents to voice their concerns and engage in
open discussion on the safety and scope of CBC. At the
core of all consultations were these premises:
The release of the control agents is safe and does not
pose any risks for native species, human health or food
production
Both successful establishment of CBC agents and the
amount of control they can exert can never be guaranteed
and these can be a reason for failure.
The general feedback most people gave was that of
cautious optimism and being in favour for biological control
provided it is safe. It was important for most people to have
the assurance that biological control does not lead to the
introduction of a species which could become problematic.
We believed that through in-depth discussions worries
and concerns could largely be dispelled. People became
willing to trial a release hoping that it would provide the
anticipated long-term solution to the earwig problem,
whilst being fully aware that there remains a certain risk
of failure. However, this was only partly driven by direct
support of CBC versus an equal measure of concern about
risks and side-effects associated with the current use of
toxic pesticides.
Compared to the amount of advertising preceding public
events, the overall turnout was ~1% of the population of
Table 2 Earwigs, parasitoids and % parasitism recorded in 2015 and 2016.
Site/orchard
Darenth
Darenth
Silwood
Silwood
EMR total
EMR total
Target farm
392
Year
2015
2016
2015
2016
2015
2016
2016
Earwigs
collected
3,000
2,800
1,000
950
46,000
10,300
4,250
T. setipennis
16
49
6
52
125
149
108
O. pallipes
3
6
0
8
234
234
36
% parasitism
T. setipennis
0.5
1.8
0.6
5.5
0.3
1.4
2.5
% parasitism
O. pallipes
0.1
0.2
0.0
0.8
0.5
2.3
0.8
Maczey, et al.: Biological control of earwig, Falkland Is
the Falklands and thus relatively low (although one might
consider drawing in 1,000 attendees in four events in a large
town of 100,000 a very good turnout). Attendance during
the first open day at the release facilities was (~30 visitors)
comparably high and attracted the attention of local radio
and television. However, this dropped significantly in
the second release year, going down to just a handful of
visitors. The same was true for other types of engagement
towards the end of the project. Once initial concerns were
dispelled there was increasingly less new information
between individual events, both from the side of release
activities and any residual concerns to be shared. This may
have resulted in a declining interest or possible increasing
acceptance by the public over time compared to the start
of the project.
Release programme
Despite intensified efforts, earwig trapping in England
during 2016 yielded less than half the numbers of earwigs
obtained in the previous year, which was mainly due to
a drastic population crash in a single cherry orchard at
East Malling Research. In addition, earwig densities at
Target farm varied strongly throughout the year with
few earwigs being collected in September. Low earwig
numbers in 2016 were offset by a recovery of parasitoid
populations, which had been very low in 2015. In both
years the quantity of T. setipennis pupae collected was
substantially lower compared to 1,000+ pupae collected
from 20,000 earwigs in 2013 when the host range testing
took place (Maczey, et al., 2016). It remains unclear
whether T. setipennis suffered a population crash in 2015
or if the collecting sites chosen in 2015/2016 were more
generally characterised by low rates of parasitism. Studies
in continental Europe recorded, on average, higher rates
of parasitism for this species (Kuhlmann, 1995). In both
years, although a few individuals emerged very early in
the season, most T. setipennis pupae were found from the
beginning of September onwards. This coincides with field
observations of some pupae very early in the season in
England indicating a more pronounced second generation
compared to its phenology on the continent, where
occurrence of pupae peaks in August (Kuhlmann, 1991).
Collecting earlier would not have yielded more pupae for
release though, as these mostly emerged early without a
hibernation period, far too early for a release in the FI.
The low number of collected parasitoids and a low
hatching rate in Stanley in 2015 was not sufficient to
enable establishment of either of the two species. One
major problem was switching the lifecycle from a northern
hemisphere rhythm to the seasons in the FI. The lack of
synchronisation of life cycles between the northern and
southern hemisphere is a well-documented problem in
biological control (Waterhouse & Sands, 2001; De ClerckFloate et al., 2008). Ocytata pallipes normally remains
in the pupal stage only for a short period and initially we
tried to delay hatching until the Falkland summer through
storage at lower temperatures, hoping to slow down
development. However, the species does not tolerate being
stored for long periods at low temperatures resulting in
poor survival rates. In 2016, this was addressed by shipping
pupae of O. pallipes to Stanley several times between
August and October. Release in Stanley during late winter
relied on creating a suitable local environment to allow it to
parasitise earwigs soon after arrival. Therefore, flies were
kept in an artificially heated polytunnel warm enough to
allow both earwigs and flies to be active during the winter
months.
The first release trial for T. setipennis also failed but
for a different reason. November was too early to break
the dormancy of this species, which hibernates in the
pupal stage, and early exposure to elevated temperatures
(20°C) only led to unsynchronised emergence in January/
February. For the second release, pupae were kept at
low temperatures until mid-January. This resulted in a
much better synchronised hatching whilst still allowing a
sufficiently long period during the summer in the FI for the
completion of a full life-cycle.
The adapted methodology led to much improved results
and both fly species were successfully released, albeit with
lower numbers than initially hoped for. For O. pallipes, this
was mitigated in 2016 by keeping hatched flies initially in
cages up to the point of eggs being deposited and releasing
larger numbers of earwigs fed with contaminated pieces
of carrots. The ecology of T. setipennis does not allow a
similar approach, but for this species hatching rates had
strongly improved compared to the previous year and the
chances for mating were increased by keeping this species
caged for six days before the release.
At this stage of the release programme we do not know
whether either or both fly species have established. If
establishment has been successful, it is still far too early
to observe an impact on earwig numbers and this will only
become apparent during future years.
ACKNOWLEDGEMENTS
We would like to thank Jeremy Poncet for his immense
contributions to the project. A big thank you goes also to
the residents of FI for their support. Funding for this project
was provided by FIG and the Darwin Initiative (Darwin+).
REFERENCES
De Clerck-Floate, R., Hinz, H.L., Heard, T., Julien, M. Wardill, T. and
Cook, C. (2008). ‘Preliminary Results of a Survey on the Role of
Arthropod Rearing in Classical Weed Biological Control’. In: M.H.
Julien, R. Sforza, M.C. Bon, H.C. Evans, P.E. Hatcher, H.L. Hinz and
B.G. Rector (eds.) Proceedings of the XII International Symposium on
Biological Control of Weeds, pp. 528–534. Wallingford, UK: CABI
Publishing.
Dib, H., Simon, S., Sauphanor, B. and Capowiez, Y. (2010). ‘The role of
natural enemies on the population dynamics of the rosy apple aphid,
Dysaphis plantaginea Passerini (Hemiptera: Aphididae) in organic
apple orchards in south-eastern France’. Biological Control 55: 97–109.
Kuhlmann, U. (1991). ‘Zur Ökologie und Biologie zweier Raupenfliegen
(Diptera: Tachinidae) als Parasitoide des Gemeinen Ohrwurms
(Forficula auricularia, Dermaptera)’. Diploma thesis. Kiel, Germany:
Christian Albrechts Universität.
Kuhlmann, U. (1995). ‘Biology of Triarthria setipennis (Fallén) (Diptera:
Tachinidae), a native parasitoid of the European earwig, Forficula
auricularia L. (Dermaptera: Forficulidae), in Europe’. Canadian
Entomologist 127: 507–517.
Kuhlmann, U., Sarazin, M.J., O'Hara, J.E., Mason, P.G. and Huber,
J.T. (2001). ‘Forficula auricularia L., European Earwig (Dermaptera:
Forficulidae)’. In: P.G. Mason and J.T. Huber (eds.) Biological Control
Programmes in Canada, 1981–2000, pp. 127–131. Wallingford, UK:
CABI Publishing.
Maczey, N., Tanner, R., Cheesman, O. and Shaw, R. (2012). Understanding
and Addressing the Impact of Invasive Non-native Species in the UK
Overseas Territories in the South Atlantic: A Review of the Potential for
Biocontrol. Project report, DEFRA CR 0492. <http://randd.defra.gov.
uk/Default.aspx?Menu=Menu&Module=More&Location=None&Proj
ectID=17564&FromSearch=Y&Publisher=1&SearchText=wc1001&S
ortString=ProjectCode&SortOrder=Asc&Paging=10#Description>
Maczey, N., Edgington, S., Moore, D. and Haye, T. (2016). ‘Biology
and host range testing of Triarthria setipennis and Ocytata pallipes
(Diptera: Tachinidae) for the control of the European earwig (Forficula
auricularia)’. Biocontrol Science and Technology 26: 447–461.
Morris, R.F. (1984). ‘Forficula auricularia, European Earwig
(Dermaptera: Forficulidae)’. In: J.S. Kelleher and M.A. Hulme (eds.)
Biological Control Programmes Against Insects and Weeds in Canada
1969–1980, pp. 39–40. Slough, UK: Commonwealth Agriculture
Bureaux.
Nicholas, A.H., Spooner-Hart, R.N. and Vickers, R.A. (2005). ‘Abundance
and natural control of the woolly aphid Eriosoma lanigerum in an
Australian apple orchard IPM program’. Biological Control 50: 271–
291.
Santini, L. and Caroli, L. (1992). ‘Damage to fruit crops by European
earwig (Forficula auricularia L.)’. Informatore-Fitopatologico 42:
35–38.
Waterhouse, D.F. and Sands, D.P.A. (2001). Classical Biological Control
of Arthropods in Australia. Monograph No. 77. Canberra, Australia:
ACIAR.
393
C.E. Parent, P. Fisher, W. Jolley, A. Alifano and K.J. Campbell
Parent, C.E.; P. Fisher, W. Jolley, A. Alifano and K.J. Campbell. Assessment of snail
exposure to the anticoagulant rodenticide brodifacoum in the Galapagos Islands
Assessment of snail exposure to the anticoagulant rodenticide
brodifacoum in the Galapagos Islands
C.E. Parent1, P. Fisher2, W. Jolley3, A. Alifano3 and K.J. Campbell4,5
Department of Biological Sciences, University of Idaho, Moscow, ID 83844, USA. <ceparent@uidaho.edu>. 2Landcare
Research, Gerald Street, PO Box 69040, Lincoln 7640, New Zealand. 3Island Conservation, 2161 Delaware Ave Suite A,
Santa Cruz, CA 95060, USA. 4Island Conservation, Charles Darwin Ave., Puerto Ayora, Galapagos Islands, Ecuador.
5
School of Geography, Planning and Environmental Management, The University of Queensland, St Lucia 4072,
Australia.
1
Abstract Eradication of invasive rodents has become a powerful tool to protect native island biota. Use of brodifacoum,
an anticoagulant rodenticide, has contributed to hundreds of successful invasive rodent eradication efforts on islands.
Application of bait containing brodifacoum for this purpose requires appropriate consideration of adverse effects on
non-target wildlife. Thus, a priori identification of non-target risks and, where needed, approaches to mitigate these
to acceptable levels, is now an essential component of eradication planning and implementation. As part of the plan
for eradicating invasive rats and mice from Floreana Island in the Galapagos, we experimentally tested the effect of
brodifacoum on the Galapagos endemic land snail species Naesiotus unifasciatus. Importantly, the trials were designed
to evaluate effects of particular components of the bait pellets, namely the active brodifacoum, the pyranine biomarker,
and a blue dye. We found no evidence for increased snail mortality following exposure to any of these bait components.
We review results of past toxicity studies on terrestrial molluscs and find that, as for our own study, there is likely to be
little impact of anticoagulant rodenticide on terrestrial mollusc survival as the result of application of brodifacoum bait.
However, given the limited taxonomic representation in the toxicity tests performed on terrestrial molluscs so far, we
recommend the continued used of captive toxicity trials to assess potential effect of any rodenticide applications on native
malacological fauna on a case-by-case basis where large-scale eradication programmes are planned and undertaken.
Keywords: anticoagulant, brodifacoum, Bulimulidae, islands, restoration, rodent eradication
INTRODUCTION
Invasive mammal eradications are powerful
conservation tools to protect biodiversity and prevent
extinctions on islands (Lorvelec & Pascal, 2005;
Bellingham, et al., 2010; Nogales, et al., 2013). Three
rat species (Rattus rattus, R. norvegicus, R. exulans)
and house mice (Mus musculus) are the most common
rodents introduced to islands worldwide (Atkinson, 1985).
These species are responsible for population declines
and extinctions of insular flora and fauna, and they are
known to interrupt ecosystem processes with negative
cascading effects (Fukami, et al., 2006; Steadman, 2006;
Towns, et al., 2006; Jones, et al., 2008; Kurle, et al., 2008;
Varnham, 2010; Dunlevy, et al., 2011; St Clair, 2011).
To recover endangered species and restore ecosystem
processes, invasive rodents on islands are increasingly
targeted for eradication, with at least 637 successful rodent
eradications to date (based on DIISE island data ranked as
good or satisfactory; DIISE, 2015). Ninety-seven percent
of successful rodent eradications have involved the use of
rodenticide, with brodifacoum having been used in 76%
of them.
The common mode of toxicity of anticoagulant
rodenticides in mammals and birds is to inhibit Vitamin K
metabolism in liver, which in turns prevents the formation
of chemical factors essential to blood coagulation (e.g.,
Rattner, et al., 2014). In mammals and birds a lethal
exposure will cause these clotting factors to deplete to a
level so that blood can no longer coagulate, resulting in
death through internal haemorrhage (MacNicoll, 1993).
‘First-generation’ anticoagulants, such as warfarin, are
most effective against rodents in multiple feeds but their
intensive use as rodenticides resulted in the development
of heritable resistance in some rodent populations (Rattner,
et al., 2014). This prompted the development of the
more potent ‘second generation’ anticoagulants, such as
brodifacoum, which are effective against target rodents in
a single feed (Rattner, et al., 2014). Sublethal or chronic
effects of anticoagulants are not well described in wildlife
(Rattner, et al., 2014), but sublethal exposure may result in
the retention of residual anticoagulant concentrations in liver
tissue. In this regard the ‘second generation’ anticoagulant
rodenticides are more persistent in animal tissues,
especially liver, than ‘first-generation’ anticoagulants
(Fisher, et al., 2003). The second generation anticoagulant
brodifacoum has been the most commonly used rodenticide
for eradicating invasive rodents from islands, with a high
success rate (Howald, et al., 2007; Parkes, et al., 2011;
DIISE, 2015). Brodifacoum, incorporated at 20–50 ppm
(0.002–0.005%) into cereal or wax baits, is applied to
every rodent territory via bait stations, or broadcasted
by hand or from a modified agricultural spreader bucket
suspended from a helicopter. Large-scale broadcast of
bait has facilitated increasingly large and complex island
restoration projects involving the eradication of invasive
rodents (e.g., Towns & Broome, 2003), but it also raises
concerns about environmental contamination and adverse
effects on non-target wildlife (Pain, et al., 2000; Eason, et
al., 2002). Thus, a priori identification of non-target risks
and the potential mitigation of these to acceptable levels
is now an essential step to inform feasibility of large-scale
eradication projects.
Invasive rats are known to prey upon terrestrial
invertebrates (e.g., St Clair, 2011). On Galapagos, endemic
land snails are particularly vulnerable (Clark, 1981), and
recent field collections of land snail shells suggest that
rats are particularly voracious snail predators on Floreana
(Parent, unpublished data). Although the eradication of
invasive rodents would likely benefit terrestrial molluscs,
the potential impact of bait on non-target species should
be evaluated. Indeed, a range of terrestrial invertebrate
species, including snails and slugs, have been found to
feed on cereal-based baits used for rodent control (e.g.,
Spurr & Drew, 1999; Johnston, et al., 2005). Reports that
bait containing brodifacoum caused mortality in captive
introduced and endemic snails (Achatina fulica and
Pachnodus silhouettanus) from the Seychelles Islands
that fed on the baits, and suspected field mortality of
Pachystyla bicolor snails following operational baiting
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
394
up to meet the challenge, pp. 394–399. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Parent, et al.: Brodifacoum and snails, Galapagos Is
(anecdotally reported in Gerlach & Florens (2000a; 2000b)
and Gerlach (2005)) raised concerns for other native and
endemic snail species on islands where rodent eradication
using anticoagulants was proposed. Limited information
suggests that invertebrates are generally less susceptible to
brodifacoum toxicity than mammals and birds (Booth, et
al., 2001; Eason & Spurr, 1995), but current knowledge of
snail physiology is insufficient to predict with confidence
its effect on snails. To assess the feasibility of using
brodifacoum to eradicate rats and mice from Floreana
Island in the Galapagos (Island Conservation, 2013),
a need to investigate risk to endemic land snails was
therefore identified.
Land snails are known for their remarkable diversity
in island systems (Cameron, et al., 2013). On Galapagos
Islands, the land snail fauna comprises 103 endemic
species distributed in 13 genera. Approximately 80 species
and subspecies belong to the genus Naesiotus (Family
Bulimulidae) and form the most species-rich adaptive
radiation of these islands (Parent, et al., 2008). Recent
field and genetic work suggests that most (if not all)
Galapagos bulimulid species are single-island endemics
(Parent & Crespi, 2006; Parent, unpublished data). Twenty
species (and eight subspecies) of endemic land snails are
known from Floreana Island. Eight of these species are
critically endangered, and three are endangered (IUCN,
2015), whereas others remain to be evaluated. Thus, given
the conservation status of Floreana endemic snails, we
identified the need to evaluate whether exposure of these
endemic snails to the brodifacoum bait type proposed for
rodent eradication was likely to cause mortality.
Four previous experimental studies, together assessing
twelve species of terrestrial molluscs, have failed to find
significant effect of brodifacoum exposure on individual
short-term mortality. The only exception to this trend is
the study reported by Gerlach & Florens (2000a; 2000b)
mentioned above. Therefore, a precautionary approach
demands that the effects of exposure to brodifacoum bait
should be tested on island endemic snails prior to largescale rodent eradication measures on islands. Importantly,
the rodenticide baits are composed of more than 99%
inert ingredients, most of which is compacted cereal
grains but may include other inert ingredients such as
dye or biomarkers. Past studies failed to explicitly test
the effect of inert ingredients on land snails. Thus, the
main objectives of the study are to: (1) test various bait
formulations to identify which component(s) of the baits
are responsible for any mortality that might be observed
in land snails, and (2) review and synthesize the literature
on experimental toxicity tests of bait-based rodenticides on
terrestrial molluscs.
MATERIALS AND METHODS
Site description
Floreana Island is part of the Galapagos archipelago,
which straddles the equator approximately 1,000 km off
the western coast of Ecuador. The islands are oceanic and
have never been connected to any continent. Floreana is
volcanic in origin, and at 17,253 ha is the sixth largest
island in the archipelago. The maximum elevation of the
island is 640 m, and its generally conical shape results in
two distinct habitat types: dry lowlands and lush central
highlands, which meet and overlap to some extent into
what is referred to as a transition zone (McMullen, 1999).
Over 98% of the island land area is Galapagos National
Park, with the remaining 2% divided between a small
town in the lowlands and agricultural and pastoral areas
in the highlands (DPNG, 2014). The island is home to an
estimated 140 residents as of 2014.
Snail population
Snails were collected at Cerro Pajas on Floreana
(Latitude: 01.2968ᵒS, Longitude: 90.4559ᵒ W) in
November 2012. The Cerro Pajas site was selected based
on the relatively high density of snails found there (Parent,
unpublished data). We collected snails opportunistically
from leaf litter on the ground and from low (< 0.5 m)
vegetation. We chose to collect snails near the ground
because those snails would be more likely to encounter
bait pellets on or near the ground. There are at least three
Galapagos endemic species of land snails occurring at
that particular site (Naesiotus nux, N. unifasciatus, and
Succinea brevior; endangered, critically endangered,
and unknown status, respectively), and these species are
expected to be either detritivores or to consume algae
and lichens scraped from the substrate. For the present
study, we used adult individuals of N. unifasciatus since
the population density of this species was the highest and
we felt confident that our sampling would not impact the
survival of the population at that particular location (100
adult individuals were collected, less than 5% of the
individuals encountered over a period of approximately
one hour). We did not collect specimens from any other
snail species because their population density was such
that we would have had to collect more than 5% of the
adult individuals encountered for our experiments. It is
important to note that the information gathered from our
toxicity experiments will by far outweigh the potential
detrimental effects of our population sampling of N.
unifasciatus. Bulimulid snails are hermaphrodites and
therefore sexing them was not applicable.
Experimental design
We housed snails in small cylindrical plastic containers
(11 cm high, 17 cm and 14 cm in diameter at the top
and base of the container, respectively) replacing lids
with tightly covering mesh secured with rubber bands to
prevent snail escapes. Two sheets of task wipe (Kimwipe
®) paper were placed on the bottom of each container and
kept moist for the duration of the experiment. A small
amount of litter from which the snails were collected was
sifted and visually inspected to remove any other small
invertebrates. Approximately 10 grams of sifted litter was
added to each container as a source of natural food and
shelter. Each container held five snails at the beginning of
the experiment.
We used three types of pelleted bait as experimental
treatments: (1) non-toxic baits containing a blue dye (well
less than 1% of pellet content) that is a standard proprietary
component of the bait formulation, (2) non-toxic baits
containing pyranine (a fluorescent marker dye allowing
easy detection of metabolized bait in snails’ bodies, faeces
and slime trails when exposed to ultraviolet (UV) light,
also representing well less than 1% of pellet content); and
(3) bait containing blue dye and 50 ppm brodifacoum. For
treatment groups, one moistened bait pellet was placed
in each container at the start of the trial. Control group
containers had no pellets, but were otherwise the same
as treatment containers. We prepared five containers per
treatment and for the control group (total n = 100 snails).
All containers were kept on Floreana, at sea level in the
shade at ambient temperature (25-28°C). In parallel, we
kept five pellets in a container under the same conditions
but without snails to evaluate the effect of the containers on
the pellets themselves. All containers were opened twice
daily to increase circulation of fresh air and to remoisten the
tissue paper and bait pellet by spraying water, as necessary.
The experiment was conducted in two parts; the first
over 10 days during which all containers with snails
were monitored daily for mortality. A 10-day period was
395
Island invasives: scaling up to meet the challenge. Ch 2D Other taxa: Invertebrates
selected as slightly longer than the four to eight days that
bait pellets are likely to be available to snail populations in
the Floreana highlands (Island Conservation, unpubl. data)
and substantially longer than the maximum of 72 hours
over which snail mortality occurred in the study reported
by Gerlach & Florens (2000a and 2000b).
Snail activity was noted twice a day by recording
whether the snails were immobile and firmly attached to
substrate (i.e., estivating) or moving in the containers.
Snails found estivating were moved onto the vegetation
and sprayed with water; this reliably caused the snails to
become active once again. Snails found dead were frozen
immediately in individual vials to preserve tissue and be
dissected later if any statistically significant mortality
effects were to be detected in our study. In addition,
containers were visually inspected daily by illumination
with a UV light to detect any fluorescent traces of pyranine
in the slime trails and faeces of the snails in containers of
treatment 2 which would have indicated ingestion of this
bait by the snails. Control containers were inspected in the
same manner. Finally, we also monitored bait consumption
by noting any changes to the surfaces of the bait pellets.
At the end of the 10-day period, living snails from the two
treatment groups using non-toxic bait types were returned
to the location where they were collected.
In the second part of the experiment, the remaining
snails in the control group and the treatment group with
brodifacoum baits were monitored for an additional 11
days (for a total of 21 days, well in excess of the period
over which signs of poisoning and mortality would
have occurred in mammals) and any snails alive were
euthanized by freezing at the end of this second part of the
experiment to use in subsequent residue content analysis if
any significant mortality was observed.
Statistical analyses
We used a logistic regression approach with a Bernoulli
(binomial) distribution to evaluate the effect of each bait
component on the survival of the snails. We used post-hoc
tests to determine whether any of the bait components had
a significant effect on snail survival. We implemented all
statistical analyses in R version 3.1.0 (R Core Team, 2014).
Fig. 1 Survival of snails exposed to different types of
pellets: non-toxic pellets with blue dye, non-toxic pellets
with biomarker, pellets with 50 ppm brodifacoum and
blue dye, and control populations without pellets.
Survival for all treatments was not significantly different
than the survival of the snails in the control populations
(P > 0.05). Error bars represent standard errors.
396
RESULTS
Snails remained active (i.e., not estivating) throughout
the experiment. Individuals were confirmed to consume
baits, either by direct observation (snails on bait pellet)
or evidence of blue dye or pyranine fluorescence in the
snails’ bodies, faeces or slime trails. We did not track each
individual snail’s consumption of bait, but the observed
evidence suggested that most snails consumed or were in
direct physical contact with bait when available.
The survival of snails over the course of the bait
treatments did not differ significantly from the survival
of snails in the control groups without baits (Fig. 1).
Because our treatment groups did not represent all possible
combinations of bait type (with/without brodifacoum,
presence/absence of blue dye, and presence/absence of
pyranine), we could not directly compare the individual
effect of each of these bait components on the snails.
However, we found that when all treatments were analysed
simultaneously, none of the components had a significant
effect on snail survival (Fig. 1). In contrast, in a post-hoc
test for individual effects of each component, we found
that the survival of snails exposed to bait containing
pyranine was greater than the survival of snails exposed
to bait without pyranine (Fig. 2; Welch two-tailed t-test for
unequal sample size, t = 3.056, d.f. = 14, P < 0.01). The
survival of the snails over 21 days in the containers with
bait containing brodifacoum did not significantly differ
from the control (Welch two-sample two-tailed t-test, t =
1.497, d.f. = 5.611, P > 0.05).
DISCUSSION
The goal of our study is to quantify the short-term
impacts of anticoagulant rodenticide bait on Galapagos
endemic land snails. Importantly, any potential short-term
impact the bait might have has to be considered against
the long-term benefits that rodenticide bait application
can bring to terrestrial malacofauna. These potential
benefits include, for example, release from invasive rodent
predation and general habitat improvement.
Our results suggest that none of the baits tested were
toxic to the snails over the 10-day exposure period (i.e.
Fig. 2 Survival of snails as a function of presence/
absence of pellet components: blue dye, biomarker,
and brodifacoum. A logistic regression approach
with a Bernoulli (binomial) distribution including all
samples at once reveals no significant effect for any of
the components (P > 0.05). However, a post-hoc t-test
indicates that snails exposed to pellets with biomarkers
have significantly higher survival than snails exposed
to pellets without biomarker (P < 0.01). Error bars
represent standard errors.
Henderson
Is, South
Pacific
Henderson
Island,
South
Pacific
Henderson
Is, South
Pacific
New
Zealand
Fregate
Island,
Seychelles
Floreana
Island,
Ecuador
Helicinidae
Achatinellidae
Pupilidae
Helicidae
Cerastidae
Achatinidae
Orobophana
solidula
Pacifella sp.,
Tornatellides sp.,
Lamellidae sp.,
Tubuaia hendersoni
Pupisoma orcula
Helix aspersa
Pachnodus
silhouettanus,
Achatina fulica
Naesiotus
unifasciatus
Bulimulidae
Hawaii,
USA
Limacidae
Limax maximus
yes
Information
not available
yes
yes
yes
yes
yes
yes
Hawaii,
USA
Zonitidae
Deroceras laeve
Ciliellopsis oglasae
Oxychilus spp.
Use of
control(s)?
Information
not available
Information
not available
yes
Locality
Montecristo
Island, Italy
Hygromiidae
Montecristo
Island, Italy
Agriolimacidae Hawaii,
USA
Family
Oxychilus oglasicola Zonitidae
Species
Pestoff 20R (0.002%
brodifacoum)
Pestoff 20R (0.002%
brodifacoum)
Talon 20P (0.002%
brodifacoum)
Information 0.01 - 0.2mg
not
brodifacoum
available
25
Bell Laboratories
Brodifacoum 50D
Conservation Blue
FP2015 (0.005%
brodifacoum)
24
1
43
Number of Anticoagulant
individuals rodenticide
tested
12
Brodifacoum 0.005%,
bromadiolone
4
Brodifacoum 0.005%,
bromadiolone
15
HACCO Ramik
® Green (0.005%
diphacinone)
15
HACCO Ramik
® Green (0.005%
diphacinone)
15
HACCO Ramik
® Green (0.005%
diphacinone)
28
Pestoff 20R (0.002%
brodifacoum)
Table 1 Terrestrial mollusc species experimentally tested for the effects of anticoagulant rodenticides.
21
4
14
10
10
10
7
7
Brodifacoum
exposure time
(max no of days)
Information not
available
Information not
available
7
Mortality was
observed but not
statistically tested
Not significant
Not significant
Not significant
Not significant
Not significant
Not significant
Not significant
Not significant
Not significant
Not significant
Effect on organism
survival
Booth et al.,
2003
Gerlach and
Florens, 2000a
& 2000b
This study
Brooke et al.,
2011
Brooke et al.,
2011
Brooke et al.,
2011
Johnston et al.,
2005
Johnston et al.,
2005
Sposimo et al.,
2011
Sposimo et al.,
2011
Johnston et al.,
2005
Reference
Parent, et al.: Brodifacoum and snails, Galapagos Island
397
Island invasives: scaling up to meet the challenge. Ch 2D Other taxa: Invertebrates
survival of snails exposed to any bait treatment was not
significantly different than 1.0). For the purposes of a
conservative risk assessment, our experiment simulated
a ‘worst case’ exposure to snails through use of 50 ppm
brodifacoum in bait, compared to the lower concentrations
proposed for the rodent eradication on Floreana (25 ppm)
and used in previous similar experiments with snails
(e.g., Booth, et al., 2003; Brooke, et al., 2011; Table 1). In
confining snails under conditions favourable for foraging
and in close proximity to bait, we also simulated a worstcase exposure potential, in comparison to the expected
availability of bait to snails following an operational
aerial application. Application rates for rodent eradication
on Floreana Island remain to be determined, nonetheless
relatively few snails are expected to encounter and consume
bait before it is removed by other animals or breaks down
naturally. Additionally, the operation will occur during the
driest time of the year, corresponding to the time when
snails are more likely to be estivating (Parent, unpublished
data).
An apparent absence of toxic effects of brodifacoum
bait on snails was further supported by survival being
significantly higher in a post-hoc test in snails exposed to
non-toxic baits that contained pyranine compared to snails
exposed to baits without the biomarker during the first 10
days of the experiment. It is possible that one or some of
inert components of the bait types used in our experiment
provided a nutritional supplement benefiting the snails.
Any such benefits from a boost in diet would become
more evident over time. However, this pattern of increased
survival did not carry over in snails that were kept for the
full-length (21 days) of the experiment.
Our results add to a growing body of research suggesting
that exposure to rodenticide bait formulations containing
brodifacoum does not cause significant mortality in snails
(Table 1). Reports by Gerlach & Florens (2000a; 2000b)
and Gerlach (2005) appear to be exceptions, but are also
brief and lacking in detail that would allow statistical
evaluation. While the absolute toxicity of brodifacoum
and its mechanism in snails remain to be established, in
the context of potential exposure to rodenticide baits it is
important to also consider the possible effects of other,
nominally inert, ingredients of specific bait formulations
(e.g., binders, preservatives, emulsifiers, pH regulating,
flavouring or colouring agents). In designing our study we
sought to account for some of these factors and recommend
that future studies of the effects of rodenticides on
invertebrates contain a mechanism to ensure any observed
mortality is in fact due to the active anticoagulant agent
and not other bait components or experimental conditions.
We caution that our tests were performed on a single
species of land snail, and are therefore not extensive
enough to confirm that exposure to brodifacoum bait would
not have adverse effects in other terrestrial malacofauna
on Galapagos or elsewhere. However, given that most
Galapagos endemic snail species are of the same genus
as the species tested here, we feel confident that at least
this important group will not be affected by exposure
to brodifacoum bait if it was applied for eradication of
invasive rodents on Floreana and other Galapagos Islands.
Most importantly, these snails are known to be consumed
by introduced rats (Clark, 1981; Parent, unpublished data),
and therefore eradication of invasive rodents is more likely
to result in positive effects on Galapagos endemic snail
populations.
Secondary exposure pathways must be considered
when assessing non-target risk and when developing
measures to prevent non-target mortality (Eason, et
al., 1999). We did not test for residual brodifacoum
398
concentrations in the bodies of exposed snails in our study,
but Booth et al. (2003) measured brodifacoum residues
in the bodies of some snails that had consumed bait. We
expect that any snails that consume bait on Floreana Island
could constitute a secondary exposure pathway for their
predators such as some of the larger land birds. The only
Galapagos birds that have been verified to be preying on
endemic snails are the Galapagos mocking birds which
have been extirpated from Floreana Island. There are
no other known potential secondary exposure pathways
for non-target species on Floreana Island involving the
endemic snails as intermediate.
Evidence to date and our results indicate that rodent
bait containing brodifacoum does not present a high risk
of non-target mortality to terrestrial snails. However, our
study is limited to the detection of mortality (i.e. we did
not monitor for other potentially negative effects) and was
over a short period of time. Given the general trend across
terrestrial molluscs of the effect of brodifacoum on snail
mortality, we recommend a re-evaluation of this effect
for the species included in the study reported by Gerlach
& Florens (2000a; 2000b) that would incorporate a more
complete set of treatments and controls. More specifically,
we recommend more toxicity tests on the invasive giant
African snails (Achatina fulica) given its broad distribution
(tests could be conducted in a range of localities on
continents and islands) and the negligible impact these
tests would have on this highly invasive species (Lowe,
et al., 2004). We conclude that it is prudent to continue to
assess toxicity risk on a species by species basis, where
rodent eradication using brodifacoum or other rodenticides
is planned. Trials to determine whether captive snails
would eat baits and whether exposure to baits results in
measurable mortality are a relatively straightforward and
low-cost means to test theoretical assessments of nontarget risk.
ACKNOWLEDGEMENTS
The authors thank Walter Gilberto Mora Mora, Jimy
Andres Mora Mora, and Francisco Morena Naula for their
help in the field. We are grateful to the Galapagos National
Park Directorate for sampling permission and the Charles
Darwin Foundation for logistical support, especially Luis
Ortiz-Catedral. We also thank Claudio Cruz Bedon and
María de Lourdes Soria Rivera for logistic support while
on Floreana Island. This work was supported by funding
from The David and Lucile Packard Foundation and The
Leona M. and Harry B. Helmsley Charitable Trust. During
this work CE Parent was supported by an NSERC PDF,
and the authorship contribution of P Fisher was supported
by the New Zealand Ministry of Business, Innovation and
Employment (Project C09X0910). Lynn Booth provided
improvements to earlier drafts of this manuscript.
REFERENCES
Atkinson, I.A.E. (1985). ‘The Spread of Commensal Species of Rattus to
Oceanic Islands and their Effects on Island Avifaunas’. In: P.J. Moors
(ed.) Conservation of Island Birds, pp. 35–81. Cambridge: International
Council for Bird Preservation.
Bellingham, P.J., Towns, D.R., Cameron, E.K., Davis, J.J., Wardle, D.A.,
Wilmshurst, J.M. and Mulder, C.P.H. (2010). ‘New Zealand island
restoration: Seabirds, predators, and the importance of history’. New
Zealand Journal of Ecology 34: 115–136.
Booth, L.H., Eason, C.T. and Spurr, E.B. (2001). ‘Literature review of the
acute toxicity and-persistence of brodifacoum to invertebrates’. Science
for Conservation 177: 1–9.
Booth, L. H., Fisher, P., Heppelthwaite, V. and Eason, C. T. (2003). ‘Toxicity
and residues of brodifacoum in snails and earthworms’. Department of
Conservation Science Internal Series 143: 5–14.
Parent, et al.: Brodifacoum and snails, Galapagos Island
Brooke, M. de L., Cuthbert, R.J., Mateo, R. and Taggart, M.A. (2011).
‘An experimental test of the toxicity of cereal pellets containing
brodifacoum to the snails of Henderson Island, South Pacific’. Wildlife
Research 38: 34–38.
Cameron, R.A.D., Kostas, A.T., Parent, C.E., Guilhaumon, F., Alonso,
M.R., Ibánez, M., Frias Martins, A.M., Ladle, R.L. and Whittaker, R.J.
(2013). ‘Snails on oceanic islands: Testing the general dynamic model of
oceanic island biogeography using linear mixed effect models’. Journal
of Biogeography 40(1): 117–130.
Clark, D.A. (1981). ‘Foraging patterns of black rats across a desertmontane forest gradient in the Galapagos Islands’. Biotropica 13:
182–194.
DIISE (2015). The Database of Island Invasive Species Eradications,
developed by Island Conservation, Coastal Conservation Action
Laboratory UCSC, IUCN SSC Invasive Species Specialist Group,
University of Auckland and Landcare Research New Zealand. <http://
diise.islandconservation.org>. Accessed 11 December 2015.
DPNG (2014). ‘Plan de Manejo de las Áreas Protegidas de Galapagos
para el Buen Vivir’. Puerto Ayora, Galapagos: Dirección Parque
Nacional Galapagos, Ministerio de Ambiente.
Dunlevy, P., Ebbert, S.E., Russell, J.C. and Towns, D.R. (2011).
‘Eradication of Invasive Predators on Seabird Islands’. In: Mulder,
C.P.H. Anderson, W.B., Towns, D.R. and Bellingham, P.J. (eds.)
Seabird Islands Ecology, Invasion and Restoration, pp. 283–316. New
York: Oxford University Press.
Eason, C.T., and Spurr, E.B. (1995). ‘A review of the toxicity and impacts
of brodifacoum on non-target wildlife in New Zealand’. New Zealand
Journal of Zoology 22(4): 371–379.
Eason, C.T., Milne, L., Potts, M., Morris, S.G., Wright, R.G. and
Sutherland, O.R.W. (1999). ‘Secondary and tertiary poisoning risks
associated with brodifacoum’. New Zealand Journal of Ecology
23(2): 219–224.
Eason, C.T., Murphy, E.C., Wright, G.R.G. and Spurr, E.B. (2002).
‘Assessment of risks of brodifacoum to non-target birds and mammals
in New Zealand’. Ecotoxicology 11: 35–48.
Fisher, P., O'Connor, C., Wright, G. and Eason, C. (2003). ‘Persistence
of four anticoagulant rodenticides in the liver of laboratory rats’.
Department of Conservation Science Internal Series 139: 1–19.
Fukami, T., Wardle, D.A., Bellingham, P.J., Mulder, C.P.H., Towns,
D.R., Yeates, G.W., Bonner, K.I., Durrett, M.S., Grant-Hoffman, M.N.
and Williamson, W.M. (2006). ‘Above- and below-ground impacts of
introduced predators in seabird-dominated island ecosystems’. Ecology
Letters 9: 1299–1307.
Gerlach, J. and Florens, V. (2000a). ‘Considering molluscs in rodent
eradication projects’. Tentacle – The Newsletter of the IUCN/SSC
Mollusc Specialist Group 9: 7–8.
Gerlach, J. and Florens, V. (2000b). ‘Toxicity of 'specific' rodenticides and
the risk to non-target taxa’. Phelsuma 8: 75–76.
Gerlach, J. (2005). ‘The impact of rodent eradication on the larger
invertebrates of Fregate Island, Seychelles’. Phelsuma 13: 44–54.
Howald, G., Donlan, C.J., Galván, J.P., Russell, J., Parkes, J., Samaniego,
A., Wang, Y., Veitch, D., Genovesi, P., Pascal, M., Saunders, A. and
Tershy, B. (2007). ‘Invasive rodent eradication on islands’. Conservation
Biology 21: 1258–1268.
Island Conservation (2013). ‘Floreana ecological restoration: Rodent and
cat eradication feasibility analysis version 6.0’. Santa Cruz, CA: Island
Conservation.
Lowe, S., Browne, M., Boudjelas, S. and De Poorter, M. (2004). ‘100 of
the world's worst invasive alien species: A selection from the Global
Invasive Species Database’. Auckland: The Invasive Species Specialist
Group (ISSG) a specialist group of the Species Survival Commission
(SSC) of the World Conservation Union (IUCN).
MacNicoll, A.D. (1993). ‘Anticoagulant rodenticides - tolerance and
resistance’. Phytoparasitica 21(3): 185–188.
McMullen, C. K. (1999). Flowering Plants of the Galápagos. Ithaca:
Cornell University Press.
Nogales, M., Vidal, E., Medina, F.M., Bonnaud, E., Tershy, B.R.,
Campbell, K.J. and Zavaleta, E.S. (2013). ‘Feral cats and biodiversity
conservation: The urgent prioritization of island management’.
Bioscience 63: 804–810.
Pain, D.J., Brooke, M. de L., Finnie, J.K. and Jackson, A. (2000). ‘Effects
of brodifacoum on the land crab, Gecarcinus lagostoma on Ascension
Island: acute effects, persistence and distribution in tissues’. The Journal
of Wildlife Management 64: 380–387.
Parent, C.E., Caccone, A. and Petren, K. (2008). ‘Colonization and
diversification of Galapagos terrestrial fauna: a phylogenetic and
biogeographical synthesis’. Philosophical Transactions of the Royal
Society of London B: Biological Sciences 363: 3347–3361.
Parent, C.E. and Crespi, B.J. (2006). ‘Sequential colonization and rapid
diversification of Galapagos endemic land snail genus Bulimulus
(Gastropoda, Stylommatophora)’. Evolution 60: 2311–2328.
Parkes, J., Fisher, P. and Forrester, G. (2011). ‘Diagnosing the cause of
failure to eradicate introduced rodents on islands: Brodifacoum versus
diphacinone and method of bait delivery’. Conservation Evidence 8:
100–106.
R Core Team (2014). R: A language and environment for statistical
computing. Vienna: R Foundation for Statistical Computing.
Rattner, B., Lazarus, R.S., Elliott, J.E., Shore, R.F. and Van Den Brink,
N.W. (2014). ‘Adverse outcome pathway and risks of anticoagulant
rodenticides to predatory wildlife’. Environmental Science &
Technology 48: 8433–8445.
Sposimo, P., Baccetti, N., Pelliccioni, E.R., Guberti, V., Giannini, F.
and Capizzi, D. (2011). ‘Piano per l’eradicazione del ratto nero Rattus
rattus nell’Isola di Montecristo, Archipelago Toscano’. <http://www.
montecristo2010.it/stealthV3_pubblica/08203500TA4264078046.pdf>.
Accessed 7 October 2014.
Spurr, E.B. and Drew, K.W. (1999). ‘Invertebrates feeding on baits used
for vertebrate pest control in New Zealand’. New Zealand Journal of
Ecology 23: 167–173.
St Clair, J.J.H. (2011). ‘The impacts of invasive rodents on island
invertebrates’. Biological Conservation 144: 68–81.
Steadman, D.W. (2006). Extinction and Biogeography of Tropical Pacific
Birds. Chicago: University of Chicago Press.
Towns, D.R. and Broome, K.G. (2003). ‘From small Maria to massive
Campbell: Forty years of rat eradications from New Zealand islands’.
New Zealand Journal of Zoology 30: 377–398.
Towns, D.R., Atkinson, I.A.E. and Daugherty, C.H. (2006). ‘Have
the harmful effects of introduced rats on islands been exaggerated?’
Biological Invasions 8: 863–891.
Varnham, K. (2010). Invasive Rats on Tropical Islands: Their History,
Ecology, Impacts and Eradication. Report No. 41. UK: Royal Society
for the Protection of Birds Research.
IUCN (2015). The IUCN Red List of Threatened Species. Version
2014.2. Gland, Switzerland and Cambridge, UK: IUCN. <http://www.
iucnredlist.org>. Accessed 24 July 2015.
Johnston, J.J., Pitt, W.C., Sugihara, R.T., Eisemann, J.D., Primus,
T.M., Holmes, M.J., Crocker, J. and Hart, A. (2005). ‘Probabilistic
risk assessment for snails, slugs and endangered honeycreepers in
diphacinone rodenticide baited areas on Hawaii, USA’. Environmental
Toxicology and Chemistry 24: 1557–1567.
Jones, H.P., Tershy, B.R., Zavaleta, E.S., Croll, D.A., Keitt, B.S.,
Finklestein, M.E. and Howald, G.R. (2008). ‘Severity of the effects of
invasive rats on seabirds: A global review’. Conservation Biology 22:
16–26.
Kurle, C.M., Croll, D.A. and Tershy, B.R. (2008). ‘Introduced rats
indirectly change marine rocky intertidal communities from algaeto invertebrate-dominated’. Proceedings of the National Academy of
Sciences 105: 3800–3804.
Lorvelec, O. and Pascal, M. (2005). ‘French attempts to eradicate
non-indigenous mammals and their consequences for native biota’.
Biological Invasions 7: 135–140.
399
C.B. Phillips, K. Brown, K. Broome, C. Green and G. Walker
Phillips, C.B.; K. Brown, K. Broome, C. Green and G. Walker. Criteria to help
evaluate and guide attempts to eradicate terrestrial arthropod pests
Criteria to help evaluate and guide attempts to eradicate terrestrial
arthropod pests
C.B. Phillips1,2, K. Brown3, K. Broome4, C. Green2,5 and G. Walker6
AgResearch, Private Bag 4749, Christchurch, New Zealand 8140. <craig.phillips@agresearch.co.nz>. 2Better Border
Biosecurity, <www.b3nz.org>. 3Department of Conservation, Private Bag 5, Nelson, New Zealand 7042. 4Department
of Conservation, Private Bag 3072, Hamilton, New Zealand 3240. 5Department of Conservation, Private Bag 68908,
Newtown, Auckland, New Zealand 1145. 6Plant & Food Research, Private Bag 92169, Auckland, New Zealand 1142.
1
Abstract Attempts to eradicate invasive terrestrial arthropods are often regarded as gambles. They offer the possibility
of long term freedom from a pest but are usually confronted with substantial uncertainty and come with a range of
technical, economic, environmental, social and political risks. Few guidelines are available for evaluating eradication
attempts against terrestrial arthropods. Here, we build on scientific literature, including six criteria previously developed
to evaluate the feasibility of vertebrate eradications, and our own experiences to define nine criteria that are intended to
both assist experts to evaluate proposed arthropod eradication attempts and guide attempts that are underway. The criteria
are straight forward and easily interpreted, though evaluating them for a particular programme relies on expert group
assessment that will often benefit from rigorous supporting statistical and/or modelling analyses.
Keywords: area wide management, decision support, eradication attempt, eradication campaign, eradication feasibility,
invasive species, pest control
INTRODUCTION
Invasive species eradications make substantial
contributions to conservation (Keitt, et al., 2015; Hoffmann,
et al., 2016), agriculture (Vreysen, et al., 2007b; Suckling,
et al., 2014a) and human health (Kay & Russell, 2013;
Monteiro, et al., 2014). They offer perpetual benefits over
long-term pest damage and associated control costs, but
are usually expensive, can be disruptive both socially and
ecologically, require whole-hearted long-term commitment
from those involved, and success is far from guaranteed
(Myers, et al., 2000; Myers, 2003; Tobin, et al., 2014;
Liebhold, et al., 2016). Thus, eradication attempts invoke
a range of technical, economic, environmental, social
and political risks. Weighing up information about the
potential benefits, costs, risks and probabilities of success
of eradication attempts can be fraught with uncertainty
(Brown, et al., 2019; Cannon, et al., 1999) and demands
that diverse technical issues and societal perspectives be
considered (Simberloff, et al., 2013).
Bomford & O’Brien (1995) defined six criteria to help
evaluate the feasibility of eradicating vertebrate pests.
They drew from lessons learnt in eradicating feral goats
from islands, coypus in England, and infectious human
diseases in various countries. These criteria have become
widely adopted (Brown & Sherley, 2002; Burbidge &
Morris, 2002; Clout & Veitch, 2002; Simberloff, 2003a),
and have been used in New Zealand by the Department of
Conservation (DOC) and Ministry for Primary Industries
to help assess the feasibility of eradicating various
vertebrate and arthropod pests (Cromarty, et al., 2002;
Ashcroft, et al., 2010). We were members of a Technical
Advisory Group convened by DOC to assist it with its
eventually successful attempt to eradicate a non-native
butterfly, Pieris brassicae (Lepidoptera: Pieridae), from
New Zealand (Phillips, et al., 2016; Brown, et al., 2019).
This Palearctic butterfly was regarded as a major risk to
New Zealand’s 79 native (mostly endemic) Brassicaceae
species, many of which were already at risk (Hasenbank,
et al., 2011; de Lange, et al., 2013), and also to cultivated
exotic brassicas. We began using Bomford & O’Brien's
(1995) criteria while evaluating the P. brassicae eradication
programme and found them useful, though not entirely
appropriate for arthropods. Moreover, we were cognisant
of valuable insights about arthropod eradications that had
been described in the literature since Bomford & O’Brien
(1995). Thus, we refined and added to their criteria near
the outset of the P. brassicae eradication attempt, then
used these modified criteria throughout the programme to
help both evaluate if the campaign should continue and
identify the improvements required to maximise its chance
of succeeding.
We present our refined set of criteria here with the aim
of assisting others with expertise in pest eradication and
arthropod ecology to evaluate and guide further arthropod
eradication attempts. Other authors have summarised
the elements and processes needed to mount an effective
eradication campaign (Cromarty, et al., 2002; Hosking,
2002a; Hosking, 2002b; Vreysen, et al., 2007b; Pacific
Invasives Initiative, 2013). However, to our knowledge,
criteria developed specifically to help evaluate attempts
to eradicate terrestrial arthropods have not previously
been documented. Information about previous eradication
programmes has recently been compiled in an on-line
database (Kean, et al., 2018), yet much valuable information
about eradication attempts remains either as grey literature
or unrecorded, which impedes improvements in eradication
methods (Myers, 2003; Hoffmann, et al., 2011). Thus, we
endeavour to document some of our own lessons here.
We do not: review any eradication programmes
(Vreysen, et al., 2007b; Hoffmann, et al., 2011; Suckling,
et al., 2014a); discuss the growing ecological knowledge
and developing technologies that are steadily increasing
the potential for eradication attempts to succeed (Vreysen,
et al., 2007b; Liebhold, et al., 2016; Alphey & Bonsall,
2017; Scott, et al., 2017); or discuss the enormous benefit
of protecting countries or regions from invasions by
new pests and diseases (Leung, et al., 2002; Hoffmann,
et al., 2011; Lovett, et al., 2016). Nor do we attempt to
provide a list of criteria that must be irrefutably met before
choosing to initiate – or persevere with – an eradication
attempt. Rather, we aim to list some readily interpretable,
easily used criteria to assist constructive discussion,
decision making and planning within the broader context
of what is at stake if the pest is allowed to persist and
spread, the pest’s priority relative to other problems, and
the availability of the resources, expertise and personnel
required to mount an effective campaign. Evaluating the
criteria will often benefit from expert group assessment and
rigorous supporting statistical and/or modelling analyses.
Eradication attempts usually involve many uncertainties,
In:
400C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 400–404. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Phillips, et al.: Criteria to eradicate terrestrial arthropod pests
and the criteria should help to identify those that are most
important to resolve as programmes proceed. Certainly, we
will have missed our goal if the criteria impede prompt,
effective action (Simberloff, 2003a; Simberloff, 2003b;
Martin, et al., 2012; Sims & Finnoff, 2013).
METHODS
We adapted the six criteria of Bomford & O’Brien
(1995) to make them clearer and more directly applicable
to arthropod eradications. We reviewed science literature
about factors that influence the success of eradication
attempts, and used the insights gained from the science
publications to further refine the six criteria. Based both
on the literature and our own experiences of eradication
attempts, we also developed three additional criteria.
RESULTS
The nine criteria are listed below, each with clarifying
comments and, where available, supporting evidence from
the literature. The list begins with criteria that deal mainly
with details of the species being considered, the tools
available to suppress it, and the physical environment in
which it occurs, and ends with those that relate more to
the societal and organisational context of the eradication
attempt. Criteria 1–3, 5, 7 and 8 are based on the six
criteria of Bomford & O’Brien (1995), though we modified
them to make them clearer and more directly applicable
to terrestrial arthropod eradications. Criterion 6 is from
Pacific Invasives Initiative (2013), and we added criteria 4
and 9 based both on recent research (Pluess, et al., 2012a;
Pluess, et al., 2012b; Tobin, et al., 2014; Buddenhagen
& Tye, 2015) and our own experiences (Cromarty, et al.,
2002; Brown & Brown, 2015; Keitt, et al., 2015; Phillips,
et al., 2016; Brown, et al., 2019).
1. The pest population can be forced to decline from
one generation to the next, irrespective of its
density.
This is a re-wording of criterion 1 of Bomford &
O’Brien (1995): Rate of removal exceeds rate of increase
at all population densities. Myers, et al. (2000), in a
review that covered eradications of species from several
phyla including arthropods, also emphasised that the pest
must be susceptible to control. Our changes recognise that
terrestrial arthropods typically have several life stages per
generation, which will likely have differing susceptibilities
to control. Thus, it may be acceptable for particular life
stages to numerically increase (e.g., egg-stage offspring
may outnumber their adult parents) provided the overall
effect of control measures is to cause inter-generational
declines. Moreover, the availability of tools to suppress the
pest population both at very high and at very low population
densities should be considered. The eradication attempt
must be capable of driving the population to extinction once
it has been suppressed to very low densities and becomes
more difficult to detect. Some commonly used pest control
methods, such as host plant removal, biological control,
and insecticide applications, do not require direct detection
of pest individuals, and have potential to be effective across
a range of population densities. Some species suffer Allee
effects at low population densities, which can drive them to
extinction once they have been suppressed to below critical
density thresholds (Blackwood, et al., 2012; Liebhold, et
al., 2016).
2. Every pest individual must be at risk of control at
some stage of its development.
This is a re-wording of criterion 3 of Bomford &
O’Brien (1995): All reproductive animals must be at risk.
Our changes recognise that terrestrial arthropods typically
have several life stages per generation, and these will likely
have differing susceptibilities to control. A combination
of control techniques targetting different life stages could
increase the likelihood the pest will be successfully
eradicated (Blackwood, et al., 2012; Suckling, et al.,
2014b; Hoffmann, et al., 2016). Control methods might
include augmenting natural enemies that already occur in
or near the treated area to increase predation or parasitism
of the pest (Montoya, et al., 2007; Hogg, et al., 2013;
Richards, et al., 2016).
3. Pest individuals can be detected at low population
densities.
This is a minor re-wording of criterion 4 of Bomford
& O’Brien (1995): Animals can be detected at low
densities. It is supported by many studies (Myers, et al.,
2000; Simberloff, 2003a; Tobin, et al., 2014). The latter
study analysed factors that influenced the outcomes of
672 arthropod eradication programs and found that high
detectability contributed to success rates. Population
declines must be measurable and, assuming the pest
population is eventually suppressed, management tools
should be adequate to confirm it has been eradicated. Are
effective lures, attractants or traps available, or can they
be developed in a timely fashion? When such tools are
unavailable and eradication attempts depend on visual
searches, they are more successful when targeting easilyobserved foliage-feeding species, rather than species that
occupy more cryptic niches such as roots, fruit or stems
(Tobin et al., 2014). Moreover, programmes without
sensitive detection tools that capitalise on citizen reports
of sightings have higher probabilities of success than those
that do not (Tobin et al., 2014). Thus, detection will be more
likely when: effective attractants are available; the pest and
its feeding damage are conspicuous and easily recognised;
the pest’s host plants are low growing, easily searched
and of interest to gardeners, commercial growers and/or
citizen ecologists; and citizen surveillance is supported
by effective outreach programmes. Some control methods
may impede detection. For example, using pheromones
to disrupt mating will reduce their efficacy as lures for
detection (Suckling, et al., 2014b).
4. Success is favoured by small spatial extent of the
population.
Bomford & O’Brien (1995) acknowledged pest
population spatial extent as important under ‘Other factors’.
The meta-analysis of Tobin, et al. (2014) considered
672 arthropod eradication programs that involved pest
infestations ranging in area from about 0.1 km2 to about
100,000 km2. Overall, there was a base rate of 59% success,
and the spatial extent of the targeted population was the
most important factor explaining variation around this rate
(Tobin, et al., 2014). Population spatial extent was also
recognised as a critically important factor in the outcomes
of 136 eradication programs against invertebrates, plants
and plant pathogens (Pluess, et al., 2012a). When infested
areas are small, eradication attempts are less expensive
and more likely to be successful (Myers, et al., 2000;
Simberloff, 2003a; Brockerhoff, et al., 2010; Pluess, et
al., 2012a; Tobin, et al., 2014). This is why “wait and see”
responses to detections of new pests are seldom justifiable
even when uncertainty is high (Sims & Finnoff, 2013).
5. Immigration and emigration can be prevented.
Here, we added ‘and emigration’ to criterion 2 of
Bomford & O’Brien (1995) (Immigration prevented)
because an attempt to eradicate a localised arthropod
population will fail if individuals disperse from the
401
Island invasives: scaling up to meet the challenge. Ch 2D Other taxa: Invertebrates
eradication zone and establish new undetected populations
nearby. Myers, et al. (2000) and Hoffmann, et al. (2016)
emphasised that immigration (reinvasion) must be
prevented, and Bomford & O’Brien (1995) acknowledged
pest dispersal rates as important under ‘Other factors’.
Attempts to eradicate isolated localised populations (e.g.
on islands or other geographically isolated areas) might
benefit from low likelihoods of natural pest dispersal in or
out of the eradication zone (Myers, et al., 2000), though an
analysis of 173 eradication programmes found no evidence
that eradication attempts were more successful on islands
(Pluess, et al., 2012b). The likelihood the pest will be
transported in or out of the infested area in association
with humans must also be considered, as should the extent
to which this risk can be mitigated (e.g. by implementing
regulatory controls on host plant movements). Pluess, et
al. (2012b) found that implementing sanitary measures to
restrict pest emigration made an important contribution
to eradication success rates. A further consideration is the
capability of the programme to identify when immigration
or emigration is occurring, and to respond effectively to
such processes. Recognising that immigration is occurring
may be challenging, though genetically characterising
the population within the eradication zone could help
to identify new immigrants if they differ genetically
from the initially targeted population (Barr, et al., 2014;
Hiszczynska-Sawicka & Phillips, 2014; Piertney, et al.,
2016). Detecting emigrants will depend on the extent,
intensity and efficacy of active and passive surveillance
outside the known infested area.
6. Environmental impacts of the programme are
acceptable.
Most methods used to manage pests will have nontarget impacts (Bomford & O’Brien, 1995; Pacific Invasives
Initiative, 2013). These include host plant removal,
biological control, synthetic pesticides, biopesticides and
traps (e.g. due to by-catch). Eradicating a pest from an
ecosystem could release other non-native species from
competition, predation or parasitism, thus solving one
problem while exacerbating another (Myers, et al., 2000).
Decision makers must consider if such impacts will be
reversible and/or socially and environmentally acceptable,
and if they will be substantially less than those likely to
be sustained if the pest became permanently established
and more widely distributed. If the infested area being
treated is small and the expected term of the programme
is short, then environmental impacts might be ephemeral
because those non-target species negatively impacted by
the eradication programme may be able to recover once the
programme ends.
7. Benefit-cost analysis favours eradication over
control.
This is a minor re-wording of criterion 5 of Bomford
& O’Brien (1995): Discounted benefit-cost analysis
favours eradication over control. It was also listed by
Pacific Invasives Initiative (2013). We omitted the word
‘discounted’ from Bomford & O’Brien's (1995) criterion
because discounting in benefit-cost analysis remains
controversial (Gollier & Hammitt, 2014; Hockley, 2014).
Myers, et al. (2000) acknowledged that evaluating the
benefits and costs of eradication is difficult, and contended
that the benefits of eradication are often over estimated
and the costs of eradication under estimated. Nevertheless,
many successful eradication programmes have been
regarded as highly cost-effective (Brockerhoff, et al., 2010;
Buddenhagen & Tye, 2015; Scott, et al., 2017). Benefitcost analyses provide useful frameworks for aggregating
information about an eradication attempt to support decision
making. However, they must often include educated
402
guesses about parameter values, struggle to quantify the
value of biodiversity and ecosystem services, and seldom
address uncertainty (Born, et al., 2005; Epanchin‐Niell &
Hastings, 2010; Simberloff, et al., 2013; Hockley, 2014;
Brown, et al., 2019).
8. Suitable social, political, legal and institutional
environment.
This is a minor re-wording of criterion 6 of Bomford
& O’Brien (1995): Suitable socio-political environment. It
was emphasised by Buddenhagen & Tye (2015) and similar
criteria were listed by Pacific Invasives Initiative (2013)
and Simberloff (2003a). Myers, et al. (2000) also stressed
that funding must be sufficient and lines of authority clear.
Our changes more clearly specify the need for eradication
programmes to be supported by every facet of society that
has an important role or stake in the programme. Those
evaluating eradication attempts must ask questions like:
Will property owners allow or support eradication activities
on their land? Would those implementing the eradication
programme have legal authority to implement control
actions on private and public land? Will the programme
be supported by stakeholders such as local and regional
authorities, farmer organisations and environmental
advocacy groups? Will all management levels of the
institution(s) attempting the eradication remain fully
committed – especially financially – to the programme for
the long haul?
9. Programme is effectively managed, and its status is
reliably monitored and accurately recorded.
Efficient, meticulous and effective planning and
management are critical to eradication success (Cromarty,
et al., 2002), as are clear lines of authority (Myers, et al.,
2000; Simberloff, 2003a). These programme attributes must
be supported by efficient and robust data collection and
analysis to enable progress to be monitored, assumptions
tested, weaknesses identified, and improvements devised
and implemented (Vreysen, et al., 2007a). Brown & Brown
(2015) suggested that systematic and persistent effort by
individuals with a “completer-finisher” personality type
(Belbin, 2010) or an “eradication attitude” can increase
the likelihood of success. For arthropod eradications, it
is clearly important to involve people with expertise in
arthropod ecology and management.
DISCUSSION
We propose that the nine criteria can help to focus
discussion and evaluate and guide attempts to eradicate
terrestrial arthropods. We repeatedly scored the criteria
throughout the P. brassicae eradication programme
(Phillips, et al., 2016) and, although our individual
assessments often differed, they always provided a valuable
basis for discussion and planning. Moreover, our individual
assessments all became progressively more optimistic
as the programme proceeded and uncertainty declined
(Phillips, et al., 2015). In fact, optimism grew even as the
pest’s known geographical distribution increased because
we also gained confidence that the pest was detectable and
controllable.
We found it useful to classify each criterion as being
either ‘not met’, ‘marginally met’ or ‘substantially met’.
These qualitative terms recognised that criteria can be met
to varying degrees and using just three classes simplified
the assessment process and eased interpretation. Criteria
were classified as ‘not met’ if the eradication attempt was
likely to fail unless improvements to that aspect of the
programme were urgently made. Criteria were considered
‘marginally met’ if there was some evidence the criterion
could be (or was being) met, but knowledge gaps caused
Phillips, et al.: Criteria to eradicate terrestrial arthropod pests
uncertainty to be high and made assessing the criterion
difficult. This classification also signalled a need for action
because important knowledge gaps had to be addressed
to ensure eradication feasibility. Criteria were scored as
‘substantially met’ when these elements of the eradication
attempt appeared (likely to be) effective. Improvements
to aspects of the eradication classified as ‘not met’ or
‘marginally met’ were regarded as critical and urgent, and
improvements to those classified as ‘substantially met’
were regarded as desirable.
In the context of vertebrate eradications, Bomford &
O’Brien (1995) considered criteria 1, 2 and 5 (numbers
as used in the main text of the Results in this paper) as
essential to success, and criteria 3, 7 and 8 as desirable,
though they emphasised that negatives in the latter
three criteria “will greatly reduce the feasibility and
desirability of eradication”. With our criteria for terrestrial
arthropod eradications, we suggest that all of the criteria
except numbers 4 (small spatial extent) and 7 (benefitcost analysis) will need to be substantially met before
eradication is eventually achieved. However, it may be
reasonable to initiate an eradication attempt before many
of the critical criteria are substantially met. This is because,
with thoughtful management, new knowledge and tools
will often be developed during the course of a programme
(Vreysen, et al., 2007b; Scott, et al., 2017) that will rectify
some or all of its deficiencies and/or enable the criteria to
be scored with more confidence. Indeed, the criteria aim
to highlight those aspects of programmes that most need
improvement. In cases where few critical criteria are
substantially met, it will be particularly important to specify
conditions under which the attempt will cease (e.g. when
a key programme deficiency is not rectified by a specified
date) in order to minimise expenditure on programmes that
are doomed to failure.
The capability to robustly evaluate programme progress
and confidently reclassify criteria is especially dependent
on criterion 9 (excellent management). It is also highly
desirable that a (proposed) programme substantially meets
criteria 4 (small spatial extent) and 7 (benefit-cost analysis).
Yet, with criterion 4, there are examples of arthropod
populations with very large spatial extents that have been
successfully eradicated (Vreysen, et al., 2007b; Monteiro,
et al., 2014; Scott, et al., 2017), thus scores of ‘not met’
or ‘marginally met’ may be acceptable in cases where the
other criteria for achieving eradication can be substantially
met and resources are available to work effectively across
large geographical areas. For criterion 7, perceptions of
the potential economic and/or environmental benefits of
an eradication attempt will strongly influence the level of
risk that is deemed acceptable when deciding whether to
initiate or persist with the attempt. However, the previously
mentioned limitations of benefit-cost analyses (Born, et
al., 2005; Epanchin‐Niell & Hastings, 2010; Simberloff, et
al., 2013; Hockley, 2014; Brown, et al., 2019) combined
with overwhelming evidence of negative impacts of
many invaders suggests that the precautionary principle
(Simberloff, et al., 2013) should be applied particularly to
criterion 7, and scores of ‘marginally met’ may be adequate
to justify action.
We intend the criteria to be used by people with
expertise in pest eradication and arthropod ecology and
management. During the P. brassicae eradication attempt,
we found it productive to discuss programme performance
against each criterion as a group because our individual
perspectives often initially differed. Our evaluations of one
or more criteria were frequently supported by statistical
and/or modelling analyses of data being collected by the
programme. Eventually we would reach a consensus that
enabled us to provide better advice to the programme
than any one of us could have alone. Thus, we advocate
using the criteria in fora similar to the ‘Technical Advisory
Groups’ that are often applied in New Zealand to support
management decision making. We believe that using the
criteria to help evaluate the feasibility of an eradication
attempt and its progress towards success will help to
improve decision making and increase programme
success rates. However, the quality of decision making
will of course also depend on the values and motivations
of decision makers, the experience and problem-solving
abilities of the expert group, the quality of data analysis,
and the preparedness of all involved to fill knowledge gaps
and take timely action.
ACKNOWLEDGEMENTS
We thank Drs John Kean (AgResearch, Hamilton),
Max Suckling (Plant and Food Research, Lincoln), Graeme
Bourdôt (AgResearch, Lincoln), George Gill (Ministry for
Primary Industries, Wellington), David Teulon (Plant and
Food Research, Lincoln), Pablo Garcia Diaz (Landcare
Research, Lincoln) and two anonymous referees for
reviewing previous versions of this manuscript. The senior
author was funded by AgResearch via its contribution to the
Better Border Biosecurity research collaboration (<www.
b3nz.org>). The authors declare no competing interests.
REFERENCES
Alphey, N. and Bonsall, M.B. (2017). ‘Genetics‐based methods for
agricultural insect pest management’. Agricultural and Forest
Entomology 20: 131–140.
Ashcroft, T., Richardson, E., Voice, D. and Dustow, K. (2010). Strategies
and Feasibility to Undertake Eradication of Pieris brassicae (GWCB)
in New Zealand (No. version 3 (final)). Wellington, New Zealand:
Ministry of Agriculture and Forestry.
Barr, N., Ruiz-Arce, R. and Armstrong, K. (2014). ‘Using Molecules to
Identify the Source of Fruit Fly Invasions’. In: T. Shelly, N. Epsky, E.B.
Jang, J. Reyes-Flores and R. Vargas (eds.) Trapping and the Detection,
Control, and Regulation of Tephritid Fruit Flies, pp. 321–378.
Dordrecht, The Netherlands: Springer.
Belbin, R. (2010). ‘Management Teams: Why they Succeed or Fail’.
Oxford, UK: Elsevier.
Blackwood, J.C., Berec, L., Yamanaka, T., Epanchin-Niell, R.S., Hastings,
A. and Liebhold, A.M. (2012). ‘Bioeconomic synergy between tactics
for insect eradication in the presence of Allee effects’. Proceedings of
the Royal Society of London B: Biological Sciences 279: 2807–2815.
Bomford, M. and O’Brien, P. (1995). ‘Eradication or control for vertebrate
pests?’. Wildlife Society Bulletin 23: 249–255.
Born, W., Rauschmayer, F. and Brauer, I. (2005). ‘Economic evaluation
of biological invasions—a survey’. Ecological Economics 55: 321–336.
Brockerhoff, E.G., Liebhold, A.M., Richardson, B. and Suckling, D.M.
(2010). ‘Eradication of invasive forest insects: concepts, methods, costs
and benefits’. New Zealand Journal of Forestry Science 40: 117–135.
Brown, K. and Brown, D. (2015). ‘Control to eradication of Tradescantia
fluminensis on Stephens Island (Takapourewa): the importance of
systematic and persistent effort’. DOC Research & Development Series
15. Wellington, NZ: Department of Conservation.
Brown, K., Phillips, CB, Broome, K., Green, C., Toft, R, and Walker,
G. (2019). ‘Feasibility of eradicating the large white butterfly (Pieris
brassicae) from New Zealand: Data Gathering to Inform Decisions
about the Feasibility of Eradication.’ In: C.R. Veitch, M.N. Clout, A.R.
Martin, J.C. Russell and C.J. West (eds.) Island invasives: scaling up to
meet the challenge, pp. 364–369. Occasional Paper SSC no. 62. Gland,
Switzerland: IUCN.
Brown, K.P. and Sherley, G.H. (2002). ‘The eradication of possums
from Kapiti Island, New Zealand’. In: C.R. Veitch and M.N. Clout
(eds.) Turning the tide: the eradication of invasive species, pp. 46–52.
Occasional Paper SSC no. 28. IUCN SSC Invasive Species Specialist
Group. IUCN, Gland, Switzerland and Cambridge, UK.
Buddenhagen, C.E. and Tye, A. (2015). ‘Lessons from successful plant
eradications in Galapagos: commitment is crucial’. Biological Invasions
17: 2893–2912.
Burbidge, A.A. and Morris, K.D. (2002). ‘Introduced mammal
eradications for nature conservation on Western Australian Islands: A
Review’. In: C.R. Veitch and M.N. Clout (eds.) Turning the tide: the
eradication of invasive species, pp. 64–70. Occasional Paper SSC
no. 28. IUCN SSC Invasive Species Specialist Group. IUCN, Gland,
Switzerland and Cambridge, UK.
Cannon, R., Pemberton, A. and Bartlett, P. (1999). ‘Appropriate measures
for the eradication of unlisted pests’. EPPO Bulletin 29: 29–36.
403
Island invasives: scaling up to meet the challenge. Ch 2D Other taxa: Invertebrates
Clout, M.N. and Veitch, C.R. (2002). ‘Turning the tide of biological
invasion: the potential for eradicating invasive species.’ In: C.R. Veitch
and M.N. Clout (eds.) Turning the tide: the eradication of invasive
species, pp. 1–3. Occasional Paper SSC no. 28. IUCN SSC Invasive
Species Specialist Group. IUCN, Gland, Switzerland and Cambridge,
UK.
Cromarty, P.L., Broome, K.G., Cox, A., Empson, R.A., Hutchinson,
W.M. and McFadden, I. (2002). ‘Eradication planning for invasive
alien animal species on islands – the approach developed by the New
Zealand Department of Conservation’. In: C.R. Veitch and M.N. Clout
(eds.) Turning the tide: the eradication of invasive species, pp. 85–91.
Occasional Paper SSC no. 28. IUCN SSC Invasive Species Specialist
Group. IUCN, Gland, Switzerland and Cambridge, UK.
de Lange, P.J., Rolfe, J.R., Champion, P.D., Courtney, S., Heenan, P.B.,
Barkla, J.W., Cameron, E.K., Norton, D.A. and Hitchmough, R. (2013).
Conservation Status of New Zealand Indigenous Vascular Plants, 2012.
Wellington, NZ: Department of Conservation.
Epanchin‐Niell, R.S. and Hastings, A. (2010). ‘Controlling established
invaders: integrating economics and spread dynamics to determine
optimal management’. Ecology Letters 13: 528–541.
Gollier, C. and Hammitt, J.K. (2014). ‘The long-run discount rate
controversy’. Annual Review of Resource Economics 6: 273–295.
Hasenbank, M., Brandon, A. and Hartley, S. (2011). ‘White butterfly
(Pieris rapae) and the white rust Albugo candida on Cook’s scurvy
grass (Lepidium oleraceum)’. New Zealand Journal of Ecology 35:
69–75.
Hiszczynska-Sawicka, E. and Phillips, C.B. (2014). ‘Mitochondrial
cytochrome c oxidase subunit 1 sequence variation in New Zealand and
overseas specimens of Pieris brassicae (Lepidoptera: Pieridae)’. New
Zealand Plant Protection 67: 8–12.
Hockley, N. (2014). ‘Cost-benefit analysis: a decision-support tool or a
venue for contesting ecosystem knowledge?’ Environment and Planning
C: Government & Policy 32: 283–300.
Hoffmann, B., Davis, P., Gott, K., Jennings, C., Joe, S., Krushelnycky,
P., Miller, R., Webb, G. and Widmer, M. (2011). ‘Improving ant
eradications: details of more successes, a global synthesis and
recommendations’. Aliens 31: 16–23.
Hoffmann, B.D., Luque, G.M., Bellard, C., Holmes, N.D. and Donlan,
C.J. (2016). ‘Improving invasive ant eradication as a conservation tool:
a review’. Biological Conservation 198: 37–49.
Hogg, B.N., Wang, X.-G., Levy, K., Mills, N.J. and Daane, K.M. (2013).
‘Complementary effects of resident natural enemies on the suppression
of the introduced moth Epiphyas postvittana’. Biological Control 64:
125–131.
Hosking, G. (2002a). An Emergency Response Guide for Forestry
Incursions. Rotorua, NZ: Forest Research.
Hosking, G. (2002b). ‘Incursion response – back to basics’. New Zealand
Journal of Forestry 47: 24–27.
Kay, B. and Russell, R. (2013). ‘Mosquito Eradication: The Story of
Killing Campto’. Collingwood, Victoria, Australia: CSIRO Publishing.
Kean, J.M., Suckling D.M., Sullivan N.J., Tobin P.C., Stringer L.D.,
Smith G.R., Kimber B, Lee D.C., Flores Vargas R., Fletcher J., Macbeth
F., McCullough D.G., Herms D.A. et al. (2018). ‘Global Eradication
and Response Database’. <http://b3.net.nz/gerda>. Accessed 20 April
2018.
Keitt, B., Griffiths, R., Boudjelas, S., Broome, K., Cranwell, S., Millett,
J., Pitt, W. and Samaniego-Herrera, A. (2015). ‘Best practice guidelines
for rat eradication on tropical islands’. Biological Conservation 185:
17–26.
Leung, B., Lodge, D.M., Finnoff, D., Shogren, J.F., Lewis, M.A. and
Lamberti, G. (2002). ‘An ounce of prevention or a pound of cure:
bioeconomic risk analysis of invasive species’. Proceedings of the
Royal Society of London (B) Biological Sciences 269: 2407–2413.
Liebhold, A.M., Berec, L., Brockerhoff, E.G., Epanchin-Niell,
R.S., Hastings, A., Herms, D.A., Kean, J.M., McCullough, D.G.,
Suckling, D.M. and Tobin, P.C. (2016). ‘Eradication of invading
insect populations: from concepts to applications’. Annual Review of
Entomology 61: 335–352.
Lovett, G.M., Weiss, M., Liebhold, A.M., Holmes, T.P., Leung, B.,
Lambert, K.F., Orwig, D.A., Campbell, F.T., Rosenthal, J. and
McCullough, D.G. (2016). ‘Nonnative forest insects and pathogens in
the United States: impacts and policy options’. Ecological Applications
26: 1437–1455.
Martin, T.G., Nally, S., Burbidge, A.A., Arnall, S., Garnett, S.T.,
Hayward, M.W., Lumsden, L.F., Menkhorst, P., McDonald‐Madden,
E. and Possingham, H.P. (2012). ‘Acting fast helps avoid extinction’.
Conservation Letters 5: 274–280.
Monteiro, F.A., Shama, R., Martins, A.J., Gloria-Soria, A., Brown, J.E.
and Powell, J.R. (2014). ‘Genetic diversity of Brazilian Aedes aegypti:
patterns following an eradication program’. PLOS Neglected Tropical
Diseases 8: e3167.
404
Montoya, P., Cancino, J., Zenil, M., Santiago, G. and Gutierrez, J. (2007).
‘The Augmentative Biological Control Component in the Mexican
National Campaign against Anastrepha spp. Fruit Flies’. In: M.J.
Vreysen, A.S. Robinson and J. Hendrichs, (eds.) Area-wide Control of
Insect Pests: From Research to Field Implementation, pp. 661–670.
Dordrecht, The Netherlands: Springer Science & Business Media.
Myers, J.H. (2003). ‘Eradication: Is it Ecologically, Financially,
Environmentally, and Realistically Possible’. In: D.J. Rapport, B.L.
Lasley, D.E. Rolston, N.O. Nielsen, C.O. Qualset and A.B. Damania
(eds.) Managing for Healthy Ecosystems, pp. 533–539. Boca Raton:
Lewis Publishers.
Myers, J.H., Simberloff, D., Kuris, A.M. and Carey, J.R. (2000).
‘Eradication revisited: dealing with exotic species’. Trends in Ecology
and Evolution 15: 316–320.
Pacific Invasives Initiative (2013). Resource kit for rodent and cat
eradication.
<http://www.pacificinvasivesinitiative.org/rk/project/2_
Feasibilty_Study.html>. Accessed 9 May 2013.
Phillips, C.B., Brown, K., Green, C., Broome, K., Toft, R., Shepherd, M.,
Bayley, S. and Rees, J. (2015). Pieris brassicae (Great White Butterfly)
Eradication Annual Report 2014/15 (Department of Conservation
Annual Report No. R77017), Pieris brassicae Eradication. Nelson,
New Zealand: Department of Conservation.
Phillips, C.B., Brown, K., Green, C., Broome, K., Toft, R., Shepherd, M.,
Bayley, S. and Rees, J. (2016). Pieris brassicae (Great White Butterfly)
Eradication Annual Report 2015/16 (Department of Conservation
Annual Report), Pieris brassicae Eradication. Nelson, New Zealand:
Department of Conservation.
Piertney, S.B., Black, A., Watt, L., Christie, D., Poncet, S. and
Collins, M.A. (2016). ‘Resolving patterns of population genetic and
phylogeographic structure to inform control and eradication initiatives
for brown rats Rattus norvegicus on South Georgia’. Journal of Applied
Ecology 53: 332–339.
Pluess, T., Cannon, R., Jarošík, V., Pergl, J., Pyšek, P. and Bacher,
S. (2012a). ‘When are eradication campaigns successful? A test of
common assumptions’. Biological Invasions 14: 1365–1378.
Pluess, T., Jarošík, V., Pyšek, P., Cannon, R., Pergl, J., Breukers, A. and
Bacher, S. (2012b). ‘Which factors affect the success or failure of
eradication campaigns against alien species?’. PLOS ONE 7: e48157.
Richards, N., Hardwick, S., Toft, R. and Phillips, C. (2016). ‘Enhancing
parasitism of Pieris brassicae pupae by Pteromalus puparum’. New
Zealand Plant Protection 69: 126–132.
Scott, M.J., Concha, C., Welch, J.B., Phillips, P.L. and Skoda, S.R. (2017).
‘Review of research advances in the screwworm eradication program
over the past 25 years’. Entomologia Experimentalis et Applicata 164:
226–236.
Simberloff, D. (2003a). ‘Eradication—preventing invasions at the outset’.
Weed Science 51: 247–253.
Simberloff, D. (2003b). ‘How much information on population biology
is needed to manage introduced species?’. Conservation Biology 17:
83–92.
Simberloff, D., Martin, J.-L., Genovesi, P., Maris, V., Wardle, D.A.,
Aronson, J., Courchamp, F., Galil, B., García-Berthou, E. and Pascal,
M. (2013). ‘Impacts of biological invasions: what’s what and the way
forward’. Trends in Ecology and Evolution 28: 58–66.
Sims, C. and Finnoff, D. (2013). ‘When is a “wait and see” approach
to invasive species justified?’. Resource and Energy Economics 35:
235–255.
Suckling, D., Kean, J.M., Stringer, L.D., Cáceres‐Barrios, C., Hendrichs,
J., Reyes‐Flores, J. and Dominiak, B.C. (2014a). ‘Eradication of
tephritid fruit fly pest populations: outcomes and prospects’. Pest
management science 72: 456–465.
Suckling, D., Stringer, L.D., Stephens, A.E., Woods, B., Williams,
D.G., Baker, G. and El‐Sayed, A.M. (2014b). ‘From integrated pest
management to integrated pest eradication: technologies and future
needs’. Pest Management Science 70: 179–189.
Tobin, P., Kean, J., Suckling, D., McCullough, D., Herms, D. and
Stringer, LD (2014). ‘Determinants of successful arthropod eradication
programs’. Biological Invasions 16: 401–414.
Vreysen, M., Gerardo-Abaya, J. and Cayol, J. (2007a). ‘Lessons from
Area-wide Integrated Pest Management (AW-IPM) Programmes with
an SIT Component: An FAO//IAEA Perspective’. In M. Vreysen, A.
Robinson and J. Hendrichs (eds.) Area-wide Control of Insect Pests:
From Research to Field Implementation, pp. 723–744. Dordrecht, The
Netherlands: Springer Science & Business Media.
Vreysen, M., Robinson, A. and Hendrichs, J. (eds.) (2007b). Area-wide
Control of Insect Pests: From Research to Field Implementation’.
Dordrecht, The Netherlands: Springer Science & Business Media.
R. Sandodden
Sandodden, R. Eradication of invasive alien crayfish: past experiences and further possibilities
Eradication of invasive alien crayfish: past experiences
and further possibilities
R. Sandodden
Norwegian Veterinary Institute, Environmental and Biosecurity Measures, Pb. 5695 Sluppen, NO-7485 Trondheim,
Norway. <roar.sandodden@vetinst.no>.
Abstract The EU regulation 1143/2014 “On the prevention and management of the introduction and spread of invasive
alien species” entered into force on 1 January 2015. On 13 July 2016, the EU list of invasive alien species that require
action was adopted. The list includes five different crayfish species. Member states will be required to take measures for
early detection and rapid eradication of these species. Except for some eradications performed in the United Kingdom and
Norway, there has not been much effort put into eradication of invasive crayfish species throughout Europe. The reasons
for this are probably complex and differ between member states. Are the main reasons legislative constraints, ability to
eradicate or lack of knowledge and experience? Is eradication of alien crayfish possible and desirable, and what is left to
save in Europe? Focus could be put into identifying or creating island populations of special concern and preserve them
for the future survival of European native crayfish populations. Eradication measures should be considered as an option
in this work. What are the experiences from completed eradication efforts in Europe? Two crayfish eradications have
been performed in Norway, and both have been successful. The eradications were performed in locations with several
ponds and small streams and performed using the synthetic pyrethroid-based pharmaceutical BETAMAX VET®. Both
legislative and funding constraints seem less prominent as successful eradications have been confirmed. Time will show
if this trend will spread throughout Europe.
Keywords: crayfish management, invasive alien species, IAS, pyrethroids, signal crayfish
INTRODUCTION
The EU regulation 1143/2014 “On the prevention and
management of the introduction and spread of invasive
alien species” (<http://data.europa.eu/eli/reg/2014/1143/
oj>) entered into force on 1 January 2015. On 13 July 2016
the EU list of invasive alien species, IAS that requires
action was adopted. The list includes five different crayfish
species, spinycheek crayfish (Orconectes limosus), virile
crayfish (Orconectes virilis), signal crayfish (Pacifastacus
leniusculus), red swamp crayfish (Procambarus clarkii),
and marbled crayfish (Procambarus fallax) (<http://data.
europa.eu/eli/reg_impl/2016/1141/oj>). Crayfish are one of
the most successful and widely distributed invasive species
in the world (Holdich, et al., 2014). Twenty eight different
crayfish species have been translocated from their native
range, and seven of them have been identified with invasive
potential (Gherardi, 2010). At least ten non-native species
of crayfish have been introduced to Europe (Souty-Grosset,
et al., 2006). The five indigenous European freshwater
crayfish species are all threatened by different factors, but
the most detrimental is probably the North American signal
crayfish Pacifastacus leniusculus and the crayfish plague
caused by the oomycete parasite Aphanomyces astaci
(Holdich & Sibley, 2009). Signal crayfish are natural hosts
for the crayfish plague (Unestam, 1972), the causal agent of
crayfish plague, and a disease lethal to European freshwater
crayfish (Alderman, et al., 1990; Souty-Grosset, et al.,
2006), causing dramatic population reduction and in many
cases extinction (Holdich, et al., 1999). The signal crayfish
exhibits a number of biological adaptions which allow it
to tolerate extreme environmental conditions (McMahon,
2002). This flexibility may facilitate the further spread of
both he crayfish and the crayfish plague.
Except for some eradications performed in the United
Kingdom (Peay, et al., 2006) and Norway (Sandodden &
Bardal, 2010; Sandodden & Johnsen, 2010), there has not
been much effort put into eradication of invasive crayfish
species throughout Europe using chemicals. The reasons
for this are probably complex and differ between countries.
Are the main reasons legislative constraints, unwillingness
or lack of knowledge and experience? Is eradication of
alien crayfish possible and desirable, and what is left to
save in Europe?
The EU regulation on invasive alien species includes
restrictions on keeping, importing, selling, breeding and
growing listed species. Taking action as early as possible
and preventing introduction will ensure that unnecessary
suffering of animals is avoided and is more cost effective
than eradication. On the other hand, member states will
be required to take measures for early detection and
rapid eradication of listed species. If a new population is
detected there is an eradication obligation, whereas for
Chemical methods of eradication include the use of
biocides, surfactants and pheromones. Ribbens & Graham
(2004) review the use of biocides for control of crayfish
populations. Organophosphates and organochlorines are
reported to be effective, but these chemicals are known
to bioaccumulate through the food chain (Holdich, et al.
1999). Crayfish can bioaccumulate organochlorines and,
as crayfish are eaten by many predators, this is obviously
important in terms of biomagnification through the food
widely spread species management measures must take
place. The list mainly contains species already present in
the EU, but future updates are expected to introduce more
species not yet present in the EU. Member states select the
measures appropriate to the local conditions and do not
have an obligation to eradicate IAS of Union concern that
are already widely spread in their territory.
Throughout Europe there have been several attempts
to eradicate different crayfish species. Reviews of possible
methods for controlling nuisance populations of alien
crayfish are available (Holdich, et al., 1999; Hiley, 2003;
Ribbens & Graham, 2004; Peay & Hiley, 2006; Freeman,
et al., 2010; Stebbing, et al., 2014). These methods include
different legislative, mechanical, biological and physical
measures, including the use of biocides and pheromones.
Mechanical methods, such as trapping, seining, and
electrofishing can control, but not eradicate crayfish
populations (Holdich, et al., 1999; Hiley, 2003; Peay &
Hiley, 2006). It seems that only chemical based treatments
offer any hope for effective eradication of invasive crayfish
species (Peay, 2001).
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 405–409. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
405
Island invasives: scaling up to meet the challenge. Ch 2D Other taxa: Invertebrates
chain (Ludke, et al., 1971) In contrast, both natural
pyrethrum (Pyblast) and synthetic pyrethroids, which have
been shown to be effective at very low doses, break down
rapidly and do not bioaccumulate (Holdich, et al., 1999;
Hiley, 2003; Peay & Hiley, 2006). Synthetic pyrethrum
is based on the chemical structure and biological activity
of natural pyrethrum, an extract of plants of the genus
Chrysanthemum (Holdich, et al., 1999).
Eversole & Seller (1997) concluded, in a comprehensive
study based on 35 different chemical groups, that synthetic
pyrethroids were the most poisonous to crayfish. Both
natural pyrethrum and synthetic pyrethroids have low
toxicity to birds, mammals, plants and many invertebrates
(Van Wijngaarden, et al., 2005). They are, however, in
varying degrees toxic to non-target fauna, including
crustaceans, insects, arthropods, fish and amphibians
(Mayer & Ellersieck, 1986; Burridge & Haya, 1997).
The environmental fate and degradation of pyrethroid
insecticides were reviewed by Leahey (1979). He concluded
that pyrethroids do not persist in the environment for long
periods, do not accumulate in the biosphere and do not
biomagnify in the food chain. Ecosystem recovery is fairly
rapid, with the toxic effect of pyrethroids lasting from
days to months, and all major animal groups recovering
within a year (Gydemo, 1995). Holdich, et al. (1999) states
that ecosystems can recover fairly rapidly from the toxic
effects of pyrethroids. Compared to natural pyrethrum the
synthetic forms are more toxic, less degradable by light,
more readily available and less expensive (Morolli, et
al., 2006). To date, no crayfish-specific biocide has been
developed.
O`Reilly (2015) suggested that lower concentrations
of the natural pyrethrum may be suitable to eradicate or
control signal crayfish in small standing waterbodies.
Where the risk of damage to non-target species is not an
issue and the water is not being used for another purpose,
cheaper alternative biocides such as synthetic pyrethroids
could be used.
Two successful signal crayfish eradications have been
performed in Norway. On the basis of these results and
EU regulation 1143/2014, more focus should be put into
identifying or creating island populations of special concern
and preserve them for the future survival of European
native crayfish populations. Eradication measures should
and must be considered as an option in this work. The
number of eradication attempts probably will increase in
Europe as both the knowledge base and environmental
impacts increase.
MATERIALS AND METHODS
Both successful eradications were performed in
southern Norway close to the capital Oslo (Fig. 1). The
locations consisted of several ponds and small streams
and involved the application of the synthetic pyrethroidbased pharmaceutical BETAMAX VET®, which is a
cypermethrin-based pharmaceutical originally developed
for treatment of salmon louse (Lepeophtheirus salmonis)
infestation of farmed Atlantic salmon (Salmo salar).
Cypermethrin is a synthetic pyrethroid and a common
agent in many insecticides licensed throughout Europe.
Both eradications involved two separate consecutive
treatments separated by two weeks and a partial drainage
of some of the ponds. The first eradication was performed
in the Dammane watershed in Telemark County, during
May 2008. The watercourse consists of a creek with five
small ponds, the largest measuring approximately 2,000 m2
(Table 1). The treatment is described in detail in Sandodden
& Johnsen (2010). The second eradication was undertaken
using the same pharmaceutical, methods and equipment
in the Oslo & Akershus County at Ostøya, an island in
Oslo-fjord, during October 2009. The treatment involved
six ponds on a golf course. The two largest ponds were
close to 2,200 m2 (Table 1). The treatment is described
in Sandodden & Bardal (2010, in Norwegian). All ponds
were treated with the help of pumps placed in a boat or
on the shore (Fig. 2). The chemical was dispersed both on
the water surface, along the pond bottom and on a 10 m
onshore belt around each pond. Continuous drip stations
were placed at the most upstream location of each creek
or seep to ensure treatment of the whole drainage basin.
This ensured a continuous, constant dosage of BETAMAX
VET® during treatment. In the smallest of seeps, enclosed
water bodies and small upstream creeks, watering cans
were used to dispense a dilution of the chemical. For more
details regarding methods, see Sandodden & Johnsen
(2010).
The requirements set by the Norwegian Food and Safety
Authority for issuing an eradication confirmation after
eradication of signal crayfish are described in Johnsen, et
al., (2010) and state: 1. No crayfish caught during trapping
five to five and a half years after eradication is performed.
2. Noble crayfish (Astacus astacus) placed in cages in
the treated area have shown no signs of crayfish plague
during the last three years of monitoring. 3. Analyses
of water and sediments show no sign of crayfish plague
spores five to five and a half years after eradication. The
methodology is described in Vrålstad, et al., (2009). Based
Table 1 Area, mean depth, volume and BETAMAX VET® used during treatment of the ponds at Dammane
and Ostøya locations.
Dammane
Ostøya
406
Dam 1
Dam 2
Dam 3
Dam 4
Dam 5
Dam 14
Dam 18
Dam 13
Dam 2
Dam 1
Dam 8
Area m2
371
697
1,146
3,154
1,346
2,242
1,400
990
2,200
1,054
370
Mean depth metres
0.82
0.92
2.27
1.92
1.73
3.00
1.80
1.80
1.80
1.80
2.00
Volume m3
303
639
2,602
6,054
1,996
6,726
2,520
1,782
3,960
1,897
370
BETAMAX litres
0.14
0.17
1.41
2.78
0.57
3.56
1.33
0.94
2.09
1.00
0.20
Sandodden: Eradication of invasive crayfish
on the investigations involved in eradication confirmation,
the Norwegian Food and Safety Authority can issue a selfdeclaration of freedom of disease (OIE, 2009).
RESULTS
Dammane
No surviving crayfish was observed or found during
the second treatment or during drainage of the ponds. On
the basis of the Norwegian Food and Safety Authority selfdeclaration of freedom for disease procedure, the County
Governor carried out trials with caged live noble crayfish
in 2010 and 2011. In 2010, a total of 31 male crayfish
were placed in three cages in three of the treated ponds.
The caged crayfish suffered a high mortality during the
trials that lasted for 136 days. Analyses performed at the
Norwegian Veterinary Institute showed that the cause of
death was not crayfish plague. In 2011, a total of 30 male
crayfish were placed in three cages in two of the treated
ponds. The caged crayfish suffered a high mortality during
the trials that lasted for 129 days. Analyses performed at
the Norwegian Veterinary Institute showed that the cause
of death was not caused by crayfish plague. In 2011 a
trapping trial for crayfish was carried out in two of the
treated ponds. No crayfish were caught.
Regarding eradication confirmation, the Norwegian
Food and Safety Authority concluded on the basis of these
results in December 2011 that they could either issue an
eradication confirmation based on the results alone or carry
out trials with caged crayfish for another year. The relatively
new method using molecular investigations based on water
samples in search of crayfish plague spores might be
carried out as an addition, but the more realistic approach
would be caged crayfish trials. The Norwegian Food and
Safety Authority’s final advice was to issue an eradication
confirmation based on the results given in Dammane.
They have not yet issued a formal letter or report declaring
eradication confirmation (Jan Egil Aronsen, Norwegian
Food and Safety Authority pers. comm., 2017).
Ostøya
No surviving crayfish was observed or found during
the second treatment or during drainage of the ponds. The
County Governor carried out trials with caged live noble
crayfish in 2013 and 2014. Cages were placed in five of
the treated ponds. No signs of disease or crayfish plague
were detected. In 2014 a trapping trial for crayfish was
carried out. No crayfish were caught. In June 2014 the
Norwegian Veterinary Institute collected water samples
for analyses of crayfish plague spores in two of the treated
ponds (unpublished data). No spores were detected. On the
basis of these results the County Governor concluded that
the signal crayfish and the crayfish plague is eradicated
form the infected ponds. These are unpublished results but
summarized in a letter from the County Governor dated 17
March 2017 (ref. 2017/1978-1 M-NA).
DISCUSSION
What is left to save in Europe?
There are still significant native crayfish populations
in Europe, which are being decimated through the spread
of introduced invasive non-native crayfish (Gherardi, et
al., 2011). Action to control invasive non-native crayfish
would protect these rare and valuable species. Equally,
the impacts from invasive non-native crayfish are wider,
ranging from damage to river and flood defence banks
(Guan, 1994), through to impact on recreational fisheries.
So, there is a case for action based on both ecological and
socio-economic factors.
Is eradication of alien crayfish possible and desirable?
As this paper shows, there are possibilities for crayfish
eradication. We have the scientific evidence base regarding
the species, their risks and impacts; we have the processes
to make a robust case, tools, techniques and expertise to
take action and now the powers under EU IAS regulations
to make that a reality. It is possible to make robust cases to
government and only by doing this can we tackle the final
funding barrier. Reporting successful eradications should
both inspire and motivate future eradication projects. To
justify the use of chemicals, it is important to conduct and
report the environmental impacts following the eradication
attempts and evaluate these in comparison to not taking
action.
Fig. 1 Dammane and Ostøya locations in southern Norway.
Fig. 2 Boat mounted pump used to apply BETAMAX VET®
during the Dammane and Ostøya treatments.
407
Island invasives: scaling up to meet the challenge. Ch 2D Other taxa: Invertebrates
Lack of eradication projects
The answer to why there have been no eradication
projects until now is probably mostly a combination
of legislative constraints and lack of experience. Not
all European countries are EU- members and most
European countries have national legislation regulating
the use of chemicals in freshwater. Both local and national
regulatory agencies seem not to know where to start and
which legislation to apply when trying to implement an
eradication strategy involving chemicals. The answer
to why not in future is, while complex, now only down
to making a strong case to secure political backing and
funding to take eradications forward.
Legislation is now a reason for crayfish control
To date, legislation has probably been one of the
greatest constraints. Many countries lack effective
legislation to carry out pro-active management/control
of invasive non-native crayfish species. Legislation
controlling import and trade of crayfish species, as well
as introduction to the wild has, on the other hand, existed
in several European countries (Edsman, 2008; Holdich &
Sibley, 2009; Stentiford, et al., 2010), although there is no
international regulatory framework for the trade of live
animals (Chucholl, 2013). At least in principle, legislation
has prevented introduction, and controlled exploitation to
reduce risk of spread. Legislation to allow action to control
spread or attempt to eradicate once invasive crayfish
have been illegally introduced has been missing in many
countries. That has all changed with the introduction of
the EU invasive alien species regulations, which provide
member states with mechanisms to issue Species Control
Orders, and the powers behind them to take direct action
to eradicate high risk invasive species. We have yet to see
how this regulation will be enforced.
the above methods have been trialled to attempt eradication
of invasive alien crayfish, but none have achieved more
than population reduction (Peay, 2001). Eradication has
not been feasible using conventional methods and longterm control is not financially or operationally sustainable,
because of the costs associated and work load necessary.
As in Norway, that is now changing, and the expertise,
tools and techniques we have developed for application
of rotenone-based pesticides are directly transferable to
application of biocides (synthetic or natural pyrethrins)
for crayfish management. These methods have been
trialled and found to be very effective if applied correctly
(Sandodden & Bardal, 2010; Sandodden & Johnsen, 2010).
Total eradication of invasive alien crayfish in Europe
is no longer feasible, but emphasis should be placed on
sustaining viable island populations of native crayfish and
creating new ones. Eradication programmes should be
made an option throughout Europe during identification
and establishment of suitable island populations and areas.
Knowledge and experience to carry out successful crayfish
eradications exists. The new EU regulation 1143/2014 is
a new tool for securing necessary local legislation and
funding.
CONCLUSION
There seems to be an increase in governmental
willingness in Norway to conduct chemical eradications
when projects are feasible and have acceptable short term
environmental impacts. The opportunities for successful
eradications should be weighed against not only the
environmental impact but also the size and complexity of
the waters holding the introduced species. Both legislative
and funding constraints seem less prominent as successful
eradications have been confirmed. Time will show if this
trend will spread throughout Europe.
Ability to eradicate
ACKNOWLEDGEMENTS
Where the threats have been recognised, there seems to
have been willingness to take action within the regulatory
agencies and conservation bodies on the ground. However,
that has been hampered by the lack of legislative powers,
scientific evidence base and funding. This has been
combined with a lack of public will (and therefore pressure
on government) to take action. In most cases, the impacts
from invasive crayfish species were not seen or understood
by the general public.
I would like to thank the Norwegian Environment
Agency for funding the eradications and the Norwegian
Food and Safety Authority, the County Governors in
Telemark and Oslo & Akershus for cooperation before,
during and after the eradications.
Lack of knowledge?
Burridge, L.E. and Haya, K. (1997). ‘The lethality of pyrethrins to larvae
and post-larvae of the American lobster (Homarus americanus)’.
Ecotoxicology and Environmental Safety 38(2): 150–154.
This has probably been an important reason for not
performing eradications. Lack of knowledge has been in
three main areas: 1. a clear understanding of the species
biology, impacts and risks; 2. leading on from this, an
understanding of recognised processes such as risk
assessments, risk management assessments, invasive
species action plans etc. to build a robust case for action;
and 3. a lack of knowledge regarding effective tools and
techniques to translate that into action. Most of these
areas we have now largely addressed or are working to
do so. When a bigger experience base has been built more
countries probably will try to address the legislative and
funding issues necessary to reduce the detrimental effects
of invasive non-native crayfish.
Lack of experience?
Historically this has been a factor. Biocide based
programmes have only fairly recently become an
alternative. Conventional means to manage aquatic
invasive fish and crustaceans have been netting, trapping,
electrofishing, draining waterways and liming etc. All of
408
REFERENCES
Alderman, D.J., Holdich, D. and Reeve, I. (1990) ‘Signal crayfish as
vectors in crayfish plague in Britain’. Aquaculture 86: 3–6.
Chucholl, C. (2013). ‘Invaders for sale: trade and determinants of
introduction of ornamental freshwater crayfish’. Biological Invasions
15: 125–141.
Edsman, L. (2008). ‘The Swedish story about import of live crayfish’.
Bulletin Francais Pêche Pesticide 372–373: 281–288.
Freeman, M.A., Turnbull, J.F., Yeomans, W.E. and Bean, C.W. (2010).
‘Prospects for management strategies of invasive crayfish populations
with an emphasis on biological control’. Aquatic Conservation Marine
Freshwater Ecosystem 20: 211–223.
Gherardi, F. (2010). ‘Invasive crayfish and freshwater fishes of the world’.
Revue Scientifique et Technique International Office of Epizootics
29(2): 241–254.
Gherardi, F., Aquiloni, L., Diéguez-Uribeondo, J. and Tricarico, E.
(2011). ‘Managing invasive crayfish: Is there a hope?’ Aquatic Sciences.
73: 185–200.
Guan, R. Z. (1994). ‘Burrowing behaviour of the signal crayfish,
Pacifastacus leniusculus (Dana), in the River Great Ouse, England’.
Freshwater Forum 4: 155–168.
Sandodden: Eradication of invasive crayfish
Gydemo, R. (1995). Effect on an Insecticide Induced Crayfish Kill. Report
to the Swedish Environmental Protection Agency and the Swedish
Board of Fisheries. Stockholm University, Sweden: Department of
Systems Ecology.
Hiley, P.D. (2003). ‘The Slow Quiet Invasion of Signal Crayfish
(Pacifastacus leniusculus) in England – Prospects for the White-clawed
Crayfish (Austropotamobius pallipes)’. In: D.M. Holdich and P.J.
Sibley (eds.) Management and Conservation of Crayfish: Proceedings
of a Conference held in Nottingham on 7 November 2002, pp. 127–138.
Bristol: Environment Agency.
Holdich, D.M., Gydemo, R. and Rogers, D.W. (1999) ‘A Review of
Possible Methods for Controlling Nuisance Populations of Alien
Crayfish’. In: F. Gherardi and D.M. Holdich (eds) Crustacean Issues
11, Crayfish in Europe as Alien Species, pp. 245–270. Rotterdam: A.A.
Balkema.
Holdich, D.M. and Sibley, P.J. (2009). ‘ICS and NICS in Britain in the
2000s’. In: J. Brickland, D.M. Holdich and E.M. Imhoff (eds.) Crayfish
Conservation in the British Isles. Proceedings of a Conference Held on
25th March 2009 in Leeds, UK, pp. 13–33. Leeds, UK.
Stebbing, P., Longshaw, M. and Scott, A. (2014). ‘Review of methods for
the management of non-indigenous crayfish, with particular reference to
Great Britain’. Ethology Ecology and Evolution 26: 204–231.
Stentiford, G.D., Oidtmann, B., Scott, A. and Peelerb, E.J. (2010).
‘Crustacean diseases in European legislation: Implications for importing
and exporting nations’. Aquaculture 306: 27–34.
Unestam, T. (1972). ‘On the host range and origin of the crayfish plague
fungus’. Report of the Institute of Freshwater Research Drottningholm
52: 192–198.
Van Wijngaarden, R.P.A., Brock, T.C.M. and Van den Brink, P.J. (2005).
‘Threshold levels for effects of insecticides in freshwater ecosystems: A
review’. Ecotoxicology 14: 355–380.
Vrålstad, T., Knutsen, A.K., Tengs, T. and Holst-Jensen, A. (2009). ‘A
quantitative TaqMan® MGB realtime polymerase chain reactionbased assay for detection of the causative agent of crayfish plague
Aphanomyces astaci’. Veterinary Microbiology 137: 146–155.
Holdich, D.M., James, J., Jackson C. and Peay, S. (2014). ‘The North
American signal crayfish, with particular reference to its success as an
invasive species in Great Britain’. Ethology Ecology and Evolution 26:
232–262.
Johnsen, S.I., Vrålstad, T. and Sandodden, R. (2010). Prosedyre ved Funn
eller Mistanke om Introduksjon av Signalkreps – Iverksetting av Tiltak
og Eventuell Friskmelding av Lokalitet. Norwegian Institute of Nature
Research Report 572. Lillehammer, Norway: Norwegian Institute of
Nature Research.
Leahey, J.P. (1979). ‘The metabolism and environmental degradation of
the pyrethroid insecticides’. Outlook on Agriculture. 10: 121–133.
Ludke, J.L., Finley, M.T. and Lusk, C. (1971). ‘Toxicity of Mirex
to crayfish, Procambarus blandinggi’. Bulletin of Environmental
Contamination and Toxicology 6(1): 89–96.
Mayer, F.L. and Ellersieck. M.R. (1986). ‘Manual of Acute Toxicity:
Interpretation and Database for 410 Chemicals and 88 Species of
Freshwater Animals’. US Department of Interior Fish and Wildlife
Resources Publication 160. Washington, D.C.
McMahon, B.R. (2002). ‘Physiological Adaptations to Environment’.
In: D.M. Holdich (ed.) Biology of Freshwater Crayfish, pp. 327–376.
Oxford, UK: Blackwell Science Ltd.
Morolli, C., Quaglio, F., Della Rocca, G., Malvisi, J. and Di Salvio, J.
(2006). ‘Evaluation of the toxicity of synthetic pyrethroids to red
swamp crayfish (Procambarus clarkia, Girard 1852) and common carp
(Cyprinus carpio, L 1758)’. Bulletin Francais Pêche Pesticide. 380–
381: 1381–1394.
OIE (2009). Aquatic Animal Health Code Twelfth Edition. Paris, France:
World Organisation for Animal Health.
O`Reilly, S. (2015). ‘Assessing the Toxicity of Biocides on the North
American Signal Crayfish Pacifastacus leniusculus (Dana) to Aid
Eradication’. MSc(R) thesis. Glasgow: University of Glasgow.
Peay, S. (2001). Eradication of Alien Crayfish Populations. Environment
Agency, Bristol: Environment Agency and English Nature, R&D
Technical Report W1-037/TR1.
Peay, S. and Hiley, P.D. (2006). ‘Biocide Trial to Eradicate Signal Crayfish
in the North Esk Catchment’. Scottish Natural Heritage Commissioned
Report No. 0122. Aberdeen, Scotland: Scottish Natural Heritage.
Peay, S., Hiley, P.D., Collen, P. and Martin, I. (2006). ‘Biocide treatment
of ponds in Scotland to eradicate signal crayfish’. Bulletin Français de
la Pêche et de la Pisciculture 380–381: 1363–1379.
Ribbens, J.C.H. and Graham J.L. (2004). ‘Strategy for the Containment
and Possible Eradication of American Signal Crayfish (Pacifastacus
leninusculus) in the River Dee Catchment and Skyre Burn Catchment,
Dumfries and Galloway’. Scottish Natural Heritage Commissioned
Report No. 014. Dumfries, Scotland: Scottish Natural Heritage.
Sandodden, R. and Bardal, H. (2010). ‘Bekjempelse av Signalkreps
(Pasifastacus leniusculus) på Ostøya i Bærum Kommune’.
Veterinærinstituttets rapportserie 1-2010. Oslo: Veterinærinstituttet.
Sandodden, R. and Johnsen, S.I. (2010). ‘Eradication of introduced signal
crayfish Pasifastacus leniusculus using the pharmaceutical BETAMAX
VET®’. Aquatic Invasions 5(1): 75–81.
Souty-Grosset, C., Holdich, D.M., Noël, P.Y., Reynolds, J.D. and Haffner,
P. (eds.) (2006). Atlas of Freshwater Crayfish in Europe. Paris: Museum
National d´Histoire Naturelle.
409
D. Da Re, E. Tordoni, Z. Negrín-Pérez , J. M. Fernàndez-Palacios, J. R. Arévalo, R. Otto, D. Rocchini and G. Bacaro
Da Re, D.; E. Tordoni, Z. Negrín-Pérez , J. M. Fernàndez-Palacios, J. R. Arévalo, R. Otto, D. Rocchini and G. Bacaro. Modelling
invasive plant alien species richness in Tenerife (Canary Islands) using Bayesian Generalised Linear Spatial Models
Modelling invasive plant alien species richness in Tenerife (Canary
Islands) using Bayesian Generalised Linear Spatial Models
D. Da Re1,2,3, E. Tordoni1, Z. Negrín-Pérez 3, J. M. Fernàndez-Palacios3, J. R. Arévalo3, R. Otto3,
D. Rocchini4,5 and G. Bacaro1
Department of Life Sciences, University of Trieste, Via L. Giorgieri 10, 34127 Trieste, Italy. 2Current address: Earth
and Life Institute, Université catholique de Louvain, Croix du Sud 2, 1348 Louvain-la-Neuve, Belgium. <daniele.dare@
uclouvain.be>. 3Departamento de Ecología, Universidad de La Laguna, Avenida Astrofísico Francisco Sánchez s/n,
38206 La Laguna, Tenerife, Spain. 4Center Agriculture Food Environment, University of Trento, Via E. Mach 1, 38010
S. Michele all’Adige (TN), Italy. 5Centre for Integrative Biology, University of Trento, Via Sommarive 14, 38123 Povo
(TN), Italy. 6Research and Innovation Centre, Department of Biodiversity and Molecular Ecology, Fondazione Edmund
Mach, Via E. Mach 1, 38010 S. Michele all’Adige (TN), Italy.
1
Abstract Biological invasions are one of the major threats to biodiversity, especially on islands where the number of
endemic species is the highest despite their small area. In the Canary Islands, the relationships among invasive alien
species (hereafter IAS) and their environmental and anthropogenic determinants have been thoroughly described but
robust provisional models integrating species spatial autocorrelation and patterns of IAS communities are still lacking.
In this study, we developed a Generalised Linear Spatial Model for Invasive Alien Species Richness (IASR) under a
Bayesian framework, using a methodological approach that encompass GIS and geostatistical analysis. In this study, we
hypothesised that the inclusion of spatial autocorrelation can improve model performance thus obtaining more IASRreliable predictions. In addition, this method provides uncertainty maps that prioritize areas where further sampling efforts
are needed. Our model showed that IASR in Tenerife is mainly driven by a combination of anthropogenic and natural
processes, highlighting favourable conditions for IAS from the coastline to about 800 m a.s.l., especially on the windward
humid aspect. Among anthropogenic factors, a clear positive relationship between road kernel density estimation and
IASR was found. Indeed, road density has recently increased especially in low to mid altitudinal zones on the Canary
Islands, strictly associated with urban expansion and it has been widely demonstrated to be one of the main IAS pathways.
Hence, higher road density can be related to increased ‘propagule pressure’ which is, together with source of disturbance,
one of the most important factors explaining richness in alien species invasion success. Our main conclusions highlight
the importance of considering spatial autocorrelation and researchers’ prior knowledge to increase the predictive power
of statistical models. From a practical perspective, these models and their related uncertainty, will serve as important
management tools highlighting those portions of territories that will be more prone to biological invasions and where
monitoring efforts should be directed.
Keywords: biodiversity, biogeography, biological invasions, geostatistics, GIS, kriging
INTRODUCTION
Islands display unique ecological and evolutionary
processes, hosting more than 20% of the terrestrial plant
and vertebrate species in the world, within less than 5%
of the global terrestrial area (Courchamp, et al., 2014).
Endemics on islands are present with a magnitude higher
than on continents (Kier, et al., 2009). In fact, more than
one third of biodiversity hotspots in the world are entirely,
or largely, within islands (Bellard, et al., 2014).
Besides their high diversity, islands host extremely
fragile environments: 50 out of 80 of the documented
plant extinctions in the last 400 years occurred on islands
and more than 2000 endemic island taxa are currently
thought to be on the verge of extinction (Ricketts, et al.,
2005; Whittaker & Fernández-Palacios, 2007; FernándezPalacios, et al., 2015). Nowhere in Europe is this pattern
more conspicuous than in Macaronesia, the biogeographic
region that encompasses the oceanic islands of the Azores,
Madeira, the Canaries and the Cape Verde archipelago
(Whittaker & Fernández-Palacios, 2007; FernándezPalacios, et al., 2015). Macaronesia is widely recognised
as an outstanding biodiversity hotspot worldwide due to its
high rates of endemism in angiosperms and in bryophytes
(40% and 6.5%, respectively, Whittaker & FernándezPalacios, 2007).
Invasive Alien Species (IAS) pose a serious threat
to the conservation of biodiversity and ecosystem
integrity worldwide (DAISIE, 2009; Scalera, et al.,
2012). Island systems, in fact, are extremely susceptible
to biological invasions due to low habitat diversity,
their simplified trophic webs and higher rate of endemic
species (Courchamp, et al., 2003; Millennium Ecosystem
Assessment, 2005; Vilà & Lopez-Darias, 2006; Barni, et
al., 2012; Bacaro, et al., 2015).
Oceanic islands perform as an open-air laboratory
in the field of invasion biology, because of their long
history of large-scale anthropogenic disturbances and the
recent introduction of non-native species (Whittaker &
Fernández-Palacios, 2007; Denslow, et al., 2009), allowing
us to generalise about the outcome of biotic invasions and
to test the consistency of invasive organisms’ behaviours
(Kueffer, et al., 2010).
Several factors may determine the composition
and the abundance of alien floras, including climate,
geology, land use, landscape context, human impact,
competition with natives and natural or anthropogenic
disturbance and residence time (Crawley, 1987; Pyšek, et
al., 2002; Arévalo, et al., 2005). Anthropogenic factors,
such as inhabitants and trade networks, were imputed
as main drivers of plant IAS introduction and spread:
most populated islands should have more opportunities
to import (and export) novel species due to the high rate
of trade and transport with mainland areas (Pyšek, et al.,
2010). Roads are anthropogenic features that can have
greater influence on the distribution of IAS, particularly
increasing the IAS propagule pressure (Lockwood, et al.,
2005) or promoting the spread of generalist species with
short life cycles and high reproductive rates (Parendes &
Jones, 2000; Pauchard & Alaback, 2004; Arèvalo, et al.,
2005; Dietz & Edwards, 2006; Arteaga, et al., 2009). In
the Canary archipelago, as well as worldwide (Pauchard, et
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
410
up to meet the challenge, pp. 410–416. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Da Re, et al.: Modelling invasive plant species richness
al., 2009), elevation and topography are factors driving the
structure and distribution patterns of alien species spread
(Arévalo, et al., 2005; Rejmánek, et al., 2005; Arteaga, et
al., 2009)
Ecologists agree on the need for preventive tools such
as early alert systems, given that control or eradication of
already-established populations is more difficult and costly
(Hobbs & Humphries, 1995; Bax, et al., 2001). Predictive
invasion models, in fact, allow for evaluating the present
and future extent of plant invasions. Furthermore, their
outcomes are useful tools supporting the development of
eradication/control programmes (Wace, 1977; Alpert, et
al., 2000; Rejmánek & Pitcairn, 2002).
Spatial autocorrelation (SAC) is rarely included in
ecological models thus potentially leading to biased
parameter estimates. Furthermore, classic geostatistical
models assume that data are Gaussian distributed, which
may be an unrealistic assumption for count data, such
as species richness. Generalised linear spatial models
(GLSMs) provide a more robust model definition able to
cope with response variables belonging to the exponential
family distribution (Diggle, et al., 1998, 2003; Zhang, 2002;
Christensen & Waagepetersen, 2002; Diggle & Ribeiro,
2007). By definition, the GLSM is a generalised linear
mixed model in which the random effects are derived from
a spatial process. The Bayesian approach allows parameter
estimation by combining information coming from the
observed data (via the likelihood function) as well as
information coming from other prior sources (i.e. previous
studies, subjective judgments) which is formalised through
prior distributions. Therefore, Bayesian GLSMs (BGLSM)
offer a flexible and robust approach for incorporating
spatial correlation and prior knowledge into the modelling
approach. In addition, the possibility of obtaining
uncertainty maps may provide useful information where
data are missing and further sampling efforts should be
addressed. In this study, we hypothesised that the inclusion
of SAC can improve model performance and therefore
more reliable predictions, assuming that a variable selection
process has been adopted. Specifically, we investigated
alien species richness distribution on Tenerife (Canary
Islands) using a multidisciplinary approach encompassing
Geographic Information Systems (GIS), geostatistical
calculation and statistical modelling. The main goals of
this study are: i) to compute an ecologically and spatially
reliable model of ASR spatial pattern in the island ii) to
test if the inclusion of SAC into the modelling framework
improves model performance.
Strong variation in elevation and aspect, which define
local mesoclimatic zones and land use, are primary factors
in structuring both native and alien plant communities
on the Canary Islands (Whittaker & Fernández-Palacios,
2007). On Tenerife, vegetation can be simplified into five
ecosystems based mainly on elevation and orientation
gradients: succulent coastal scrub (0–700 m a.s.l.),
thermophilous forest (200–600 m), laurel forest or
laurisilva (500–1,000 m), Canarian pine forest (800–2,000
m), and summit or high-mountain scrub (> 2,000 m)
(Fernández-Palacios, 1992; del Arco Aguilar, et al. 2006).
Statistical methods
Response variable
The distribution of Invasive Alien Species on the
Canary Islands is available at ATLANTIS (Gobierno de
Canarias, 2016). This database contains the occurrences of
alien species within a grid of 500 m × 500 m square cells
covering the entire archipelago. Species records span from
1970 to 2013. Invasive Alien Species Richness (IASR) on
Tenerife was obtained by aggregating species occurrences
in those ATLANTIS grid cells covering Tenerife Island land
(5,514 cells out of 8,519 selected). Seventy-two species are
present in the dataset (out of 701 alien species reported for
the entire archipelago; Arechavaleta, et al., 2010).
Predictor variables
Three sets of abiotic variables, namely landscape,
anthropogenic and climatic predictors, were derived
in order to take into account all the the potential factors
MATERIALS AND METHODS
Study area
The study was carried out on Tenerife, the largest
(2,033 km2) island of the volcanic Canary archipelago
situated in the subtropics ca. 70 km off the northwest coast
of Africa (27–29° N, 13–18° W; Fig. 1). It is characterised
by steep altitudinal gradient and it has a triangle-based
pyramid shape with a truncated apex at 2,000 m a.s.l. at
Las Cañadas, from which the volcano Teide rises (3,718
m a.s.l.)
The climate on Tenerife is semiarid to humid
Mediterranean type (Arteaga, et al., 2009), with mean
annual temperature reaching 19° C on the windward aspect
and 21 °C on the leeward one. Mesoclimate is affected by
trade winds that create a contrast between the northern
and windward aspect (more humid and cloudy) and the
southern and leeward aspect (more arid and cloudless).
Fig. 1 Canary Islands and position of Tenerife Island within
the Canary archipelago.
411
Island invasives: scaling up to meet the challenge. Ch 2E Other taxa: Plants
influencing alien species richness. Specifications of the
variables chosen are addressed below.
Landscape predictors
The Digital Elevation Model (DEM) was downloaded
from Cartográfica de Canarias S.A. (GRAFCAN, <https://
www.grafcan.es/>). Aspect and slope were derived from
the DEM for each 10 ×10 m pixel using QGIS 2.16.0 with
GRASS 7.0.4 (Quantum GIS Development Team, 2016).
The standard deviation of slope was calculated as an index
of roughness (Grohmann, et al., 2010).
All the predictors were resampled to 500 m of
spatial resolution using the nearest neighbour algorithm,
accordingly to the spatial resolution of the species
abundance grid. The relative abundance of vegetation
classes (del Arco Aguilar, et al., 2010) within each cell
was used to classify each grid cell, while the percentage
of protected area per cell was used as a proxy of landscape
nature conservation.
Anthropogenic predictors
As a proxy of anthropogenic impacts (e.g.
fragmentation, Bacaro, et al., 2011) the Shannon index
based on the relative abundance of land use classes within
each cell was computed using the R package “vegan”
(Oksanen, et al., 2017). We calculated a density proxy for
roads using a Kernel density estimation (Rosenblatt, 1956;
Parzen, 1962) using four regularly distributed classes of
sample points on the road network distant from each other
5, 10, 20 and 50 km. As above, data were downloaded from
Cartográfica de Canarias S.A.
Climate predictors
Climatic data were obtained from Agencia Estatal de
Metereologia (AEMET) spanning from 2005 to 2014.
Since recorded data showed many gaps throughout the
entire time series of every single weather station, we used
only those weather stations having records covering at
least 80% of the full-time series for Precipitation (P) and
60% for Temperatures (T).
For each dataset mean annual (ma), mean seasonal
(Winter: December, January, February (DJF); Spring:
March, April, May (MAM); Summer: June, July, August
(JJA)) were calculated. In order to obtain continuous
representation of the phenomena, the co-kriging spatial
interpolation technique (Myers, 1984) was applied using
elevation, slope and aspect as covariates using “geoR” R
package (Ribeiro & Diggle, 2001).
Data analysis and modelling
Spatial autocorrelation in explanatory variables was
checked by computing Moran’s I, using R package “spdep”
(Bivand & Piras, 2015). In order to avoid multicollinearity, a
forward variable selection with a double-stopping criterion
approach (Blanchet, et al., 2008) was adopted in order
to select the reduced set of predictors using “adespatial”
(Dray, et al., 2017).
This procedure consists of computing the global model
explained by all explanatory variables via a constrained
ordination such as Redundancy Analysis and, if the
resulting model is significant, calculating the adjusted
coefficient of multiple determination (R2adj). Then variables
were added to a null model (including only the intercept)
using a forward procedure: the procedure stops when no
more significant variables were founded (for a given alpha
level) or when the R2adj of the model is greater than the
global model R2adj. This double-stopping criterion should
prevent the selection method from being too liberal and
consequently inflating type I error rates. Once the reduced
set of predictors was obtained, this was further evaluated
via AIC comparisons using an iterative automatic routine
(package “glmulti”, Calcagno & de Mazancourt, 2010).
The set of predictors thus obtained was then used for
computing the BGLSM. The resulting model was used as a
starting point for the BGLSM.
Unfortunately, probably as an effect of the high number
of predictors retained in model selection, we came across
issues in algorithm convergence. For this reason, we
decided to further reduce the number of predictors chosen,
among the reduced list previously obtained, to three which
are known to be important drivers of the alien species
community along the elevation gradient (Arévalo, et al.,
2005, 2010; Barni, et al., 2012, Bacaro, et al., 2015). Thus,
only the roads 10 km kernel density, PMAM and elevation
were included in the final model (Table 1).
To take into account the spatial correlation of count
data, a BGLSM using the Langevin-Hastings Markov
Chain Monte Carlo (MCMC) algorithm was computed
using the “geoRglm” R package (Christensen & Ribeiro,
2002). To complete the Bayesian model formulation of the
geostatistical models, a strong-informative uniform prior
distribution (Rocchini, et al., 2017a) based on the result
of the geostatistical model was specified. Simulations
were run with the following specifications: four chains,
20,000 iterations, burn-in period of 6,000 iterations and
a thinning rate of 100. To ensure a good mixing of the
chains, convergences were assessed both visually and
with Geweke's diagnostic (Geweke, 1992), along with the
autocorrelation within the chains through “coda” package
(Plummer, et al., 2006). The Bayesian framework also
allows uncertainty of the model to be taken into account,
that is the uncertainty of the prediction in the sampling
units (Gelman & Hill, 2006). This statistic is crucial for
correctly interpreting results and avoiding inappropriate
decision-making.
Finally, the linear relationship between Predicted
vs Observed IASR values was evaluated and the R2 was
calculated as a measure of goodness of fit.
All analyses were performed using the R 3.4
environment (R Core Team, 2017).
RESULTS
A total of 72 IAS were present in the dataset, with a
mean of 4.18 species per cell (range: 1–27). The most
common species on the island are Opuntia maxima (3,161
Table 1 Summary statistics of predictors used in the MAM. The variables units are shown in the last column.
Roads 10 km kernel density
PMAM2005-2014
Elevation
412
Mean
1st quantile
3rd quantile
0.06523
25.808
578.9934
0.02160
14.878
240.0000
0.08953
36.546
794.0460
Min
0.00010
1.647
0.5685
Max
Units
0.41928
63.103
2421.3621
mm
m
Da Re, et al.: Modelling invasive plant species richness
occurrences), Ageratina adenophora (2,239 occurrences)
and Ricinus communis (1,615 occurrences). On average,
the northern (and windward) part of the island has higher
values than the southern (Fig. 2), where the biggest cities
are (Santa Cruz de Tenerife and San Cristóbal de La Laguna
on the NNE coast, Puerto de La Cruz on the NW coast).
Furthermore, a decreasing altitudinal trend in IASR was
also observed, with higher values of IASR near the coast
and lower values above 1,500 m a.s.l. SAC in IASR values
were confirmed by the Moran’s Index value (I = 0.873, p
< 0.001).
Table 2 Model output derived from the
maximum likelihood analysis: τ2 is the
nugget, σ2 is the sill, Φ is the range and gives
information about the spatial autocorrelation
of the sampling units.
The inclusion of SAC in the minimum adequate model
resulted in a consistent improvement in general model
performance (ΔAIC 7,366). In the BGLMS, Markov
chains show good mixing and convergence as highlighted
by Geweke’s diagnostics. Positive linear relationships of
IASR were observed with road kernel density estimation
(10 km) and PMAM2005-2014, while elevation showed a negative
trend (Table 2). Suitable areas for IAS appear to be located
in urban areas, especially on the humid leeward aspect of
the island (Fig. 3). Model output summarised in Table 3
and Fig. 4 shows the uncertainty in the predicted IASR
Variables
Intercept
Roads 10 km kernel density
PMAM2005-2014
Elevation
τ2
σ2
Φ
Coefficients
3.4966
18.7518
0.0126
-0.0017
5.7100
10.7700
2.8950
Table 3 Descriptive statistics of model outputs.
3rd
Mean 1st
quantile quantile Min
3.47
1.29
4.44
0.33
Predicted
0.40
0.90
0.14
Uncertainty 0.75
Max
27.39
4.48
Fig. 2 Spatial pattern of alien species richness in Tenerife
island.
Fig. 4 Spatial pattern of uncertainty of the prediction of
invasive alien species richness by BGLSM.
Fig. 3 Spatial pattern of predicted invasive alien species
richness by BGLSM.
Fig. 5 Predicted vs. observed alien species richness. The
solid line represents best prediction line, dashed line the
fitted linear model.
413
Island invasives: scaling up to meet the challenge. Ch 2E Other taxa: Plants
values. Figure 5 shows the predicted versus observed
values scatterplot suggesting a good performance of model
fit.
DISCUSSION
The approach used in the selection of covariates and
the incorporation of the spatial autocorrelation leads us to
build a reliable ecological model to understand the IASR
behaviour on Tenerife Island (Fig. 5). The outcomes of the
model largely agree with most of the results previously
published in the literature, taking into consideration
both natural and anthropogenic processes. However, the
ΔAIC suggests that incorporating SAC into GLM allows
a consistent improvement in general model performance.
Moreover, it allows us to obtain maps of the predictions
that can be easily consulted by local governments. The map
of uncertainty of the prediction provided in the Bayesian
framework represents a powerful tool to highlight those
areas where sampling efforts should be directed, providing
valuable guidance in the decision-making process. On
average, uncertainty in the model was quite low and
evenly dispersed across the island. The areas where the
uncertainty was higher are where human-related land uses
occur, mainly in the arid coastal belt at low elevations
(Fernández-Palacios & Nicolás, 1995; Rocchini, et al.,
2017b).
The Canary Islands, particularly Tenerife Island, are
chiefly characterised by a steep altitudinal gradient causing
potential variations in several abiotic conditions such as
water availability, temperature, precipitation, and solar
radiation even over relatively short distances (Alexander,
et al. 2009). IASR is inversely proportional with elevation
as already observed in Arévalo, et al. (2005), Arteaga, et
al. (2009) and Bacaro, et al. (2015), among others. The
positive relationship between elevation and limiting factors
such as drought, low temperatures and solar radiation were
thoroughly investigated. Accordingly, it has been observed
that at higher elevations, thermic and hydric stresses reduce
the number of successful colonisations of alien species in
different regions of the world (Fernandez-Palacios, 1992;
Alpert, et al., 2000; Godfree, et al., 2004; Pauchard &
Alaback, 2004; Becker, et al., 2005).
In general, mild environmental conditions associated
with reduced drought stress enhance alien establishment
and spread (Whittaker & Heegaard, 2003). These conditions
were found at ca. 800–1,000 m a.s.l. (Arévalo, et al., 2005;
Arteaga, et al., 2009). It has been observed that invasion
success is mainly linked to the biogeographical affinities
and environmental tolerances of the species (Wilson, et al.,
1992; Arévalo, et al., 2005; Daehler, 2005). Accordingly,
we found a positive relationship between PMAM and IASR,
and the BGLSM highlights as suitable the humid areas
below 1000 m a.s.l, especially on the windward aspect,
whereas the model did not predict suitable areas above
1500 m, except where roads are present. These findings
reflect well the same pattern already observed in other
studies performed both in Tenerife and in other oceanic
islands (e.g. Arévalo, et al., 2005, Pauchard, et al., 2009;
Bacaro, et al., 2015). In addition, Daehler (2005) observed
similar patterns on the islands of Hawaii, where the
relative importance of temperate species on the community
composition increased strongly above 1,400 m a.s.l to
detriment of the tropical ones.
Other authors (e.g. Nogués-Bravo, et al., 2008; Marini,
et al., 2013) pointed out that relationships between
IASR and anthropogenic factors are concentrated at low
elevations, consequently increasing the opportunities
414
for the introduction and establishment of propagules.
Accordingly, our results showed a peak of alien species
richness at a relatively low elevation. Species might have
been introduced in the lowlands from different sources
and in several historical periods. The kernel road density
estimation shows a clear positive relationship with IASR.
Bacaro, et al. (2015) reported that alien species were absent
from plots located at higher elevation in plots sampled near
the main Tenerife road network, consistent with previous
observations (Pauchard & Alaback, 2004). Road density
has increased especially in low to mid elevation belts of the
Canary Islands, strictly associated with urban expansion
and, consequently, to the spread of exotic plants. Roads
may facilitate the dispersal of propagules of alien species
via three main mechanisms: 1) as a source of disturbance
that creates new environmental conditions that are suitable
to ruderal and pioneer species; 2) they may facilitate the
dispersal of propagules via air movement associated with
the transit of vehicles; and 3) they may boost the rate of
invasion by reducing competitiveness of native species
that can cause the potential disappearance of even entire
stands (Trombulak & Frissel, 2000; Bacaro, et al., 2015).
In this study, we assessed the incorporation of SAC
into an ecological model built using ecologically reliable
predictors. The incorporation of SAC improved general
model performance and allows for for uncertainty to be
accounted for in the model framework, providing a way
to prioritize areas where more survey is needed along with
further monitoring actions in order to reduce uncertainty.
Mild environmental conditions may be responsible
for quick establishment and dispersal of aliens on islands.
Accordingly, compared with current literature, our results
showed higher alien species richness in mild environmental
conditions and at a relatively low elevation. This can be
also due to the fact that human land use is concentrated at
low elevations, consequently increasing the opportunities
for the introduction and establishment of propagules. To
cope with plant alien species invasion, local governments
have tried different approaches (Foxcroft, et al., 2007)
but the most effective method still remains mechanical or
hand removal (Gobierno de Canarias, 2014). In a global
warming scenario, a modelling approach that takes into
account spatial autocorrelation of data may play an even
more crucial role in alien species monitoring, highlighting
those portions of territories that are more prone to biological
invasions, especially in fragile ecosystems such as in the
Canary Islands.
REFERENCES
Alexander, J.M., Naylor, B., Poll, M., Edwards, P.J. and Dietz, H. (2009).
‘Plant invasions along mountain roads: The altitudinal amplitude of
alien Asteraceae forbs in their native and introduced ranges’. Ecography
32(2): 334–344.
Alpert, P., Bone, E. and Holzapfel, C. (2000). ‘Invasiveness, invasibility
and the role of environmental stress in the spread of non-native plants’.
Perspectives in Plant Ecology, Evolution and Systematics 3(1): 52–66.
Arechavaleta, M., Rodríguez, S., Zurita, N. and García, A. (2010).
Lista de Especies Silvestres de Canarias. Hongos, Plantas y Animales
Terrestres. Santa Cruz de Tenerife, España: Gobierno de Canarias.
Arévalo, J.R., Delgado, J.D., Otto, R., Naranjo, A., Salas, M. and
Fernández-Palacios, J.M. (2005). ‘Distribution of alien vs. native
plant species in roadside communities along an altitudinal gradient in
Tenerife and Gran Canaria (Canary Islands)’. Perspectives in Plant
Ecology, Evolution and Systematics 7(3): 185–202.
Arévalo, J.R., Otto, R., Escudero, C., Fernández-Lugo, S., Arteaga, M.,
Delgado, J.D. and Fernández-Palacios, J. M. (2010). ‘Do anthropogenic
corridors homogenize plant communities at a local scale? A case studied
in Tenerife (Canary Islands)’. Plant Ecology 209(1): 23–35.
Da Re, et al.: Modelling invasive plant species richness
Arteaga, M.A., Delgado, J.D., Otto, R., Fernández-Palacios, J.M. and
Arévalo, J.R. (2009). ‘How do alien plants distribute along roads on
oceanic islands? A case study in Tenerife, Canary Islands.’ Biological
Invasions 11(4): 1071–1086.
Bacaro, G., Santi, E., Rocchini, D., Pezzo, F., Puglisi, L. and Chiarucci,
A. (2011). ‘Geostatistical modelling of regional bird species richness:
Exploring environmental proxies for conservation purpose’. Biodiversity
and Conservation 20(8): 1677–1694.
Bacaro, G., Maccherini, S., Chiarucci, A., Jentsch, A., Rocchini, D.,
Torri, D., Gioria, M., Tordoni, E., Martellos, S., Altobelli, A., Otto, R.,
Escudero, C.G., Fernández-Lugo, S., Fernández-Palacios, J.M. and
Arévalo, J.R. (2015). ‘Distributional patterns of endemic, native and
alien species along a roadside elevation gradient in Tenerife, Canary
Islands’. Community Ecology 16(2): 223–234.
Barni, E., Bacaro, G., Falzoi, S., Spanna, F. and Siniscalco, C. (2012).
‘Establishing climatic constraints shaping the distribution of alien plant
species along the elevation gradient in the Alps’. Plant Ecology 213(5):
757–767.
Bax, N., Carlton, J.T., Mathews‐Amos, A., Haedrich, R.L., Howarth, F.G.,
Purcell, J.E., Rieser, A. and Gray, A. (2001). ‘The control of biological
invasions in the world's oceans’. Conservation Biology 15(5): 1234–
1246.
Becker, T., Dietz, H., Billeter, R., Buschmann, H. and Edwards, P.J.
(2005). ‘Altitudinal distribution of alien plant species in the Swiss
Alps’. Perspectives in Plant Ecology, Evolution and Systematics 7(3):
173–183.
Bellard, C., Leclerc, C. and Courchamp, F. (2014). ‘Impact of sea level
rise on the 10 insular biodiversity hotspots’. Global Ecology and
Biogeography 23(2): 203–212.
Bivand, R. and Piras, G. (2015). ‘Comparing implementations of
estimation methods for spatial econometrics’. Journal of Statistical
Software 63(18): 1–36.
Blanchet, F.G., Legendre, P. and Borcard, D. (2008). ‘Forward selection
of explanatory variables’. Ecology 89(9): 2623–2632.
Calcagno, V. and de Mazancourt, C. (2010). ‘glmulti: An R package
for easy automated model selection with (generalized) linear
models’. Journal of Statistical Software 34(12): 1–29.
Christensen, O.F., and Waagepetersen, R. (2002). ‘Bayesian
prediction of spatial count data using generalized linear mixed
models’. Biometrics 58(2): 280–286.
Christensen, O.F. & Ribeiro Jr., P.J. (2002) ‘geoRglm: A package for
generalised linear spatial models’. R News 2 (2): 26–28.
Courchamp, F., Chapuis, J.L. and Pascal, M. (2003). ‘Mammal invaders
on islands: impact, control and control impact’. Biological Reviews
78(03): 347–383.
Courchamp, F., Hoffmann, B.D., Russell, J.C., Leclerc, C. and Bellard,
C. (2014). ‘Climate change, sea-level rise, and conservation: keeping
island biodiversity afloat’. Trends in Ecology and Evolution 29(3):
127–130.
Crawley, M.J. (1987). ‘What Makes a Community Invasible?’. In: A.J.
Gray, M.J. Crawley, and P.J. Edwards (eds.) Colonization, Succession
and Stability, 26th BES symposium 1987, pp. 429–453. Oxford:
Blackwell Scientific.
Daehler, C.C. (2005). ‘Upper-montane plant invasions in the Hawaiian
Islands: patterns and opportunities’. Perspectives in Plant Ecology,
Evolution and Systematics 7(3): 203–216.
DAISIE. (2009). Handbook of Alien Species in Europe. Berlin: Springer.
del Arco Aguilar, M.J., Pérez-de-Paz, P.L., Acebes, J.R., GonzálezMancebo, J.M., Reyes-Betancort, J.A., Bermejo, J.A., de Armas, S. and
González-González, R. (2006). ‘Bioclimatology and climatophilous
vegetation of Tenerife (Canary Islands)’. Annales Botanici Fennici
43(3): 167–192.
del Arco Aguilar, M.J., González-González, R., Garzón-Machado, V. and
Pizarro-Hernández, B. (2010). ‘Actual and potential natural vegetation
on the Canary Islands and its conservation status’. Biodiversity and
Conservation 19(11): 3089–3140.
Denslow, J.S., Space, J.C. and Thomas, P.A. (2009). ‘Invasive exotic
plants in the tropical Pacific islands: patterns of diversity’. Biotropica
41(2): 162–170.
Dietz, H. and Edwards, P.J. (2006). ‘Recognition that causal processes
change during plant invasion helps explain conflicts in evidence’.
Ecology 87(6): 1359–1367.
Diggle, P.J. and Ribeiro Jr., P.J. (2007) Model-based Geostatistics. New
York: Springer.
Diggle, P.J., Tawn, J.A. and Moyeed, R.A. (1998). ‘Model‐based
geostatistics’. Journal of the Royal Statistical Society: Series C (Applied
Statistics) 47(3): 299–350.
Diggle, P.J., Ribeiro Jr, P.J. and Christensen, O.F. (2003). ‘An Introduction
to Model-based Geostatistics’. In: J. Møller (ed.) Spatial Statistics and
Computational Methods, Lecture Notes in Statistics Vol.173. pp. 43–86.
New York: Springer.
Dray, S., Blanchet, G., Borcard, D., Guenard, G., Jombart, T., Larocque,
G., Legendre, P., Madi, N. and Wagner H.H. (2017). ‘adespatial:
Multivariate Multiscale Spatial Analysis’. R package version 0.0-8.
<https://CRAN.R-project.org/package=adespatial>.
Fernández‐Palacios, J.M. (1992). ‘Climatic responses of plant species
on Tenerife, the Canary Islands’. Journal of Vegetation Science 3(5):
595–603.
Fernández‐Palacios, J. M., and Nicolás, J. P. (1995). Altitudinal pattern
of vegetation variation on Tenerife. Journal of Vegetation Science 6(2),
183–190.
Fernández‐Palacios, J.M., Rijsdijk, K.F., Norder, S.J., Otto, R.,
Nascimento, L., Fernández‐Lugo, S., Tjørve, E. and Whittaker, R. J.
(2015). ‘Towards a glacial‐sensitive model of island biogeography’.
Global Ecology and Biogeography 25(7): 817–830.
Foxcroft, L.C., Hoffmann, J.H., Viljoen, J.J. and Kotze, J.J. (2007).
‘Environmental factors influencing the distribution of Opuntia stricta,
an invasive alien plant in the Kruger National Park, South Africa’. South
African Journal of Botany 73(1): 109–112.
Gelman, A. and Hill, J. (2006). Data Analysis Using Regression and
Multilevel/Hierarchical Models. Cambridge: Cambridge University
Press.
Geweke, J. (1992). ‘Evaluating the Accuracy of Sampling-Based
Approaches to the Calculation of Posterior Moments’. In: J.M.
Bernardo, J.O. Berger, A.P. Dawid and A.F.M. Smith (eds.) Bayesian
Statistics, Vol. 4., pp. 169–193. Oxford: Clarendon Press.
Gobierno de Canarias (2014). ‘Base de Datos de Especies Introducidas en
Canarias’. <http://www.interreg-bionatura.com/especies/>. Accessed
31 March 2016.
Gobierno de Canarias (2016). ‘Banco de Datos de Biodiversidad de
Canarias’. <http//:www.biodiversidadcanarias.es/>. Accessed 31 March
2016.
Godfree, R., Lepschi, B. and Mallinson, D. (2004). ‘Ecological filtering
of exotic plants in an Australian sub‐alpine environment’. Journal of
Vegetation Science 15(2): 227–236.
Grohmann, C.H., Smith, M.J. and Riccomini, C. (2010). ‘Multiscale
analysis of topographic surface roughness in the Midland Valley,
Scotland’. Institute of Electrical and Electronics Engineers Transactions
on Geoscience and Remote Sensing 49(4): 1200–1213.
Hobbs, R.J. and Humphries, S.E. (1995). ‘An integrated approach to the
ecology and management of plant invasions’. Conservation Biology
9(4): 761–770.
Kier, G., Kreft, H., Lee, T.M., Jetz, W., Ibisch, P.L., Nowicki, C. and
Barthlott, W. (2009). A global assessment of endemism and species
richness across island and mainland regions. Proceedings of the
National Academy of Sciences 106(23): 9322–9327.
Kueffer, C., Daehler, C.C., Torres-Santana, C.W., Lavergne, C., Meyer,
J.Y., Otto, R. and Silva, L. (2010). ‘A global comparison of plant
invasions on oceanic islands’. Perspectives in Plant Ecology, Evolution
and Systematics 12(2): 145–161.
Lockwood, J.L., Cassey, P. and Blackburn, T. (2005). ‘The role of
propagule pressure in explaining species invasions’. Trends in Ecology
and Evolution 20(5): 223–228.
Marini, L., Bertolli, A., Bona, E., Federici, G., Martini, F., Prosser, F. and
Bommarco, R. (2013). ‘Beta‐diversity patterns elucidate mechanisms of
alien plant invasion in mountains’. Global Ecology and Biogeography
22(4): 450–460.
Millennium Ecosystem Assessment. (2005). Ecosystems and Human
Well-being: Biodiversity Synthesis. Washington DC: World Resources
Institute.
Myers, D.E. (1984). ‘Co-kriging – New Developments’. In: G. Verly, M.
David, A.G. Journel and A. Marechal (eds.) Geostatistics for Natural
Resources Characterization, pp. 295–305. Netherlands: Springer.
415
Nogués-Bravo, D., Araújo, M.B., Romdal, T. and Rahbek, C. (2008).
‘Scale effects and human impact on the elevational species richness
gradients’. Nature 453(7192): 216–219.
Oksanen, J., Blanchet, F.G., Friendly, M., Kindt, R., Legendre, P.,
McGlinn, Minchin, P.R., O'Hara, R.B., Simpson, G.L., Solymos, P.,
Henry, M., Stevens, H., Szoecs, E. and Wagner, H. (2017). ‘vegan:
Community Ecology Package’. R package version 2.4-3. <https://
CRAN.R-project.org/package=vegan>.
Parendes, L.A. and Jones, J.A. (2000). ‘Role of light availability and
dispersal in exotic plant invasion along roads and streams in the HJ
Andrews Experimental Forest, Oregon’. Conservation Biology 14(1):
64–75.
Parzen, E. (1962). ‘On estimation of a probability density function and
mode’. The Annals of Mathematical Statistics 33(3): 1065–1076.
Pauchard, A. and Alaback, P.B. (2004). ‘Influence of elevation, land
use, and landscape context on patterns of alien plant invasions along
roadsides in protected areas of South‐Central Chile’. Conservation
Biology 18(1): 238–248.
Pauchard, A., Kueffer, C., Dietz, H., Daehler, C.C., Alexander, J.,
Edwards, P.J. and Jakobs, G. (2009). ‘Ain't no mountain high enough:
Plant invasions reaching new elevations’. Frontiers in Ecology and the
Environment 7(9): 479–486.
Plummer, M., Best, N., Cowles, K. and Vines, K. (2006). ‘CODA:
Convergence diagnosis and output analysis for MCMC’. R News 6:
7–11.
Pyšek, P., Jaros̆ı́k, V. and Kuc̆era, T. (2002). ‘Patterns of invasion in
temperate nature reserves’. Biological Conservation 104(1): 13–24.
Pyšek, P., Jarošík, V., Hulme, P. E., Kühn, I., Wild, J., Arianoutsou, M.,
Bacher, S., Chiron, F., Didziulis, V., Essli, F., Genovesi, P., Gherardi, F.,
Hejda, M., Kark, S., Lambdon, P.W., Desprez-Loustaun, M., Nentwig,
W., Pergl, J., Poboljšaj, K., Rabitsch, W., Roques, A., Roy, D.B. Shirley,
S., Solarz, W., Vilà, M. and Winter, M. (2010). ‘Disentangling the role
of environmental and human pressures on biological invasions across
Europe’. Proceedings of the National Academy of Sciences 107(27):
12157–12162.
Quantum GIS Development Team (2016). ‘Quantum GIS Geographic
Information System’. Open Source Geospatial Foundation Project.
< http://qgis.osgeo.org>.
R Core Team (2017). ‘R: A Language and Environment for Statistical
Computing’. R Foundation for Statistical Computing, Vienna, Austria.
< http://www.R-project.org/>.
Rejmánek, M. and Pitcairn, M.J. (2002). ‘When is eradication of exotic
pest plants a realistic goal’. In: C.R. Veitch and M.N. Clout (eds.)
Turning the tide: the eradication of invasive species, pp. 249–253. IUCN
ISSC Invasive Species Specialist Group. IUCN, Gland, Switzerland and
Cambridge, UK.
Rejmánek, M., Richardson, D.M. and Pyšek, P. (2005). ‘Plant invasions
and invasibility of plant communities’. Vegetation Ecology 20: 332–355.
Ribeiro Jr, P.J. and Diggle, P.J. (2001). ‘geoR: A package for geostatistical
analysis’. R News 1(2): 14–18.
Ricketts, T.H., Dinerstein, E., Boucher, T., Brooks, T.M., Butchart, S.H.,
Hoffmann, M., Lamoreux, J.F., Morrison, J., Parr, M., Pilgrim, J.D.,
Rodrigues, A.S.L., Sechrest, W., Wallace, G.E., Berlin, K., Bielby, J.,
Burgess, N.D., Church, D.R., Cox, N., Knox, D., Loucks, C., Luck,
G.W., Master, L. L., Moore, R., Naidoo, R., Ridgely, R., Schatz, G.E.,
Shire, G., Strand, H., Wettengel, W. and Wikramanayake, E. (2005).
‘Pinpointing and preventing imminent extinctions’. Proceedings of the
National Academy of Sciences of the United States of America 102(51):
18497–18501.
Rocchini, D., Garzon-Lopez, C.X., Marcantonio, M., Amici, V.,
Bacaro, G., Bastin, L., Chiarucci, A., Foody, G.M., Hauffe, H.C., He,
K. S., Ricotta, C., Rizzoli, A. and Rosà, R. (2017a). ‘Anticipating
species distributions: Handling sampling effort bias under a Bayesian
framework’. Science of the Total Environment 584: 282–290.
Rocchini, D., Bacaro, G., Chirici, G., Da Re, D., Feilhauer, H., Foody,
G.M., Galluzzi, M., Garzon-Lopez, C.X., Gillespie, T.W., He, K.S.,
Lenoir, J., Marcantonio, M., Nagendra, H., Ricotta, C., Rommel, E.,
Schmidtlein, S., Skidmore, A.K., Van De Kerchove, R., Wegmann,
M., Rugani, B. (2017b). ‘Remotely sensed spatial heterogeneity as
an exploratory tool for taxonomic and functional diversity study’.
Ecological Indicators 85: 983–990.
Rosenblatt, M. (1956). ‘Remarks on some nonparametric estimates of a
density function’. The Annals of Mathematical Statistics 27(3): 832–
837.
416
Scalera, R., Genovesi, P., Essl, F. and Rabitsch, W. (2012). ‘The Impacts
of Invasive Alien Species in Europe’. European Environment Agency
Technical Report 16. Luxembourg: Publications Office of the European
Union.
Trombulak, S.C. and Frissell, C.A. (2000). ‘Review of ecological effects
of roads on terrestrial and aquatic communities’. Conservation Biology
14(1): 18–30.
Vilà, M. and López-Darias, M. (2006). ‘Contrasting biogeography of
endemic and alien terrestrial species in the Canary Islands’. Orsis:
Organismes y Sistemes 21: 91–101.
Wace, N. (1977). ‘Assessment of dispersal of plant species—the carborne flora in Canberra’. Proceedings of the Ecological Society of
Australia 10: 167–186.
Whittaker, R.J. and Heegaard, E. (2003). ‘What is the observed
relationship between species richness and productivity? Comment’.
Ecology 84(12): 3384–3390.
Whittaker, R.J. and Fernández-Palacios, J.M. (2007). Island
Biogeography: Ecology, Evolution, and Conservation. Oxford: Oxford
University Press.
Wilson, J.B., Rapson, G.L., Sykes, M.T., Watkins, A.J., and Williams, P.A.
(1992). Distributions and climatic correlations of some exotic species
along roadsides in South Island, New Zealand. Journal of Biogeography
19(2): 183–193.
Zhang, H. (2002). ‘On estimation and prediction for spatial generalized
linear mixed models’. Biometrics 58(1): 129–136.
M.J.B. Dyer, G. Keppel, D. Watling, M. Tuiwawa, S. Vido and H.J. Boehmer
Dyer, M.J.B.; G. Keppel, D. Watling, M. Tuiwawa, S. Vido and H.J. Boehmer. Using expert knowledge and field
surveys to guide management of an invasive alien palm in a Pacific Island lowland rainforest
Using expert knowledge and field surveys to guide management of an
invasive alien palm in a Pacific Island lowland rainforest
M.J.B. Dyer1, G. Keppel1,2, D. Watling3, M. Tuiwawa4, S. Vido5 and H.J. Boehmer6,7
School of Natural and Built Environments and Barbara Hardy Institute, University of South Australia, Mawson
Lakes Campus, GPO Box 2471, Adelaide, SA 5001, Australia. <dyemj003@mymail.unisa.edu.au>. 2Biodiversity,
Macroecology and Conservation Biogeography Group, Faculty of Forest Sciences and Forest Ecology, University of
Goettingen, Göttingen, Germany. 3Environment Consulting Fiji, Suva, Fiji. 4South Pacific Regional Herbarium, Faculty
of Science, Technology and Environment, University of the South Pacific, Suva, Fiji. 5Forestry Department, Ministry
of Fisheries and Forests, Suva, Fiji. 6School of Geography, Earth Science and Environment, Faculty of Science,
Technology and Environment, University of the South Pacific, Suva, Fiji. 7Institute for Applied Ecological Studies
(IFANOS), Baerenschanzstrasse 73, Nuremberg, Germany.
1
Abstract Invasive alien ornamental plants are a global problem, especially on oceanic islands, and can have severe
impacts on native biodiversity. Pinanga coronata, is an ornamental palm tree that can form mono-dominant stands in its
native habitat and is widely cultivated throughout the tropics. Here we investigate the introduction, spread, impact and
management of this invasive palm in the Fiji Islands, using extensive discussions with local experts and field surveys.
Pinanga coronata was introduced in the 1970s to the Colo-i-Suva area, eastern Viti Levu island, Fiji´s principal island,
and has since become invasive in mahogany plantations and lowland rainforest. It has also been introduced and is
becoming invasive on the western side of that island. However, the distribution of P. coronata remains geographically
limited to the immediate vicinity of introduction sites but it is rapidly spreading. In each location, the species has formed
mono-dominant stands in the understorey and appears to be displacing native plant species, as suggested by a negative
correlation of its abundance with that of native tree ferns. This highlights the need for rapid control of P. coronata in
Fiji. Local experts state management should involve manual removal of seedlings and saplings, killing of adult palms
by injection of herbicide, and education and legislation to prevent the further spread of the species. Based on these
recommendations and field data, management actions to control P. coronata are proposed and steps to develop these into
a management plan are discussed. Given P. coronata threatens native biodiversity in Fiji and has the potential to invade
other rainforest ecosystems in the tropics, proposed management approaches are urgent and relevant for other tropical
countries.
Keywords: biodiversity loss, biological invasion, island biology, mahogany plantation, management plan, monodominant protected area, ornamental
INTRODUCTION
The tropical Pacific islands are highly vulnerable to
the impacts of invasive alien plant species because of the
region’s geographic isolation and evolutionary traits that
leave species vulnerable to competition (Gurevitch &
Padilla, 2004; Denslow, et al., 2009; Caujapé-Castells, et
al., 2010; Woinarski, 2010; Minden, et al., 2010b). The
region has the greatest rate of increase in the number of
invasive alien plant introductions with respect to area in the
world (Van Kleunen, et al., 2015). Furthermore, invasive
species are a major cause of extinction on Pacific Islands
(Tye, 2009). With increasing economic development in
nations of the Pacific, the diversity and impact of invasive
plants in the region is likely to increase (Kueffer, et al.,
2010).
Pinanga coronata, or ivory cane palm, is an ornamental
palm tree that is cultivated and traded throughout the
tropics (Palmpedia, 2017). The palm is native to lowland
rainforests in Java and Sumatra (Kimura & Simbolon,
2002) and has been identified as a potentially invasive
alien plant species on Pacific and other oceanic islands
(Meyer, et al., 2008). Introduced to Fiji in the 1970s for
its ornamental properties, the palm has become invasive
in lowland rainforest and mahogany plantations (Keppel
& Watling, 2011) but the current extent of its distribution
is unknown. The Fiji Islands have a unique and highly
diverse biota that is severely threatened by habitat loss,
exploitation, pollution and invasive species (Myers, et al.,
2000; Keppel, et al., 2014).
Pacific small island developing states (SIDS) have
limited information, funding and trained professionals
for invasive species management and conservation in
general (Tye, 2009; Keppel, et al., 2012). However, local
communities and experts often have extensive knowledge
about their environment (Lefale, 2010; Keppel, 2014;
Keppel, et al., 2015). In Fiji, expert knowledge plays a
crucial role in conservation and protected area management
(Keppel, 2014) and has provided important information
about the conservation status of rare trees (Keppel, et al.,
2015).
Using results from a quantitative field survey and
qualitative expert knowledge, we demonstrate that P.
coronata is threatening biodiversity and displacing native
tree ferns in lowland rainforests and mahogany plantations.
We then combine these two lines of evidence as the basis
for a management framework to address the P. coronata
invasion and its impact on native biodiversity in Fiji.
Acknowledging the current challenges in the Pacific for
invasive species management (Tye, 2009), the framework
incorporates methods that are economically viable,
develops capacity building and promotes awareness for all
stakeholders related to the P. coronata invasion.
METHODS
Site description
The Fiji Islands include over 300 islands and islets in
the western South Pacific (Mueller-Dombois & Fosberg,
1998). Fiji has a tropical climate with a wetter season
from November to April and a drier season from May to
October. Due to the south-east trade winds orographic
rainfall produces higher precipitation in the south-east
of topographically more complex islands (Mataki, et al.,
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 417–423. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
417
Island invasives: scaling up to meet the challenge. Ch 2E Other taxa: Plants
2006). The capital of Fiji, Suva, is in the south-east of the
archipelago’s largest island, Viti Levu, and has an average
annual rainfall of about 3,000 mm and an average surface
temperature of 25.4°C (Mataki, et al., 2006).
The Colo-i-Suva area is approximately 12 km north of
Suva (Fig. 1) and has four protected areas. The Colo-i-Suva
Forest Park has no legal status but comprises the Colo-iSuva Forest Reserve (FR; 370 ha) and the MaranisaqaWainiveiota FR (77 ha). Adjacent are the Savura FR (448
ha) and the Vago FR (24.7 ha). The area is mountainous
and the vegetation communities are fragmented, lowland
rainforest and mahogany plantations amongst agricultural
and urban landscapes. The Savura and Vago FRs are mostly
comprised of lowland native rainforest with minimal
disturbance. These FRs constitute a major catchment for
the Savura Creek, which secures the water supply for the
capital city (Keppel, et al., 2005). Colo-i-Suva Forest Park
is a mahogany plantation that has not been commercially
logged since its establishment in the 1960s (Tuiwawa &
Keppel, 2012). The Forest Park has conservation values
because the mahogany plantations support a rich native
understorey (Tuiwawa & Keppel, 2012), and is also
frequented for recreational activities including local and
international tourism (Malani, 2002).
Study species
Pinanga coronata, is native to western Indonesia, on
Java and Sumatra where it is abundant in the rainforest
understorey and occurs from sea-level to about 1800 m
(Kimura & Simbolon, 2002). The species can be found
on steep hillsides, lowland flats and exposed ridges, and
juvenile palms are found at higher densities on lower
slopes and moist areas (Kimura & Simbolon, 2002).
Tolerating low light conditions, P. coronata forms monodominant clusters in the rainforest understorey, where
it can reproduce sexually and asexually from vegetative
shoots (Kimura & Simbolon, 2002; Kimura & Simbolon,
2003; Witono & Kondo, 2007). Pinanga coronata has rapid
growth rates, reaches fecundity at <1 m in height, and can
then continuously reproduce (Kimura & Simbolon, 2002).
Pinanga coronata has shown signs of becoming
invasive in Hawaii and Tahiti (Daehler & Baker, 2006;
Meyer, et al., 2008), but is not believed to currently be
threatening biodiversity (US Forest Service, 2015). In Fiji,
Pinanga coronata was introduced to the Colo-i-Suva area
on Viti Levu during the 1970s for its ornamental properties
(Keppel & Watling, 2011) and was first recognised as a
potentially invasive species in 1992 (Watling & Chape,
1992). Since its introduction the palm has become dominant
in the understorey of mahogany plantations (Watling,
2005) and invasive in native lowland rainforest, forming
mono-dominant stands (mature palms and saplings) of
several metres in diameter (Keppel & Watling, 2011).
Interviews
Experts were consulted by the lead author through
informal discussions during fieldwork, and in semistructured interviews in the office from July to September
2016, regarding the invasion history of P. coronata.
Discussions were open ended but the key themes were the
introduction history, distribution and dispersal, impact on
native flora and recommended management of P. coronata.
Field survey
A systematic field survey was conducted in the Colo-iSuva Forest Park and Savura FR. The aim was to identify
areas of management priority and to determine if P.
coronata is spreading. Using the Fishnet Tool in ArcMap
10.2.2, a 300 x 300 m grid with 11 columns and 15 rows
was overlaid on the boundaries of Colo-i-Suva Forest Park
and Savura FR. The centre point of each grid cell was
imported into a Garmin Etrex 30® as waypoints. A 5 x 5 m
plot was placed at each waypoint within the boundaries of
the two forest reserves.
The abundance of P. coronata was recorded in ninetytwo 5 x 5 m plots in Colo-i-Suva Forest Park (54 plots)
and Savura FR (38 plots). The abundance of the palm
was determined as the number of mature (stem > 1 m in
height), juvenile (>0.5–<1.0 m) and seedling (< 0.5 m)
palms, calculated by counting their numbers in each plot.
The abundance of tree ferns was estimated by counting
the number of mature (caudex > 1 m), juvenile (0.1–1
m) and tree fern saplings (< 0.1 m) (Ash, 1986; Ash,
1987). Additionally, opportunistic sightings of isolated P.
coronata palms were recorded on a GPS Etrex 30®. Palms
were only considered isolated if they were not near other
P. coronata palms and were not a part of a mono-dominant
stand.
RESULTS
Introduction history
Although there are no official records about the exact
location and year P. coronata was introduced to Fiji, it
was likely first introduced to a quarantine station north of
Fiji’s capital Suva (Fig. 2), for the propagation and trade of
exotic palm trees. Palms were likely sold from this location
to horticulturalists around Fiji. P. coronata is believed to
have spread from the site through the surrounding, now
cleared, mahogany plantation. Although the quarantine
site has been abandoned and is now surrounded by an
agricultural landscape, P. coronata is still present around
the remains of the buildings.
Fig. 1 Study locations on Fiji’s largest island, Viti Levu.
418
The first official record of P. coronata in Fiji is a
specimen in the South Pacific Regional Herbarium
(number DA 18579) collected from the former Emperor
Gold Mine guesthouse at Colo-i-Suva (about 2 km north
of the former quarantine station) by Saula Vodonivalu on
16 February 1975. The habitat was described as a roadside
and the specimen was flowering, with the tallest palm
being approximately seven feet. This specimen originated
from plantings around the mine’s guesthouse, which were
planted for ornamental purposes. This guesthouse was on
the site of what is now an agricultural property that grows
fresh produce for Joe’s Farm supermarkets (Fig. 2). At
Dyer, et al.: Management of an invasive palm in lowland rainforest
some stage P. coronata was also introduced to a residential
property within the interior of Colo-i-Suva Forest Park,
which had a diverse collection of exotic ornamental plant
species. This garden is still private property but the lease
will return to the Ministry of Fisheries and Forests.
Distribution and dispersal
We believe that the distribution of P. coronata is
currently restricted to the Colo-i-Suva area but is spreading
rapidly. Our observations record that the species has now
spread through the Savura and Vago FRs, occupying a total
area of about 1,500 ha, and is most dense in the mahogany
plantations near Joe’s Farm and in the north of Colo-i-Suva
Forest Park (Fig. 2).
P. coronata is cultivated ornamentally in several gardens
in Suva, including the University of the South Pacific,
Laucala Campus. It has escaped from cultivation in and
around the Garden of the Sleeping Giant near Nadi Airport,
on the western side of Viti Levu (Fig.1) and is distributed
as an ornamental by landscapers and horticulturists in Suva
and across Fiji, especially to hotels and tourism resorts. No
estimates of the numbers of palms dispensed is available.
Within Colo-i-Suva Forest Park, we observed P.
coronata to be most dense along streams and watercourses.
We believe that one means of dispersal for the species is
by seeds falling into waterways leading to establishment
downstream. Once established near streams, P. coronata
probably expands its distribution by moving up slopes
bordering water courses.
Birds are believed to disperse P. coronata seeds.
DW found a P. coronata seedling sprouted in his garden
approximately nine kilometres from the introduction
locations and main infestations. In the Colo-i-Suva area
(Colo-i-Suva Forest Park and Savura FR) dense patches
of P. coronata seedlings are commonly found below the
canopy of tall native (especially Gymnostoma vitiense;
Casuarinaceae) and exotic trees (Maesopsis eminii;
Rhamnaceae) used as perching locations by native members
of the Columbidae family, suggesting that fruits are eaten
by birds that forage in the lower canopy and understorey
. The island thrush, (Turdus poliocephalus) and the redvented bulbul (Pycnonotus cafer) are likely dispersers in
mature and open/edge forests, respectively.
Impact on native flora
We found mono-dominant stands of P. coronata in
the understorey around all introduction sites in the Coloi-Suva area. In the north of Colo-i-Suva Forest Park, P.
coronata comprises up to 70% of the understorey and
is outcompeting native understorey plants, especially
tree ferns, and reducing their sapling regeneration
(Mathieu 2015). Similarly, the palm is also abundant and
outcompeting native species in the understorey of lowland
rainforests. Therefore, P. coronata is considered to have
the potential to become dominant and outcompete native
plant species in Fiji’s native lowland rainforests.
Pinanga coronata was present in 54 % of the plots
surveyed in Colo-i-Suva Forest Park and 17 % of plots
in Savura FR. It was mono-dominant in the understorey
of 19 plots (21 % of all plots), 18 of which were in the
north of Colo-i-Suva Forest Park (dominated by mahogany
plantations) and the other was in the north of Savura FR
(consisting of native lowland rainforest). Visual inspection
of the distribution map (Fig. 2), shows the highest density
near putative source locations and several isolated
populations in both forest reserves.
Palm cover in the understorey displayed a strong
negative correlation with all three tree fern classes, tree fern
saplings (ρ ≥ -0.26, p < 1.2 x 10-2), juvenile tree ferns (ρ ≥
-0.38, p < 3.7 x 10-4) and mature tree ferns (ρ ≥ -0.33, p <
1.7 x 10-3). With increasing palm cover in the understorey,
the abundance of tree ferns decreased, especially when
palm cover exceeded 50 % (Fig. 3). Therefore, results from
the plots surveyed reinforce our field observation-based
belief that P. coronata is displacing native species.
Fig. 2 The distribution of palm seedlings in plots and
isolated populations in Colo-i-Suva Forest Park and
Savura Forest Reserve.
Fig. 3 The comparison of tree fern saplings and palm
cover in the understorey. Tree fern abundance on the
y-axis and palm cover on the x-axis. n = number of plots
in the class.
419
Island invasives: scaling up to meet the challenge. Ch 2E Other taxa: Plants
Recommendation to manage Pinanga coronata
We believe that P. coronata must be listed as a pest
species and be controlled as a matter of urgency. Control
could be via manual removal of P. coronata, starting with
seedlings and juvenile palms in isolated populations.
Chemicals may be required to kill adult palms, because
they are difficult to remove and the species can reproduce
vegetatively from the base. The exotic tree Maesopsis
eminii, which is spread throughout the Colo-i-Suva Forest
Park, may need to be controlled concurrently, as this tree
attracts frugivorous birds and P. coronata seedlings are
often abundant at its base.
There also needs to be action to stop further propagation
and reintroductions by horticulturalists. Furthermore,
education for communities, tourism operators and the
Biosecurity Authority of Fiji about the palm and its threats
to native biodiversity is essential to solicit maximum
support from the public and the government.
DISCUSSION
In our study, qualitative and subjective data are
supported and reinforced by quantitative and objective
field data. Combined, these results make a strong case that
P. coronata is continuing to spread through mahogany
plantations and native lowland rainforest in the Colo-iSuva area. Both observations and a negative correlation
between the abundance of palms and native tree ferns
suggest that the introduced palm is displacing native
tree ferns. Hence, both expert opinion and field data
demonstrate the detrimental impact and potential threat of
P. coronata, highlighting the need for swift and effective
management actions.
However, before any management can be effectively
implemented, knowledge of the exact distribution of P.
coronata (Panetta & Lawes, 2005) and consensus among
key stakeholders about the need for urgent management is
needed. Knowing the palm’s distribution in the Colo-i-Suva
area will not only define the target area for management
but also determine the stakeholders that need to be
involved. Support from the most influential stakeholders
will be essential for establishing and implementing a
successful conservation and management plan (Keppel,
et al., 2012; Moon, et al., 2015; Lenz, 2016). All major
stakeholders, especially the Fiji Forestry Department and
Biosecurity Authority of Fiji, need to agree that P. coronata
is a major threat to native biodiversity and an urgent
management priority. Assuming that these pre-requisites
will be attained, we propose a management framework
(Table 2) using a decision and risk analysis based on our
knowledge and available quantitative data (Maguire, 2004;
Stohlgren & Schnase, 2006; Lenz, 2016). We hope that
proposing this framework will hasten the development and
implementation of an effective management plan.
Management framework
Considering the high threat of P. coronata to native
biodiversity and that its distribution is still relatively
restricted, the overarching aim of a management plan
should be to eradicate the species (Keppel & Watling,
2011). Eradication is defined as the total removal of
all individuals, including seeds, and ensuring that
reintroduction will not occur (Myers & Bazely, 2003;
Meyer, 2014). However, there have been very few
successful invasive plant eradications in the Pacific
Islands and these were restricted to species confined
to small geographic areas (Meyer, 2014). Additionally,
eradication is not achievable without containment (Panetta
& Lawes, 2005). Therefore, a feasibility study combining
the known biological information of P. coronata with the
total extent of the invasion (invasion syndrome) will need
to be conducted to determine if eradication is achievable
with the resources available (Panetta, 2015). Due to this
uncertainty about the feasibility of eradication, we focus
our discussion about management on control measures to
reduce the abundance and spread of P. coronata.
Prior to control, stakeholders including the Ministry of
Fisheries and Forests, NatureFiji-MareqetiViti (NFMV)
and the University of the South Pacific (USP) should
develop a management plan. The framework presented
here could serve as a starting point. Coordinated efforts
by multiple parties will be more efficient (Stohlgren &
Schnase, 2006) and have a greater chance of success when
decision makers for protected areas support the strategy
(Foxcroft, et al., 2008). The involvement of NFMV is
important because they have a strong record of effectively
engaging with communities and decision-makers to achieve
positive conservation outcomes (Morrison, et al., 2012). It
is recommended that NFMV be the primary coordinators
because invasive species management facilitated by a nongovernmental organisation (NGO) that promotes education
and stakeholder communication has greater chances of
success (Epanchin-Niell, et al., 2010).
The second stage of management should aim to
investigate the best method to control P. coronata through a
feasibility trial. Currently there is a paucity of information
on best practice for palm control (Meyer, et al., 2008)
and most methods are species-specific (Langeland, et
al., 2011). A pre-control feasibility study is necessary to
determine which method will be the most effective (Meyer,
2014), as we have outlined in Table 1. Physical removal of
palm seedlings is one of the methods to be trialled, as this
has been shown to be successful at reducing seedling and
juvenile abundance (Langeland, et al., 2011).
Table 1 Recommended control measures for P. coronata that should be trialled in a feasibility study based on
literature on invasive palm management (Dovey, et al., 2004, Langeland & Stocker, 1997, Langeland, et
al., 2011) and opinions from the authors. Suggestions from literature = LT and opinion from experts = AU.
*Biocontrol is mentioned because it could be a successful control method if an appropriate control agent is found.
However, biocontrol is not recommended at this stage and should only be considered if all other methods are
ineffective and not feasible.
Management aim
Control
Control
Age target
Seedlings and juveniles
Seedlings and juveniles
Method
Hand pulling and removed from the area AU
Hand pulling and tied to a tree AU
Control
Mature individuals and clumps
Mature individuals and clumps
All fruiting palms
Crown removal and apply herbicide to the stem LT and AU
Inject herbicide into the apical bud LT
Removal of flowers LT
Biocontrol*LT
Reduce seed load
Control
420
Dyer, et al.: Management of an invasive palm in lowland rainforest
Table 2 Summary of the recommended management framework proposed to control P. coronata in the Colo-i-Suva
area, on Viti Levu, Fiji. NFMV=NatureFiji-MareqetiViti.
Aim: Control P. coronata in the Colo-i-Suva area, on Viti Levu, Fiji
Objectives
Actions
STAGE 1: Producing a management plan through stakeholder communication
Create a management plan through stakeholder
engagement
NFMV to formalise the management plan and
education programmes
Facilitate a formal discussion between stakeholders to develop a
management plan.
Develop an education programme for stakeholders and the
community.
Formalise a regular method for communication between
stakeholders.
STAGE 2: Pre-control feasibility study
Conduct a pre-control feasibility study
Trial different control methods (Table 1).
STAGE 3: Control P. coronata in the Colo-i-Suva area
Control P. coronata in the Colo-i-Suva area
Target isolated and juvenile P. coronata populations.
Remove the low-density populations in the centre and south-east
of Colo-i-Suva Forest Park.
Progressively control palms towards the dense populations in the
north of the Colo-i-Suva area
Simultaneously reduce the seed load and foliage area in the dense
populations in the north of the reserves.
STAGE 4: Post-control monitoring and reducing the threat of reinvasion
Post-control monitoring**
Reduce the threat of reestablishment**
Periodically monitor areas where P. coronata has been controlled
and investigate responses in native vegetation.
Plant native tree ferns in areas that are vulnerable to reinvasion
and monitor propagation success.
STAGE 5: Prohibiting the trade of P. coronata in local horticulture
Ensure that P. coronata is not reintroduced into
the natural environment
Ban the trade of P. coronata in the horticulture and tourism
industries. This will require the species to be listed as a pest with
involvement from the Biosecurity Authority of Fiji.
Find a native non-invasive palm that can replace the trade of P.
coronata.
**Stages three and four should be conducted simultaneously. After control efforts have removed isolated palms monitoring should
take place before the dense P. coronata populations are managed.
There are two approaches commonly used with
herbicide applications for managing clonal palms that
should be trialled. The first method is cutting the palm
below the crown and treating the cut stem with herbicide
and the second method is injecting herbicide directly into
the palm’s apical bud (Langeland & Stocker, 1997). In
Indonesia, densities of the invasive palm Arenga obtusifolia
decreased and native rainforest vegetation successfully
regenerated, when the palm was injected with herbicide
(Konstant, 2014; Nardelli, 2016). A combined approach
of applying chemical herbicides and physical removal
could be implemented but has had varied success for palm
species, like P. coronata, that can vegetatively reproduce
(Langeland & Stocker, 1997; Langeland, et al., 2011).
Although not recommended as an initial control
method for P. coronata, biocontrol may be required if
physical efforts fail to control the spread. Dovey, et al.,
(2004) recommends the use of biocontrol in the Pacific
Islands because it is resource efficient and can strengthen
stakeholder partnerships. Biocontrol is typically applied
when the distribution of an invasive plant is too large to
be controlled by physical methods but is expensive and
requires time-consuming host-specificity tests to ensure
native plants from the same family will not be negatively
affected (Meyer, 2014). However, reduced foliage cover
of an invasive tree due to biocontrol has resulted in the
regeneration of understorey species in other Pacific Island
rainforests (Meyer, et al., 2012).
After determining the most effective method, the
third stage of management should attempt to control the
spread and reduce the distribution of P. coronata. We
consider that juvenile plants in isolated populations should
be controlled first. Prioritising low-density populations
will be the most efficient at containing the invasion and
reducing the threat to endemic, rare and threatened plant
species without significantly increasing the cost (Higgins,
et al., 2000). Targeting mature and juvenile individuals in
isolated populations is a recommended strategy for other
invasive plants in the Pacific Islands and likely to be more
successful than removing large stands (Meyer, et al., 2011).
Monitoring (stage 4) should take place as control efforts
(stage 3) of the different P. coronata populations progress.
It is critical that management efforts are long-term and
control sites are periodically monitored to ensure that the
palm does not regenerate from its seed bank, the longevity
of which is currently not known, and to understand
changes in the vegetation community in response to efforts
(Blossey, 1999; Foxcroft, et al., 2008). When invasive
alien plants are removed from Pacific rainforests follow-up
control efforts are often required (Minden, et al., 2010a, b).
Management should aim to stop P. coronata reinvading
controlled areas. In healthy native forest ecosystems,
421
Island invasives: scaling up to meet the challenge. Ch 2E Other taxa: Plants
the succession by native flora will naturally occur, but
management may be required to reduce the likelihood
of invasive plants re-establishing (Awanyo, et al., 2011).
Planting native species is an expensive but effective method
of reducing the risk of reinvasion (Langeland, et al., 2011).
In the Colo-i-Suva region, planting tree ferns could be a
novel and appropriate approach to reduce the likelihood of
P. coronata re-establishing, because tree ferns are native
and abundant in the area, especially on disturbed sites
(Tuiwawa, 1999; Keppel, et al., 2005).
The final stage is to ensure that P. coronata is not
introduced into the environment again. This would
require the species to be listed as a pest plant under
legislation outlined in the Biosecurity Promulgation
2008 act (Biosecurity Authority of Fiji, 2008), ideally
at the beginning of the management process to provide
legal support for any efforts (Lenz, 2016). Adequate
enforcement of the legislation may require training and
improved technical expertise within Biosecurity Authority
of Fiji. Current palm stocks in local nurseries and
ornamental plantings should be identified and controlled to
ensure that P. coronata is not reintroduced (Meyer, et al.,
2008; Lenz, 2016). The latter will be difficult on privately
owned properties. Involving the horticultural industry
is fundamental for success because they are integral in
preventing continuous reintroduction through ongoing
plantings (Meyer, et al., 2008).
CONCLUSION
Our opinions and field data agree on the considerable
threat that Pinanga coronata is posing to native biodiversity.
They also show that the palm is expanding its distribution
and spreading into native rainforest ecosystems. There
is little doubt that it will continue to do so, unless it is
effectively and swiftly managed. Such management
would require a thorough and effective management plan
suitable to the SIDS in the Pacific and developed through
participation by all key stakeholders. Given the evidence
that the palm is threatening biodiversity, we propose a
framework that could serve as a roadmap for developing
and implementing a management plan.
ACKNOWLEDGEMENTS
We would like to thank the following people and
organisations: NatureFiji-MareqetiViti and Environmental
Consulting Fiji for support and design of the management
framework. Robert Aebi from University of South
Australia for logistics, occupational health and safety
advice. Sherri Lodhar, Geon Hanson and Sunil Gopaul for
support and fieldwork assistance. University of the South
Pacific and staff from the Institute of Applied Science
(IAS) for resources. The research was conducted with
permission from and support of the Ministry of Fisheries
and Forestry, Fiji. A New Colombo Plan Scholarship
through the Department of Foreign Affairs and Trade of the
Australia Federal Government provided financial support
for Michael Dyer to live and study in Fiji. Gunnar Keppel
was supported by an Alexander von Humboldt fellowship.
REFERENCES
Ash, J. (1986). ‘Demography and production of Leptopteris wilkesiana
(Osmundaceae), a tropical tree-fern from Fiji’. Australian Journal of
Botany 34: 207–215.
Ash, J. (1987). ‘Demography of Cyathea hornei (Cyatheaceae), a tropical
tree-fern in Fiji’. Australian Journal of Botany 35: 331–342.
Awanyo, L., Attuah, E.M. and McCarron, M. (2011). ‘Rehabilitation of
forest-savannas in Ghana: The impacts of land use, shade, and invasive
species on tree recruitment’. Applied Geography 31: 181–190.
422
Blossey, B. (1999).’Before, during and after: The need for long-term
monitoring in invasive plant species management’. Biological Invasions
1: 301–311.
Caujapé-Castells, J., Tye, A., Crawford, D.J., Santos-Guerra, A., Sakai,
A., Beaver, K., Lobin, W., Vincent Florens, F.B., Moura, M. and Jardim,
R. (2010). ‘Conservation of oceanic island floras: Present and future
global challenges’. Perspectives in Plant Ecology, Evolution and
Systematics 12: 107–129.
Daehler, C.C. and Baker, R.F. (2006). ‘New records of naturalized and
naturalizing plants around Lyon Arboretun, Manoa Valley, O’ahu’.
Bishop Museum Occasional Papers 87: 3–18.
Denslow, J.S., Space, J.C., & Thomas, P.A. (2009). ‘Invasive exotic
plants in the tropical Pacific islands: patterns of diversity’. Biotropica
41(2): 162–170.
Dovey, L., Orapa, W. and Randall, S. (2004). ‘The Need to Build
Biological Control Capacity in the Pacific’. In: J.M. Cullen, D.T.
Briese, D.J. Kriticos, W.M. Lonsdale, L. Morin, and J.K. Scott, (eds.)
Proceedings of the XI International Symposium on Biological Control
of Weeds, pp. 146–153. Canberra, Australia.
Epanchin-Niell, R.S., Hufford, M.B., Aslan, C.E., Sexton, J.P., Port, J.D.
and Waring, T.M. (2010). ‘Controlling invasive species in complex
social landscapes’. Frontiers in Ecology and the Environment 8: 210–
216.
Biosecurity Authority of Fiji. (2008). Biosecurity Promulgation 2008, No.
28 of 2008, Republic of Fiji Islands Government Gazette, Vol. 8, No. 94.
Foxcroft, L.C., Richardson, D.M. and Wilson, J.R. (2008). ‘Ornamental
plants as invasive aliens: Problems and solutions in Kruger National
Park, South Africa’. Environmental Management 41: 32–51.
Gurevitch, J. and Padilla, D.K. (2004). ‘Are invasive species a major
cause of extinctions?’. Trends in Ecology and Evolution 19: 470–474.
Higgins, S.I., Richardson, D.M. and Cowling, R.M. (2000). ‘Using a
dynamic landscape model for planning the management of alien plant
invasions’. Ecological Applications, 10: 1833–1848.
Keppel, G., Navuso, J.C., Naikatini, A., Thomas, N.T., Rounds, I.A.,
Osborne, T.A., Batinamu, N. and Senivasa, E. (2005). ‘Botanical
diversity at Savura, a lowland rain forest site along the PABITRA
Gateway Transect, Viti Levu, Fiji’. Pacific Science 59: 175–191.
Keppel, G. and Watling, D. (2011). ‘Ticking time bombs – current and
potential future impacts of four invasive plant species on the biodiversity
of lowland tropical rainforests in south-east Viti Levu, Fiji’. The South
Pacific Journal of Natural and Applied Sciences 29: 42–45.
Keppel, G., Morrison, C., Watling, D., Tuiwawa, M.V. and Rounds, I.A.
(2012). ‘Conservation in tropical Pacific Island countries: why most
current approaches are failing’. Conservation Letters 5: 256–265.
Keppel, G. (2014). ‘The importance of expert knowledge in conservation
planning – Comment to an article by C.J. Klein, et al.’. Marine Policy
48: 202–203.
Keppel, G., Morrison, C., Meyer, J.-Y. and Boehmer, H.J. (2014).
‘Isolated and vulnerable: The history and future of Pacific island
terrestrial biodiversity’. Pacific Conservation Biology 20: 136–145.
Keppel, G., Naikatini, A., Rounds, I.A., Pressey, R.L. and Thomas, N.T.
(2015). ‘Local and expert knowledge improve conservation assessment
of rare and iconic Fijian tree species’, Pacific Conservation Biology 21:
214–219.
Kimura, M. and Simbolon, H. (2002). ‘Allometry and life history of
a forest understory palm Pinanga coronata (Arecaceae) on Mount
Halimun, West Java’. Ecological Research 17: 323–338.
Kimura, M. and Simbolon, H. (2003). ‘Patterns of leaf and inflorescence
production of Pinanga coronata (Arecaceae) in a tropical montane
rainforest in West Java’. Tropics 12: 179–184.
Konstant, B. (2014). ‘Indonesia: If you cut it, they will come’. The Horn
Spring 12: 4.
Kueffer, C., Daehler, C.C., Torres-Santana, C.W., Lavergne, C., Meyer, J.Y., Otto, R. and Silva, L. (2010). A global comparison of plant invasions
on oceanic islands’. Perspectives in Plant Ecology, Evolution and
Systematics 12: 145–161.
Langeland, K. and Stocker, R. (1997). ‚Control of non-native plants in
natural areas of Florida’. University of Florida. Cooperative Extensions
Services, report no. ST 242, Gainesville, Florida.
Langeland, K., Ferrell, J., Sellers, B., Macdonald, G. and Stocker, R.
(2011). Integrated management of non-native plants in natural areas
of Florida. Institute of Food and Agricultural Sciences, University of
Florida, Gainseville, FL. SP 242: 1–27.
Dyer, et al.: Management of an invasive palm in lowland rainforest
Lefale, P.F. (2010). ‘Ua ‘afa le Aso Stormy weather today: traditional
ecological knowledge of weather and climate. The Samoa
experience’. Climatic Change 100: 317–335.
Tuiwawa, S.H. and Keppel, G. (2012). ‘Species diversity, composition and
the regeneration potential of native plants at the Wainiveiota Mahogany
Plantation, Viti Levu, Fiji Islands’. The South Pacific Journal of Natural
and Applied Sciences 30: 51–57.
Lenz, M. (2016): Risk Assessment of the Invasive Alien Ivory-cane
Palm (Pinanga coronata) in the Islands of the Tropical South Pacific.
The University of the South Pacific, Faculty of Science, Technology
& Environment (FSTE) and University of Goettingen, Department of
Geography, Germany.
Tye, A. (2009). Guidelines for Invasive Species Management in the
Pacific: A Pacific Strategy for Managing Pests, Weeds and Other
Invasive Species. Apia, Samoa: Secretariat of the Pacific Environment
Programme.
Maguire, L.A. (2004). ‘What can decision analysis do for invasive species
management?’. Risk Analysis 24: 859–868.
US Forest Service. (2015). ‘Pinanga coronata’. Pacific Island Ecosystems
at Risk (PIER) <http://www.hear.org/pier/species/pinanga_coronata.
htm> Accessed 30 March 2016.
Malani, M. (2002). ‘Ecotourism in Fiji’. In: T. Hundloe (ed.) Linking
green productivity to ecotourism: Experiences in the Asia-Pacific
Region, 6: 45–55 . Tokyo: Asian Productivity Organization.
Mataki, M., Koshy, K.C. and Lal, M. (2006). ‘Baseline climatology of
Viti Levu (Fiji) and current climatic trends’. Pacific Science 60: 49–68.
Mathieu, J. (2015): ‘Diversity of Understory Vegetation in a Submontane
Tropical Rainforest under Impact of an Invasive Alien Palm Species.
(Bachelor dissertation), Fiji: University of the South Pacific and
Canada: University of Quebec.
Meyer, J-Y., Lavergne, C. and Hodel, D.R. (2008). ‘Time bombs in
gardens: Invasive ornamental palms in tropical islands, with emphasis
on French Polynesia (Pacific Ocean) and the Mascarenes (Indian
Ocean)’. Palms 52: 71–83.
Van Kleunen, M., Dawson, W., Essl, F., Pergl, J., Winter, M., Weber,
E., Kreft, H., Weigelt, P., Kartesz, J. and Nishino, M. (2015). ‚Global
exchange and accumulation of non-native plants’. Nature 525: 100–103.
Watling, D. (2005). Palms of the Fiji Islands, Suva, Fiji: Environmental
Consultants Fiji.
Watling, D. and Chape. S. (1992). Fiji: State of the Environment Report.
Suva: Government of Fiji.
Witono, J.R. and Kondo, K. (2007). ‘Genetic relationships of Pinanga
coronata accessions by RAPD markers’. Chromosome Botany 2: 93–97.
Woinarski, J. (2010). ‘Biodiversity conservation in tropical forest
landscapes of Oceania’. Biological Conservation 143: 2385–2394.
Meyer, J-Y., Loope, L. and Goarant, A. (2011). ‘Strategy to control the
invasive alien tree Miconia calvescens in Pacific Islands: eradication,
containment or something else’. In: C.R. Veitch, M.N. Clout and
D.R. Towns (eds.), Island invasives: eradication and management.
Occassional Paper SSC no. 42. Gland, Switzerland: IUCN and
Auckland, New Zealand: CBB.
Meyer, J-Y., Fourdrigniez, M. and Taputuarai, R. (2012). ‘Restoring
habitat for native and endemic plants through the introduction of a
fungal pathogen to control the alien invasive tree Miconia calvescens in
the island of Tahiti’. BioControl 57: 191–198.
Meyer, J.-Y. (2014). ‘Critical issues and new challenges for research
and management of invasive plants in the Pacific Islands’. Pacific
Conservation Biology 20: 146–164.
Minden, V., Hennenberg, K.J., Porembski, S. and Boehmer, H.J. (2010a).
‘Invasion and management of alien Hedychium gardnerianum (kahili
ginger, Zingiberaceae) alter plant species composition of a montane
rainforest on the island of Hawai’i’. Plant Ecology 206: 321–333.
Minden, V., Jacobi, J.D., Porembski, S. and Boehmer H.J. (2010b):
‘Effects of invasive alien kahili ginger (Hedychium gardnerianum) on
native plant species regeneration in a Hawaiian rainforest’. Applied
Vegetation Science 13 (1): 5–14.
Moon, K., Blackman, D.A. and Brewer, T.D. (2015). ‘Understanding
and integrating knowledge to improve invasive species management’.
Biological Invasions 17: 2675–2689.
Morrison, C., Rounds, I. and Watling, D. (2012). ‘Conservation and
management of the endangered Fiji sago palm, Metroxylon vitiense, in
Fiji’. Environmental Management 49: 929–941.
Mueller-Dombois, D. and Fosberg, R. (1998). Vegetation of the tropical
Pacific Islands. New York: Springer.
Myers, N., Mittermeier, R.A., Mittermeier, C.G., Da Fonseca, G. and
Kent, J. (2000). ‚Biodiversity hotspots for conservation priorities’.
Nature 403: 853–858.
Myers, J.H. and Bazely, D. (2003). Ecology and control of introduced
plants. Cambridge: Cambridge University Press.
Nardelli, F. (2016). ‘Current status and conservation prospects for the
Javan rhinoceros Rhinoceros sondaicus Desmarest’. International Zoo
News 63: 180–202.
Palmpedia. 2017. ‘Pinanga coronata’. Palmpedia – Palm Grower’s
Guide. <http://www.palmpedia.net/wiki/Pinanga_coronata> Accessed
5 November, 2016.
Panetta, F.D. and Lawes R. (2005). ‘Evaluation of weed eradication
programs: the delimitation of extent’. Diversity and Distributions 11:
435–442.
Panetta, F. D. (2015). ‘Weed eradication feasibility: Lessons of the 21st
century’. Weed Research 55: 435–442.
Stohlgren, T.J. and Schnase, J.L. (2006). ‘Risk analysis for biological
hazards: What we need to know about invasive species’. Risk Analysis
26: 163–173.
Tuiwawa, M. (1999). ‘The Flora, Ecology and Conservation of the
Botanical Biodiversity of Waisoi and the Southeasterm Slopes of
the Korobasaga Range in Namosi Province, Fiji’. MSc thesis, Suva:
University of the South Pacific.
423
K. Floyd, K. Passfield, S. Poncet, B. Myer and J. Lee
Floyd, K.; K. Passfield, S. Poncet, B. Myer and J. Lee. Persistence, accuracy and timeliness: finding,
mapping and managing non-native plant species on the island of South Georgia (South Atlantic)
Persistence, accuracy and timeliness: finding, mapping and managing
non-native plant species on the island of South Georgia (South Atlantic)
K. Floyd1,2, K. Passfield1,2, S. Poncet1,2, B. Myer1,2 and J. Lee2
1
Indigena Biosecurity International, PO Box 54, Nelson, New Zealand. <brad@indigena.co.nz>. 2Government of South
Georgia & the South Sandwich Islands, Government House, Stanley, Falkland Islands FIQQ 1ZZ. <env@gov.gs>.
Abstract The South Georgia ecosystem-based habitat restoration project is a major project that began with the
eradication of invasive rats (Rattus norvegicus) and reindeer (Rangifer tarandus), 2011–2017. As part of this restoration
programme a non-native plant management strategy was developed and implemented. With only 8% of the whole South
Georgia landmass suitable for vascular plants (ca. 283 km²) due to permanent ice and bare rock, there have been 25
indigenous vascular plants and 41 non-native plants recorded from earlier surveys. Following removal of grazing pressure
from introduced mammals, surveys were conducted to quantify the current status and distribution of non-native plant
populations and enable a non-native plant control strategy to be developed for the island. Due to the vast scale of the
island, multiple seasons were required to carry out rapid surveys of key indicators such as species, area of plant coverage
in square metres and age class (mature or juvenile). Survey and control data were entered into a spatial database to enable
analysis, allow data-informed management decisions and be used for long-term control-based monitoring of outcomes.
During this series of surveys, 44 naturalised, non-native plant species were identified and mapped. Of these, 34 species
are now being managed at zero density with 56,851 m2 at 184 sites controlled to date; four are managed at specific sites
with 22,443 m2 controlled to date, three require confirmation of species and the remaining three species are widely
established and receive limited control. Spatially quantifying the distribution and control of non-native plants has enabled
the development and implementation of an effective management strategy which contributes to the restoration of South
Georgia’s native biodiversity.
Keywords: Atlantic Ocean, control-based monitoring, habitat restoration, non-native plants, South Atlantic, South
Georgia
INTRODUCTION
South Georgia (3,533 km², 54°21’ S, 36°42’ W)
is located in the South Atlantic Ocean approximately
1,450 km south-east of the Falkland Islands (Fig. 1).
South Georgia is a United Kingdom Overseas Territory
(UKOT) managed by the Government of South Georgia
and the South Sandwich Islands (GSGSSI). The island
is mountainous and glaciated, and it is only the coastal
fringes that are snow free in the summer months and able
to support vegetation. An estimated 8% of the land mass
of South Georgia (i.e. 283 km²) provides suitable habitat
for vascular plants (GSGSSI GIS, 2007) and, in spite of
the sub-Antarctic climate, many non-native species have
naturalised or persisted for many years.
The first non-native plant species recorded on South
Georgia was Poa annua in 1902 (Walton, et al., 1973)
and this may have been introduced with early sealing
expeditions. Increasing disturbance due to the activity of
shore-based whaling operations after their establishment
from 1904 (Burton, 2012) likely contributed to many of the
later introductions. Greene (1964) classified 51 vascular
species for the island with 24 as listed as native and 27 as
non-native or introduced. There are now considered to be
25 native vascular species with the addition to Greene's list
of the hybrid Acaena magellanica × A. tenera (Galbraith,
2011; Burton, 2012)
Osborne, et al. (2009) recorded 24 introduced vascular
plant species during the survey undertaken in 2009 as part
of the Royal Society for the Protection of Birds (RSPB)
South Atlantic Invasive Species Project.
Local management of selected non-native plant species
on South Georgia has been undertaken since 2004 when
efforts to control bittercress (Cardamine glacialis) were
initiated. In 2010 the efforts to control bittercress were
increased and other non-natives were targeted at selected
sites (GSGSSI, 2016).
In 2014, GSGSSI obtained funding from the UK
Government-funded Darwin Plus initiative (www.
darwininitiative.org.uk) for the project ‘Strategic
Management of Invasive Alien Plants on South Georgia’.
This project enabled a more strategic approach to islandwide non-native plant control. As well as on-going control
of low incidence species, comprehensive surveys were
completed and the distribution and range of non-native
plant species on the island were mapped. This paper
outlines the processes that were undertaken to determine
the extents of non-native plant species to support the
development of a non-native plant management strategy
for the island.
METHODS
Desktop review
Fig. 1 Location of South Georgia Island and main whaling
station locations on the island.
The first step to determine the plant species present
was a desktop review of all documents available that had
location information for non-native species recorded on
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
424
up to meet the challenge, pp. 424–429. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Floyd, et al.: Non-native plants on South Georgia
South Georgia. These included published and unpublished
reports, herbarium records, and personal communications
with researchers, staff and visitors. Many of these records
had limited spatial accuracy and were recorded by general
location only, which restricted the ability to map these
records. Osborne, et al. (2009) had good spatial data for the
sites they visited although they had not been able to access
the restricted areas in some of the historic whaling stations.
All the records found in the desktop review were compiled
into a filterable dataset which generated a species list by
locality which was then checked for presence/absence
during any visits to that area. This dataset provided the
initial non-native plant species list and search targets, with
more recent records given a higher search priority. While
not many of these records were very accurate spatially they
gave a good starting point for surveying.
assess distribution, and the flower structures provided
the diagnostic characteristics to differentiate non-native
species from closely related native plants.
Field surveys
Along with surveys, control was undertaken on selected
known sites and all control activities were recorded using
the same key indicators that had been used during survey
data collection (GPS coordinates, coverage in square metres
and age class), with the addition of the type of herbicide,
the application rate used, and the volume of water used.
Using the compiled historical records, surveys were
undertaken to obtain accurate information on the current
abundance and distribution of non-native plant species
in each search area. Priority survey locations were given
to those with recent records, areas of human disturbance
particularly whaling stations, and where reindeer had
been present. All the surveys were conducted on foot with
access to some areas, where required, provided by ship or
small boat.
A rapid GPS survey technique was used to cover the
large areas involved. The resolution for survey points
was based on plant population size whereby the larger
the population, the further apart points were recorded to
reduce the time required for recording. GPS waypoints
were taken at the centre of each non-native plant infestation
with separate waypoints for each species present. Key
indicators for each waypoint were recorded i.e. the plant
species, the area of plant coverage in square metres, and
age class (mature or juvenile not capable of reproduction).
Coverage was estimated from the ground cover of the
infestation, and is the ground area covered by the plant if
forming a monoculture. In the case of scattered plants the
percentage cover is used to estimate the total square meters
of the infestation.
From the spatial information collected non-native
species were classified into 5 classes depending on
population size and distribution (Table 1). Rather than
determine whether a non-native species was likely to be
invasive, the precautionary principle (Williams, 1997)
was adopted and all non-native vascular plants have been
classified as part of the strategy.
The main surveys were conducted in February–
March 2015 and in February–March 2016, when most
non-native plants were in flower; this made it easier to
Accurate geographical coordinates were essential for
relocating infestations; coordinates were recorded using
hand-held global positioning system receivers (Garmin
GPS62 & 64). To manage these data a GPS Exchange Format
file (gpx) import and export capability was developed in
the recording database to facilitate data storage and display.
Data were collected continuously during different control
and survey visits as weather and logistics allowed more
time to check areas more thoroughly. Additional data were
collected where necessary during these visits to improve
spatial knowledge of the infested areas and outlier plants.
Control based monitoring
Management units
In order to manage the site-led control of Class Two
species, South Georgia was divided into 117 management
units. The management units were determined by a twostep process; firstly the island was divided into eight ecogeographic zones, defined primarily by climate, vegetation
and the historic presence of introduced mammals
(Martin, et al., 2009). For the purpose of non-native plant
management, these zones were further divided into smaller
units based on the level of historic human disturbance,
presence of non-native plants, geographical features and
ease of logistical access.
RESULTS
South Georgia’s vegetation is mostly short grassland
or low-growing rush and sedge communities, apart from
the tall stands of coastal tussac (Parodiochloa flabellata).
Many of the non-native plants are also low-growing which
makes detection very difficult; persistent surveying is
required to locate all individuals. Sometimes a number of
visits are needed as many species are not very visible until
flowering, and timing is critical to finding and controlling
these species before seed becomes viable. New infestations
and new non-native species have also continued to be found
which highlights the need for persistent surveys. Due to the
size of the island, multiple seasons were required to survey
the priority areas. Following repeated surveys between
Table 1 Classification of non-native plants on South Georgia and number of non-native plants in each weed strategy class.
Class
One
Class Two –
Site-Led
Class Three –
Site-Led
Research
Historic
Description
Number of species
Priority species; require species-led control at the island-wide level, to control
34
all plants before they reach maturity. All sites with these species have a 'Site
Tag' in the Weeds Database, for management of follow-up visits.
Species of moderate distribution, requiring site-led control. Priority populations
4
are those at high-use visitor sites, and sites with small infestations where control
will reduce further dispersal.
Species which are widespread and abundant, and require management at high3
use visitor sites and at some remote outlier sites where appropriate.
More information required before classification, to confirm status.
3
Historic species, not seen for at least 10 years. A re-sighting promotes the
35
species to Species-Led – Class One.
425
Island invasives: scaling up to meet the challenge. Ch 2E Other taxa: Plants
2014 and 2017, we consider there to be 44 non-native
species present on the island or that have recent records
from the last 10 years, with a further 35 species recorded
historically but no longer present (Appendix 1).
There have now been 4,245 non-native survey locations
recorded to date. Following the survey these non-native
species have been classified depending on population size
and distribution (Table 1).
From the survey results a non-native plant management
strategy (GSGSSI, 2016) and an associated environmental
impact assessment were developed. In line with GSGSSI
requirements, these documents were peer-reviewed
to ensure they met best practice standards. After their
finalisation, more widespread control of non-native plant
species was undertaken across the island.
There are 34 Class One species occurring at 184 control
sites; these are managed on a species-led basis by targeting
them across their entire known range on South Georgia.
Each of the Class One species is managed at zero density
whereby all plants are controlled where found.
Fig. 2 shows the small increase in new Class One sites
found and treated each season, along with the proportion
that were active (some plants found) and not active (no
plants seen at that site that season).
Control-based monitoring data show that 49,202 m
of Class One species have now been controlled on South
Georgia with 850 m2 of follow-up required in 2016/17
(Fig. 3). The majority of the treated species controlled in
2015/16 were Rumex acetosella since this was the most
widespread Class One species and control was undertaken
only once the full extent of the infested area of this species
was known after surveys that season.
2
There are currently four Class Two species with 221
control locations, these records total 44,903 m2 of plants
treated over the seasons shown in Fig. 3.
DISCUSSION
Spatially quantifying the distribution and control
of non-native plants has enabled the development and
implementation of the ‘South Georgia Non-Native Plant
Management Strategy 2016-2020’ (GSGSSI, 2016) which
contributes to the restoration of South Georgia’s native
biodiversity.
Control of Class Two species is prioritised according to
the potential dispersal risk posed by small populations and
the threat they present to surrounding areas. Spatial data
from the surveys overlaid with the units was essential in
Fig. 2 Number of Class One non-native plant sites on
South Georgia 2014–2017.
426
presenting this information to enable decision-making for
the strategic management of the surveyed species.
There were 44 non-native plant species detected during
the surveys and of these, 34 are currently being controlled
using a range of methods with the aim being to eradicate
them from South Georgia. Many remote areas have not
been able to be visited yet, and although the risk of non–
native infestations at these sites is considered to be low,
based on historic records, all the vegetated areas of the
island will eventually need to be surveyed. This may take
many years due to the logistical difficulties of accessing the
island’s remote areas.
To ensure success of the non-native plant strategy,
persistence is required in treating all target plants until their
seed bank is fully diminished. Control-based monitoring
will assist in determining success by utilising the data
recorded on plant coverage, age class, herbicide rates and
volumes in order to measure progress season by season.
We are confident that most of the non-native species
and infestations have now been located. However, due to
the large size of the areas to be searched, new records are
not unexpected and the weed strategy has been designed to
be adaptive based on the data available.
While all high priority areas for non-native species have
now been surveyed, continued checks will be required to
ensure all infestations are located around the island. Also,
as vegetation communities are likely to recover from
grazing following the reindeer eradication, further searches
for non-native plants will be required across the estimated
4,500 ha of vegetated landscape on the Barff Peninsula and
in the Stromness Bay area (3,250 ha) where the reindeer
were present.
Monitoring new incursions and unknown infestations
will be ongoing and this persistence can be achieved only
if there is a long-term commitment to providing necessary
resources, as is currently the case with the present control
programme funded by the Government of South Georgia
and the South Sandwich Islands.
Timeliness is also vital for ensuring that populations are
successfully controlled. All control operations and surveys
must take place during the optimum time for locating and
treating non-native targets. For South Georgia, this is
between December and February.
Accuracy is also essential, all target species need to be
spatially documented using GPS waypoint data to aid in relocating plants. While there are some small differences in
estimating plant coverage by observers, regular comparisons
between people improve accuracy and consistency of
Fig. 3 Area of Class One and Class Two non-native plant
sites controlled on South Georgia over the last three
seasons (2014/15–2016/17).
Floyd, et al.: Non-native plants on South Georgia
measuring. Having data in quantifiable measures allows
changes in the size and number of infested areas to be
monitored as control efforts are undertaken. Controlbased monitoring provides quantitative information for
managing the target species and enables the comparison of
control and survey data. This information will assist with
further refinement of the management strategy and enable
data driven decision making.
Finally, as with all eradication projects, strong
biosecurity to prevent new introductions to South Georgia
and the movement of already established non-native plant
species between areas is essential. In South Georgia,
there is a wide range of biosecurity measures in place
from cargo packing facilities in the UK and mandated
equipment cleaning before every landing to a bespoke
biosecurity facility on the island itself. Ongoing education
and awareness raising is key to ensure that all visitors to
the island are aware of their biosecurity obligations and the
vital role it plays in protecting native biodiversity.
ACKNOWLEDGEMENTS
We would like to thank South Georgia based
Government Officers, particularly Keiron Fraser and Pat
Lurcock, for their support and Stanley based GSGSSI
staff, in particular Richard McKee. Thanks also to Sarah
Browning for assistance in the field, to FPV Pharos SG for
ship support and Sanfords New Zealand for transporting
equipment. The contributions of British Antarctic Survey
staff at South Georgia and Cambridge, in particular Helen
Peat, and those of Rebecca Upson and Colin Clubbe of
Royal Botanic Gardens, Kew and Eilidh Young of the
Darwin Initiative Fund are gratefully acknowledged, as is
the logistical support provided by HMS Clyde and IAATO
member vessels who kindly assisted in getting project
members to and from South Georgia.
REFERENCES
Burton, R. (2012). South Georgia. Government of South Georgia and the
South Sandwich Islands.
Galbraith, D. (2011). A Field Guide to the Flora of South Georgia. Great
Britain: South Georgia Heritage Trust.
Greene, S.W. (1964). The Vascular Flora of South Georgia. British
Antarctic Survey Scientific Report No. 45.
GSGSSI GIS (2007). GIS Vegetation Layer. NDVI data composed of
NDVI calculations made on Landsat 8 Images. Government of South
Georgia and the South Sandwich Islands.
GSGSSI (2016). South Georgia Non-Native Plant Management Strategy
2016–2020. Government of South Georgia and the South Sandwich
Islands.
Martin, A.R., Poncet, S., Barbraud, C., Foster, E., Fretwell, P. and Rothery
P. (2009). ‘The white-chinned petrel (Procellaria aequinoctialis) on
South Georgia: population size, distribution and global significance’.
Polar Biology 32: 655.
Osborne, J., Borosova, R., Briggs, M. and Cable, S. (2009). Introduced
Vascular Plants: Survey for Baseline Information on Introduced
Vascular Plants and Invertebrates: South Georgia. Unpublished report
to South Atlantic Invasive Species Project. The Lodge, Bedfordshire,
UK: RSPB.
Walton, D.W.H. and Smith, R.I.L. (1973). ‘Status of the Alien Flora of
South Georgia.’ British Antarctic Survey Bulletin 36: 79–97.
Williams, P. A. (1997). ‘Ecology and Management of Invasive Weeds’.
Conservation Sciences 7. Wellington, New Zealand: Department of
Conservation.
427
Island invasives: scaling up to meet the challenge. Ch 2E Other taxa: Plants
Appendix 1 Naturalised non-native vascular plants on South Georgia species list, 2017.
428
Latin Name
Common Name
Family
Achillea millefolium
Achillea ptarmica
Agrostis vinealis
Allium schoenoprasum
Anthoxanthum odoratum
Anthriscus sylvestris
Capsella bursa-pastoris
Cardamine glacialis
Carex aquatilis
Carex nigra
Carex sp.
Carex vallis-pulchrae
Dactylis glomerata
Deschampsia cespitosa
Deschampsia flexuosa
Elytrigia repens
Empetrum rubrum
Festuca rubra
Juncus filiformis
Leptinella scariosa
Lobelia pratiana
Luzula multiflora var congesta
Nardus stricta
Ranunculus acris
Ranunculus repens
Rumex acetosella
Rumex crispus
Sagina procumbens
Scorzonerioides autumnalis
Stellaria media
Trifolium repens
Tripleurospermum inodorum
Vaccinium vitis-idaea
Veronica serpyllifolia
Agrostis capillaris
Deschampsia parvula
Poa pratensis
Trisetum spicatum
Cerastium fontanum
Poa annua
Taraxacum officinale
Agrostis? unknown
Galium saxatile
Holcus lanatus
Aegilops sp.
Alchemilla monticola
yarrow
sneezewort
brown bent
chives
sweet vernal grass
cow parsley
shepherd's purse
bittercress
water sedge
common sedge
sedge unknown (not flowering)
marsh sedge
cocksfoot
tufted hair-grass
wavy hair-grass
couch grass
diddle dee
red fescue
thread rush
feathery buttonweed
berry lobelia
heath wood-rush
mat grass
meadow buttercup
creeping buttercup
sheep’s sorrel
curled dock
pearlwort (procumbent)
autumn hawkbit
common chickweed
white clover
scentless mayweed
cowberry
thyme-leaved speedwell
common bent
punk grass
smooth meadow grass
spike trisetum
common mouse-ear
annual meadow grass
dandelion
unknown grass - TBC
heath bedstraw
Yorkshire fog
goat grass
velvet lady’s mantle
Asteraceae
Asteraceae
Poaceae
Amaryllidaceae
Poaceae
Apiaceae
Brassicaceae
Brassicaceae
Cyperaceae
Cyperaceae
Cyperaceae
Cyperaceae
Poaceae
Poaceae
Poaceae
Poaceae
Ericaceae
Poaceae
Juncaceae
Asteraceae
Campanulaceae
Juncaceae
Poaceae
Ranunculaceae
Ranunculaceae
Polygonaceae
Polygonaceae
Caryophyllaceae
Asteraceae
Caryophyllaceae
Fabaceae
Asteraceae
Ericaceae
Scrophulariaceae
Poaceae
Poaceae
Poaceae
Poaceae
Caryophyllaceae
Poaceae
Asteraceae
Poaceae
Rubiaceae
Poaceae
Poaceae
Rosaceae
Strategy Class
One
One
One
One
One
One
One
One
One
One
One
One
One
One
One
One
One
One
One
One
One
One
One
One
One
One
One
One
One
One
One
One
One
One
Two
Two
Two
Two
Three
Three
Three
Research
Research
Research
Historic
Historic
Floyd, et al.: Non-native plants on South Georgia
Appendix 1 (continued) Naturalised non-native vascular plants on South Georgia species list, 2017.
Latin Name
Common Name
Family
Alopecurus geniculatus
Artemisia sp.
Avena fatua
Brassica cf. napus
Carum carvi
Centella sp.
Cerastium arvense
Daucus carota
Festuca ovina
Hypericum tetrapterum
Lactuca sp.
Lamium purpureum
Lolium multiflorum
Lolium temulentum
Lotus corniculatus
Lupinus sp.
Matricaria discoidea
Phleum pratense
Pisum sativum
Plantago sp.
Poa trivialis
Raphanus sp.
Rorippa islandica
Rumex alpinus
Senecio vulgaris
Sinapis arvensis
Solanum tuberosum
Sonchus sp.
Stellaria graminea
Thlaspi arvense
Trifolium hybridum
Urtica dioica
Urtica urens
marsh foxtail
mugwort
wild-oat
rape
caraway
centella
field mouse-ear
carrot
sheep's fescue
square-stemmed St John's-wort
wild lettuce
red dead-nettle
Italian rye grass
darnel ryegrass
bird's foot trefoil
lupin
pineapple weed
timothy grass
pea
hoary plantain
rough meadow grass
radish
northern yellow-cress
alpine dock
common groundsel
charlock
potato
sow thistle
grass leaf starwort
field penny-cress
alsike clover
common nettle
annual nettle
Poaceae
Asteraceae
Poaceae
Brassicaceae
Apiaceae
Apiaceae
Caryophyllaceae
Apiaceae
Poaceae
Clusiaceae
Asteraceae
Lamiaceae
Poaceae
Poaceae
Fabaceae
Fabaceae
Asteraceae
Poaceae
Fabaceae
Plantaginaceae
Poaceae
Brassicaceae
Brassicaceae
Polygonaceae
Asteraceae
Brassicaceae
Solanaceae
Asteraceae
Caryophyllaceae
Brassicaceae
Fabaceae
Urticaceae
Urticaceae
Strategy Class
Historic
Historic
Historic
Historic
Historic
Historic
Historic
Historic
Historic
Historic
Historic
Historic
Historic
Historic
Historic
Historic
Historic
Historic
Historic
Historic
Historic
Historic
Historic
Historic
Historic
Historic
Historic
Historic
Historic
Historic
Historic
Historic
Historic
429
R. Neville, J.Y. Fujikawa and M. Halabisky
Neville, R.; J.Y. Fujikawa and M. Halabisky. Eradication programmes complicated by long-lived seed banks: lessons learnt from 15 years of miconia control on O’ahu Island, Hawai’i
Eradication programmes complicated by long-lived seed banks: lessons
learnt from 15 years of miconia control on O'ahu Island, Hawai'i
R. Neville1, J.Y. Fujikawa1 and M. Halabisky2
O'ahu Invasive Species Committee, Pacific Cooperative Studies Unit, University of Hawai'i at Mānoa, 743 Ulukahiki
St. Kailua, HI 96734, USA. <oiscmgr@hawaii.edu>. 2University of Washington, Remote Sensing and Geospatial
Analysis Laboratory, School of Forest Resources, Seattle, WA 98195, USA.
1
Abstract The invasive tree Miconia calvescens (Melastomataceae) is a priority for control on the Hawaiian Island of
Oʽahu due to its potential to replace native ʽōhiʽa (Metrosideros polymorpha, Myrtaceae) forests and degrade watershed
function if allowed to establish. The Oʽahu Invasive Species Committee (OISC) is attempting to eradicate this species
from the island of Oʽahu. OISC uses a buffer strategy based on estimated seed dispersal distance to determine the area
under surveillance. This strategy has worked well enough to suppress the number of trees reaching reproductive age. The
number of mature trees removed annually is now less than the number initially removed when the programme started in
2001. In 2016, just 12 mature trees were removed from 54.71 km2 surveyed compared to 2002, when 40 mature trees were
removed from 8.26 km2 surveyed, a 96% drop in mature trees per square kilometre surveyed. However, miconia has a
long-lived seed bank and can germinate after 20 years of dormancy in the soil. Funding shortages and gaps in surveys due
to refusal of private property owners to allow access have resulted in some long-range extensions. OISC’s results suggest
that seed bank longevity is an important factor when prioritising invasive species risk and that allocating more resources
at the beginning of a programme to eradicate a species with long-lived seed banks may be a better strategy than starting
small and expanding.
Keywords: invasive species, invasive plants, watershed, outreach, cloud water interception, Miconia calvescens,
Metrosideros polymorpha
INTRODUCTION
The tree miconia (Miconia calvescens –
Melastomataceae) has been recognised as a threat to forests
on Pacific Islands where it has been introduced (Meyer, et
al., 2011; Medeiros, et al., 1997). Native to tropical Central
and South America, it is under control programmes in
French Polynesia, New Caledonia and Hawaiʽi (Meyer,
et al., 2011). In areas where miconia has invaded, it has
formed monospecific stands, shading out all plant species
beneath it (Meyer, 1996). A miconia-dominated forest
would likely not perform watershed services as well as
Hawaiʽi’s multi-layered native forests. Runoff and water
would likely increase and replenishment of the islands’
freshwater aquifer through cloud water interception would
likely decrease. (Nanko, et al., 2013; Takahashi, et al.,
2011). Because of its potential to outcompete native forest
flora and its potential deleterious effects on watershed
function, miconia has been prioritised for eradication on
the Hawaiian Island of Oʽahu. Miconia was introduced to
Oʽahu at the Wahiawā botanical garden in 1961 (Medeiros,
et al., 1997). It was not until the late 1990s that its invasive
potential became known and efforts to control it began
(Medeiros, et. al., 1997). Here we describe the results
of the island-wide eradication programme for miconia
implemented by the Oʽahu Invasive Species Committee
(OISC) since 2002.
et al., 2011). To put those numbers in perspective, the
forested area of the Koʽolau Mountains is approximately
40,469 ha and its highest peak is 960 m (Koʽolau Mountain
Watershed Partnership, 2017). The rainfall and elevation
of the Koʽolau Range are similar to those areas in Tahiti
where miconia has formed monospecific stands and is
therefore vulnerable to the transformative effects of a
miconia invasion (Fig. 1).
Miconia leaves can reach up to one metre in length
(Chimera, et al., 2000) (Fig. 2). These large leaves
reduce light levels so dramatically that understorey and
groundcover vegetation under a miconia canopy are
severely reduced (Meyer 2004; Nanko, et al., 2013).
Rainwater collects on the large leaves and funnels it to
The Koʽolau Range forms the eastern spine of the
island of Oʽahu and is the location of the island’s primary
aquifers supplying water to the urban centre of Honolulu
(Board of Water Supply, 2016). Data from miconia’s
native and invaded ranges shows that this species occurs in
tropical areas with more than 1,500 mm of rainfall (Libeau,
et al., 2017). Oʽahu rainfall data indicates that most of the
Koʽolau Range, including the areas encompassing the
island’s most important aquifers, could support miconia
(Giambelluca, et al., 2013).
Miconia’s potential to replace forest ecosystems with
monospecific stands is evident from its invasion history in
French Polynesia (Meyer, et al., 2011). There, dense stands
occur over 80,000 ha from sea level to 1,400 m (Meyer,
Fig. 1 Miconia calvescens occurs in areas with more than
1,500 mm rainfall annually. The area shown encompasses
almost all of the Ko’olau mountain range on the eastern
side of the island. (Rainfall data from Giambelluca, et al.,
2011)
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
430
up to meet the challenge, pp. 430–434. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Neville, et al.: Miconia control on O’ahu Island, Hawai’i
the bare ground with a velocity high enough to accelerate
erosion when hitting bare ground (Nanko, et al., 2013).
Water recharge of the island’s aquifers may also
be at risk. A study on Hawaiʽi Island found that nativedominated ‘ōhi’a forest intercepted 454 mm more cloud
water than strawberry guava (Psidium cattleyanum –
Myrtaceae) dominated forests due to the differences in bark
structure and tree shape (Takahashi, et al., 2011). Miconia
has smooth bark similar to strawberry guava and would
likely have similar rates of cloud water interception. This
is important as cloud water interception may contribute
up to 32% of total precipitation in Hawaiʽi’s montane wet
forests (Giambelluca, et al., 2011). Based on these studies,
we surmise that a structurally complex, native forest is
likely better at condensing fog and cloud drip and directing
rain into the islands’ aquifer than a forest dominated by
monospecific stands of miconia.
The frequency interval of every three years is necessary
since miconia can mature in as little as four years (Meyer,
1996). Areas within this 800 m ground buffer that are too
steep to survey by ground are surveyed by helicopter every
two years.
Ground crews locate miconia visually during both
ground and air surveys. In addition to their large size,
miconia’s leaves have vibrant purple undersides and this
makes it fairly easy to detect on both types of surveys
CONTROL OF MICONIA ON OʽAHU
Control of miconia in the Hawaiian Islands began in
1991 after scientists and conservationists saw the damage
it was causing in Tahiti (Medeiros, et al., 1997). On
Oʽahu, miconia was planted at three botanical gardens,
at two private residences and a commercially operated
park (Medeiros, et al., 1997). All voluntarily destroyed
their trees when requested by the state Departments of
Agriculture (HDOA) and Land and Natural Resources
(DLNR). Follow-up surveillance and control were
conducted by volunteers and HDOA and DLNR employees
until the Oʽahu Invasive Species Committee was formed as
a project of the University of Hawaiʽi in 2001.
The miconia eradication project strategy is based on
delimitation, defined as conducting enough surveillance
to be sure that we know how far the invasion extends;
containment, defined as containing the population by
removing plants before they can mature; and extirpation,
defined as removing immatures until the seed bank is
exhausted (Panetta & Lawes, 2005; Panetta, 2007). In order
to achieve the benchmarks of delimitation, containment
and extirpation, OISC designates areas within a certain
radius around reproductive trees for ground or helicopter
surveys and conducts outreach to property owners and
outdoor enthusiasts. The search area is currently at 91.39
km2 and encompasses 4,000 different private property lots
for which we must acquire permission to access in order to
survey (Fig. 3).
Fig. 3 The O'ahu Invasive Species Committee search area
for Miconia calvescens.
Ground surveys are conducted for 800 m around every
mature tree and 500 m around every immature tree every
three years. The 500 m or 800 m radius around trees is
called the ground buffer. An analysis of OISC’s miconia
field data shows that 99% of immature trees fall within
350 m of a mature tree (Fujkawa, pers. comm. 2017),
confirming that the size of the search area is large enough.
Fig. 2 Typical leaf size for Miconia calvescens.
Fig. 4 A sapling growing out of a patch of Clidemia hirta on
O'ahu. The large leaves and striking purple undersides
make miconia fairly detectable, although in heavy
vegetation, trees can still be missed.
431
Island invasives: scaling up to meet the challenge. Ch 2E Other taxa: Plants
(Fig. 4). However, on ground surveys trees are sometimes
missed due to dense vegetation that limits visibility. Steep
terrain can also make trees difficult to detect as simply
getting up a vertical slope safely may distract the surveyor
from finding trees. Large trees that have already breached
the canopy are also difficult to spot from the ground.
Surveyors can visually find miconia trees from a
helicopter but the helicopter must fly very low and slow
above the canopy. The large leaves are visible from the
air and the rotor wash from the helicopter often blows
the leaves around so the purple undersides are visible.
Immature trees as tall as four metres have been spotted
on helicopter surveys. One disadvantage of helicopter
surveys is that in areas with a thick canopy, trees growing
beneath may be missed. OISC’s observations are that once
a large tree has broken through the canopy it often matures
very quickly, so areas designated as too steep for ground
surveys within the ground buffer are flown every two years
to compensate for the fact that trees will not be found until
they are older.
Another 800 m from all mature trees is flown by
helicopter to check for outliers. Despite the high cost of
paying for helicopter time, the per-hectare cost is actually
less than ground surveys because so much area can be
covered quickly. Residential areas designated for helicopter
surveys are done by ground or road in order not to disturb
the residents. If a tree is found during an outlier survey,
then an 800 m buffer is drawn around it and it becomes part
of the area that is searched by ground.
Outreach to hikers, hunters and other outdoor
recreationalists has been helpful in receiving reports of
miconia. OISC engages organised groups of hikers and
hunters through presentations, educational materials and
social media with the aim of informing people how to
identify and report miconia. We also present to schools
and set up educational booths at community festivals in
the areas where we are surveying. We believe that outreach
also assists in gaining entry to private land. Our observation
is that property owners who have heard about the invasive
species problem before we call and ask their permission to
survey, are more likely to let us on, although we have not
specifically measured this.
from the nearest mature tree. The 6,900 m and 2,400 m
extensions were found while the crew was surveying for
other plant species.
DISCUSSION
Having the source trees removed from the botanical
gardens and the few private properties, as well as detection
of mature trees by agencies and volunteers before OISC
was even formed in 2001, was a tremendous help to the
eradication project. OISC was able to apply its strategy
around the historical points and get a head start on
delimitation. By 2010, the surveillance and delimitation
phase of the project was complete. OISC did not have
the resources to survey all suitable habitat, however, we
took the steps described below to ensure that all known
populations were mapped. We interviewed fellow natural
resource agencies and hiker groups working in suitable
habitats to ensure there was not a population in areas we
did not have the time to look at. We conducted binocular
surveys outside our survey areas in prime miconia habitat
looking for large patches. We also calculated the distance
of immature trees to the nearest mature tree in 2009. We
found that 99% of trees fell within 400 m of a mature tree
and maximum distance of an immature tree was just short
of 1,600 m (Fujikawa, et al., 2009). This gave us confidence
that by 2010, delimiting was complete.
After 2010, the project moved into the containment
phase, but it has been difficult to achieve containment as
defined by eliminating all mature plants. Although OISC
has been able to achieve a significant decrease in the
number of mature plants per square kilometre surveyed,
we have not been able to completely suppress maturation.
Trees that are missed during one survey cycle are
sometimes missed due to human error—thick vegetation
and steep terrain are two factors that may decrease the
efficacy of surveillance. Although our success rate with
getting property owners to agree to let field crews survey
their property is 95%, there are some property owners who
have been reluctant to agree to surveys. In one case, it took
Table 1 Number of mature, >2m tall trees and km2
surveyed 2002–2016.
RESULTS
OISC hired its first staff in November of 2001 and
surveys started in 2002. The number of square kilometres
surveyed per year has grown as more funding became
available. Since 2002, OISC has been able to achieve a
96% reduction in mature plants from 4.8 mature trees per
km2 surveyed to 0.2 mature trees per km2 surveyed (Table
1). There were 40 mature trees found and removed over
8.26 km2 in 2002 and just 12 found and removed over
54.71 km2 in 2016.
OISC also counts and takes GPS points for trees that
are immature but over two metres tall. Trees over two
metres that are missed will likely be mature the next time
the field crew surveys. Therefore, the number of trees over
two metres should also be at zero in order to achieve and
ensure containment. OISC has achieved an 81% decrease
in trees over two metres from 6.8 to 1.3 trees per km2.
Three significant range extensions have recently
occurred. As stated above, OISC’s data shows that 99%
of immature plants fall within 350 m of a mature plant.
However, in 2015, one immature tree was found 6,900 m
from the nearest mature tree. In 2016, another was found
1,600 m from the nearest mature tree. In 2017, a small
patch of mature and immature plants was found 2,400 m
432
Year
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
No. mature
trees
40
4
7
9
6
6
0
4
1
3
5
2
5
12
12
No. trees >2m
94
21
14
54
27
25
37
89
48
27
83
94
97
123
94
Km2
surveyed
8.26
9.37
9.00
21.00
25.16
29.37
20.53
14.07
23.25
27.00
14.87
22.20
21.59
39.44
54.71
Neville, et al.: Miconia control on O’ahu Island, Hawai’i
several years to acquire access from a property owner who
owned an entire valley. By the time the crew was able to
survey, trees had matured. Sometimes the 5% that say no
or take a long time to say yes can be critical. In some years
decreases in funding meant we did not have the resources to
survey the area required by our strategy. The combination
of funding fluctuations and time spent negotiating property
access allowed some trees to mature. The presence of
mature trees may have resulted in range extensions into
new watersheds from long-distance dispersal events.
A review of the distances between immature and mature
trees conducted in 2017 resulted in 99% of immature
miconia falling within 350 m of a mature tree (Fujikawa
pers. comm. 2017), which was similar to our 2009 results.
However, the furthest immature miconia was now 6,900 m
away from the nearest mature tree.
we believe outreach has helped OISC get access to survey
on private land. People seem more willing to grant access
if they have heard of the invasive species problem and
miconia before they receive the call asking for permission
to survey.
OISC’s outreach is a combination of talking directly to
community groups, hiking and hunting groups and schools
through presentations, participating at events and social
media. We have not had the ability to scientifically test
the outreach and see which methods or messages work
best. Anecdotally, we believe that explaining the larger
ecosystem effects and the possible effects on the island’s
water supply will persuade a wider group of people than
those who might be motivated by preserving the native
flora of Hawaiʽi. Research about which messages would
be received best would be welcomed.
Miconia’s long-lived seed bank is a complicating factor
in achieving containment. Research by Meyer (2010) has
estimated the seed bank at 16 years in French Polynesia,
but observations on Oʽahu suggest it may be as long as 20
years. The Wahiawā botanical garden where miconia was
originally introduced has found seedlings and reported them
to OISC as late as 2016. They removed their mature tree in
1996 (Medeiros, et. al., 1997) and OISC has surveyed the
entire area at least three times without finding any miconia
whatsoever, so the likelihood that the seedlings are from
the 1996 mature tree is very high.
Although completely suppressing maturation of miconia
has proved difficult due to the long-lived seed bank, OISC
has been able to achieve a 96% reduction in mature trees
per square kilometre since the programme began in 2002.
In 2016, only 12 mature trees were found across 54.71 km2.
OISC has been able to keep the density of mature trees
very low, but the long-lived seed bank means that a missed
tree due to human error, an area that the field crew cannot
survey due to lack of funding or the inability to get access
to private property will likely result in a mature tree once
the crew has access to the area. Containment for species
with long-lived seed banks will be long-term projects.
Lessons learnt
For this reason, when evaluating feasibility, prioritising
species for control and planning eradications, seed bank
longevity should always be taken into account (Panetta
and Timmins 2004). Policymakers deciding which species
should be restricted for import should also consider seed
bank longevity as a critical factor. While a long-lived seed
bank is certainly not the only factor that makes a species
invasive, if a species with a long-lived seedbank starts to
become a problem, eradication will be a long-term and
expensive project. Seed-bank longevity is not known for
many species, and conducting longevity studies to answer
that question for species under management would be very
helpful to plan eradication efforts.
The long-term work done on miconia in French
Polynesia is key to OISC’s success in preventing a fullscale miconia invasion on Oʽahu. The research was critical
to raising the alarm about the species and mobilising
control efforts early. Miconia is one of the few species that
OISC has taken on where the seed-bank longevity is known
thanks to the long-term studies done by Jean-Yves Meyer
(Meyer, et al., 2011). Research from Tahiti also formed the
basis of the outreach narrative. For example, one of the key
talking points for outreach was the enormous area of forest
that had been turned into monotypic stands of miconia and
a photograph of a landslide in Tahiti was a mainstay of
state-wide outreach to explain the potential erosion effects.
Both Tahiti and Oʽahu offer lessons to islands where
miconia might be dispersed in the future.
If possible, having adequate funds at the beginning
of an eradication project to complete delimiting as soon
as possible may shorten the containment phase. It took
OISC eight years to be sure where the miconia population
was. Private property was a complicating factor. A small
percentage of larger landowners would not let the field
crew survey in a timely manner and trees were allowed
to mature while we negotiated access. Delimiting and
containing a species before it spreads to additional private
property owners will be immensely helpful in achieving
eradication as quickly as possible. Taking on additional
species can also be helpful in detecting long-distance
dispersal events. OISC volunteered to do aerial surveys for
a forest pathogen because it would require us to fly over
all habitat suitable for miconia. The small patch of mature
and immature trees 2,600 m away from the nearest known
mature tree was discovered during this survey.
Outreach has also been helpful for the survey effort
and programme managers should consider dedicating
funds and employees for outreach from the beginning of
the programme. OISC did not have a full-time outreach
specialist until four years after the programme started.
OISC has had trees reported to us but more importantly,
ACKNOWLEDGEMENTS
The authors would like to thank the Department of
Land and Natural Resources/Division of Forestry and
Wildlife, Hawaiʽi Invasive Species Council, Honolulu
Board of Water Supply, US Forest Service and the Hawaiʽi
Tourism Authority for funding this project; the members of
OISC’s steering committee for guidance and moral support
over the years; OISC’s Principal Investigator Dr. David
Duffy and the staff of the Pacific Cooperative Studies Unit
and OISC volunteers, field crew and staff past and present.
The Oʽahu Invasive Species Committee is a project of
the Pacific Cooperative Studies Unit at the University of
Hawaiʽi at Mānoa.
REFERENCES
Board of Water Supply (2016). Water Master Plan. Honolulu, USA:
Board of Water Supply. <www.boardofwatersupply.com/waterresources/water-master-plan>.
Chimera, C.G., Medeiros, A.C., Loope, L.L. and Hobdy, R.H. (2000).
‘Status of management and control efforts for the invasive alien tree
Miconia calvescens DC. (Melastomataceae) in Hana, East Maui’.
Pacific Cooperative Studies Unit Technical Report #128. <https://
scholarspace.manoa.hawaii.edu/handle/10125/7122>.
433
Island invasives: scaling up to meet the challenge. Ch 2E Other taxa: Plants
Fujikawa, J.Y., Iott, S., Luechauer, Z., Beebe, C., Kanakaokai, G., Conley,
A., Arnott, C. and Neville, R. (2009). ‘Delimiting surveys, control and
eradication of Miconia calvescens using pre-determined buffer areas on
the island of Oʽahu, Hawaiʽi’. Poster presented at the 2009 International
Miconia Conference, Keanae, Maui, Hawaii, May 4–7, 2009.
Giambelluca, T.W., DeLay, J.K., Nullet, M.A., Scholl, M.A. and Gingerich,
S.B. (2011). ‘Canopy water balance of windward and leeward Hawaiian
cloud forests on Haleakalā, Maui, Hawaiʽi’. Hydrological Processes 25:
438–447.
Giambelluca, T.W., Chen, Q., Frazier, A.G., Price, J.P., Chen, Y.-L., Chu,
P.-S., Eischeid, J.K, and Departe, D.M. (2013). ‘Online rainfall atlas of
Hawaiʽi’. Bulletin of the American Meteorological Society 94: 313–316.
Koʽolau Mountain Watershed Partnership (KMWP) (2017). Koʽolau
Mountain Watershed Partnership Action Plan 2016–2017. <http://
koolauwatershed.org/wp-content/uploads/KMWP-2016-2017-ActionPlan.pdf>.
Libeau, M., Pouteau, R., Taputurai, R. and Meyer, J.-Y. (2019). ‘Predicting
the risk of plant invasion on islands: the case of Miconia calvescens in
the Marquesas, French Polynesia (South Pacific)’. In: C.R. Veitch, M.N.
Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) Island invasives:
scaling up to meet the challenge, p. 716. Occasional Paper SSC no. 62.
Gland, Switzerland: IUCN.
Medeiros, A.C., Loope, L.L., Conant, P. and McElvaney, S. (1997).
‘Status, ecology, and management of the invasive plant Miconia
calvescens DC. (Melastomataceae) in the Hawaiian Islands’. Records
of the Hawaii Biological Survey for 1996. Bishop Museum Occasional
Papers 48: 28–36.
Meyer, J.-Y. (1996). ‘Status of Miconia calvescens (Melastomataceae),
a dominant invasive tree in the Society Islands (French Polynesia)’.
Pacific Science 50: 66–76.
Meyer, J.Y. (2004). ‘Threat of invasive alien plants to native flora and
forest vegetation of Eastern Polynesia’. Pacific Science 58: 355–375.
Meyer, J.-Y. (2010). ‘The Miconia Saga: 20 Years of Study and Control
in French Polynesia (1998–2008)’. In: L.L Loope, J.-Y. Meyer, B.D.
Hardesty and C.W. Smith (eds.) Proceedings of the International
Miconia Conference, Keʽanae, Maui, Hawaiʽi, May 4–7, 2009, Maui
Invasive Species Committee and Pacific Cooperative Studies Unit,
University of Hawaiʽi at Mānoa. <http://www.hear.org/conferences/
miconia2009/proceedings/>.
Meyer, J-Y., Loope, L. and Gourant, A-C. (2011). ‘Strategy to control the
invasive alien tree Miconia calvescens in Pacific islands: eradication,
containment or something else?’ In: C.R. Veitch, M.N. Clout and
D.R. Towns (eds.) Island invasives: eradication and management, pp.
91–96. Occassional Paper SSC no. 42. Gland, Switzerland: IUCN and
Auckland, New Zealand: CBB.
Nanko, K., Giambelluca, T.W., Sutherland, R.A., Mudd, R.G., Nullet,
M.A. and Ziegler, A.D. (2013). ‘Erosion potential under Miconia
calvescens stands on the island of Hawaiʽi’. Land Degradation &
Development 26: 218–226.
Panetta, F.D. and Timmins, S.M. (2004). ‘Evaluating the feasibility of
eradication for terrestrial weed incursions’. Plant Protection Quarterly
19: 5-11.
Panetta, F.D. and Lawes, R. (2005). ‘Evaluation of weed eradication
programs: the delimitation of extent’. Diversity and Distributions 11:
45–442.
Panetta, F.D. (2007). ‘Evaluation of weed eradication programs:
containment and extirpation’. Diversity and Distributions 13: 33–41.
Takahashi, M., Giambelluca, T.W., Mudd, R.G., DeLay, J.K., Nullet,
M.A. and Asner, G.P. (2011). ‘Rainfall partitioning and cloud water
interception in native forest and invaded forest in Hawaiʽi Volcanoes
National Park’. Hydrological Processes 25: 448–464.
434
C.J. West and D. Havell
West, C.J. and D. Havell. Weed eradication on Raoul Island, Kermadec Islands, New Zealand: progress and prognosis
Weed eradication on Raoul Island, Kermadec Islands, New Zealand:
progress and prognosis
C.J. West1 and D. Havell2
1
Department of Conservation, P.O. Box 10-420, Wellington 6143, New Zealand. <cwest@doc.govt.nz>.
2
Department of Conservation, Private Bag 68908, Newton, Auckland 1145, New Zealand.
Abstract During the 45 years that the Raoul Island weed eradication programme has been underway, eleven species have
been eradicated. To complete the restoration of Raoul Island’s unique ecosystems supporting significant seabird biodiversity
and endemic biota, nine further transformer weeds must be eradicated. In this review of progress to date, we examine the
feasibility of eradication of these transformers and identify that four species are on target for eradication: African olive
(Olea europaea subsp. cuspidata), yellow guava (Psidium guajava), castor oil plant (Ricinus communis) and grape (Vitis
vinifera). However, for four more species more staff resources are required to achieve eradication as currently infestations
are establishing faster than they are being eliminated: purple guava (Psidium cattleianum), black passionfruit (Passiflora
edulis), Brazilian buttercup (Senna septemtrionalis) and Mysore thorn (Caesalpinia decapetala). The ninth species,
Madeira vine (Anredera cordifolia), is being contained but presents logistical difficulties for effective control – herbicide
resistant tubers and cliff locations requiring rope access in unstable terrain. Increasing the resources for this programme
now to enable eradication of these transformer weeds will reduce the total long-term cost of the programme. Eradication
of rats, the 2006 eruption, recent greater cyclone frequency, increased tourism requiring biosecurity management, and
staffing reductions have all impacted progress on weed eradication. Myrtle rust (Austropuccinia psidii), confirmed in
March 2017 as the latest invasive species on Raoul Island, is establishing on Kermadec pohutukawa (Metrosideros
kermadecensis), the dominant canopy species. The impact of this species on the weed eradication programme is unknown
at this point.
Keywords: conservation, dispersal, eradication feasibility, invasive species, rats, seabirds, seedbank, transformer
INTRODUCTION
Raoul Island, the largest and northernmost of the main
islands in the Kermadec Group (29ᵒ 15' S, 177ᵒ 55' W),
was once home to vast seabird colonies. However, the
impacts of whalers and settlers from 1800 AD through the
introduction of goats (Capra hircus), pigs (Sus scrofa),
cats (Felis catus), and Norway rats (Rattus norvegicus)
extirpated most indigenous seabird species and a number of
indigenous land birds (Veitch, et al., 2004). The goats had
a major impact on the endemic vascular plants too (Parkes,
1984). Many vascular plant species were introduced for
food and animal forage (Sykes et al., 2000). Twenty-five
vascular plant species are endemic to Raoul Island (Sykes,
et al., 2000), and most of these make up the forest that
clothes the island, dominated by Kermadec pohutukawa
(Metrosideros kermadecensis). The latest invasive species
to arrive on Raoul Island is myrtle rust (Austropuccinia
psidii). This species was first detected in March 2017
and noticed because of canopy die-off of a small area of
mature Kermadec pohutukawa in Denham Bay. Myrtle
rust has the potential to alter the dynamics of many native
and introduced biota on the island by releasing plants
from suppression by the pohutukawa canopy and reduced
flowering and nectar production which will impact some
land birds. Raoul Island is an active volcano, last erupting
in 2006, and it is located in the path of seasonal cyclones
(December to May).
Because of its unique ecosystems, Raoul Island was
declared a Flora and Fauna Reserve (now Nature Reserve)
in 1934. The New Zealand government has funded the
eradication of all introduced feral mammals: goats were
eradicated in 1984 (Sykes & West, 1996), and rats and cats
were eradicated in 2002 and 2004, respectively (Broome,
2009). The eradication of goats greatly assisted recovery of
endemic plant species, rescuing several from the brink of
extinction. As a consequence of the rat and cat eradications
indigenous seabirds and land birds are returning to Raoul
Island (Veitch, et al., 2011), significantly beginning the
recovery of this ecosystem. Several terrestrial birds now
occupy extensive areas of Raoul and are likely to have
significant impacts on ecosystem dynamics.
However, a small suite of transformer weed species
(sensu Pyšek, et al., 2004) currently impedes full
restoration of ecosystem functioning on Raoul Island. The
vascular plant flora of Raoul Island currently comprises
118 indigenous species and 196 introduced species (of
which c.10% are transformer species). A weed eradication
programme has been underway since 1972 (West, 2011)
and, to date, 11 species have been eradicated (Table 1), the
majority of which were transformers (West & Thompson,
2013). New incursions or detection of exotic species are
evaluated for impact and eradication potential as per
DOC weed-led systems (Owen, 1998). Biosecurity to
prevent new incursions is a priority and weed control to
protect threatened plant species in non-forested, coastal
ecosystems is important.
The eradication programme is now focussed on nine
transformer species that have a major impact on forest
ecosystems, four of which are vines (Table 2, and see
West, 1996 for more background on these species). Given
that seabirds are now beginning to return to Raoul Island
to breed, it is particularly important to ensure that vines,
which can entangle landing seabirds as well as smother
native vegetation including forest, are eliminated.
Weed eradication programmes can take a long time,
and many have failed (Panetta, 2015). The Raoul Island
weed eradication programme has been formally reviewed
twice since it began 45 years ago (West, 1996; West &
Havell, 2013). Each time, the species being targeted have
been evaluated to understand impacts of the species, and
effectiveness of the eradication methodology. Both reviews
have resulted in changes to the management programme
and revised lists of species to focus on for eradication. The
latest review restricted the focus to nine species where
eradication will have the biggest impact on biodiversity.
Changes to staffing were also recommended, so that there
would be six months overlap of some experienced staff
with new staff. This recommendation has been actioned
for the contracted staff but the volunteer programme (sixmonth term) was discontinued in 2015 and replaced with
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 435–442. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
435
Island invasives: scaling up to meet the challenge. Ch 2E Other taxa: Plants
seconded staff (three-month term), effectively halving this
additional effort for weed control.
In between the two formal reviews, the programme is
constantly evaluated in relation to all management on the
island. For example, grape (Vitis vinifera) was added to the
list of target transformer species before rats were eradicated
because the two rat species present were preventing fruit
development on the grape vines (West and Havell, 2011).
The year in which eradication commenced for each of the
nine transformer species is shown in Table 3.
The option of eradication of transformer species is more
appealing in the long-term than ongoing control to zerodensity, as eradication means that financial investment in
weed detection and control can cease once the species have
been eliminated. To achieve this, sufficient resourcing is
required to not only achieve the goal but also reduce total
costs (Panetta, 2015). The feasibility of eradication of alien
plants from Raoul Island was evaluated 30 years after the
programme began (West, 2002). At that time, all necessary
conditions (listed in West, 2002) appeared to be met, and
application to the task was what was needed.
Preventing reinvasion is entirely achievable for all
nine remaining target species, given the remoteness of
Raoul Island and the strict biosecurity protocols that are
in place. But how well are the species being extirpated and
contained within Raoul Island as eradication proceeds?
Panetta (2015) describes a model for categorising species in
terms of the ‘technical’ feasibility of eradication by taking
into consideration the relative feasibilities of extirpation
and containment. He notes that eradication occurs via two
processes: (i) extirpation (the elimination of the target in
both space and time) and (ii) containment, which is the
prevention of further occupancy of space (i.e. spread).
This approach is a useful one to apply to the nine target
transformer species on Raoul Island as the work is done
and reported on a plot basis, and it is an advancement
on the methodology proposed by Holloran (2006) for
reporting progress. There are currently 13 weeding blocks
comprising 153 plots of varying size (0.1–83.2 ha),
covering almost 834 ha which is 28.3% of the total area
of Raoul Island (Fig. 1). Plot size varies based on terrain
and travelling time; typically each plot can be carefully
grid-searched in one day (see West, 2002 for more detail
Table 1 Species eradicated from Raoul Island. For each species, the year eradication began and the
year in which the species was last recorded are given. Eradication was formally declared in 2013
by West and Thompson (2013).
Species
Cortaderia selloana
Ficus macrophylla
Foeniculum vulgare
Furcraea foetida
Gomphocarpus fruticosus
Macadamia tetraphylla*
Phoenix dactylifera†
Phyllostachys aurea
Populus nigra
Senecio jacobaea
Vitex lucens
Common name
Eradication began
Last recorded
pampas grass
Moreton Bay fig
fennel
Mauritius hemp
swan plant
macadamia
date
bamboo
poplar
ragwort
puriri
1984
1996
1969
1974
1979
1996
1995
1996
1995
1980
1997
1993
1999
1999
2002
2002
2003 (2015)
1999
2001
2003
1980
1997
* One macadamia seedling was found in 2015 in the same location as the original small stand of trees.
† Wild dates have been eradicated but the species is still present at two historic sites as apparently nonreproductive individuals.
Table 2 Transformer weeds currently being eradicated on Raoul Island. The juvenile period, seed persistence and
dispersal mechanism of each species is used to estimate the feasibility of eradication (Panetta 2015). Species are listed
in order of feasibility of eradication: most to least.
Species
Common name
Growth
form
Juvenile
period
Seed persistence
Dispersal
Olea europaea subsp.
cuspidata
African olive
Tree
5 years
2.4 years1
bird
Psidium cattleianum
P. guajava
Passiflora edulis
Ricinus communis
Senna septemtrionalis
Caesalpinia decapetala
purple guava
yellow guava
black passionfruit
castor oil plant
Brazilian buttercup
Mysore thorn
Small tree
Shrub
Vine
Small tree
Shrub
Vine
2–3 years 6–7 months2
1–2 years c. 1 year3
9 months a few weeks4
5–6 months >19 years5
c. 2 years >16 years6
4–6 months >12 years7
bird
3
bird
4
bird
4
explosive*
6
explosive* 5 or 7
explosive*
6
Anredera cordifolia
Vitis vinifera
Madeira vine
grape
Vine
Vine
< 1 year
1 year
gravity*
bird
15 years (tubers)8
5 years9
Feas- Goal
ibility
3
eradicate
eradicate
eradicate
eradicate
eradicate
eradicate
eradicate
6 or 8 contain
8
eradicate
*Occasional long-distance dispersal by wind, bird, water or accidental-human vectors: for Brazilian buttercup and Madeira vine this
occasional longer distance dispersal has resulted in considerable range extension, therefore, two feasibility estimates are given to
cover both the normal and not uncommon dispersal events. 1Cuneo, et al., 2010; 2Uowolo & Denslow, 2008; 3CABI, 2017b; 4CABI,
2017a; 5Kammili & Jatothu, 2015; 6Ewart, 1908; 7no published data: this estimate is from an isolated infestation of known age on Raoul
Island; 8Harden, et al., 2004; 9no published data found for Vitis vinifera: this estimate is for Vitis aestivalis (Haywood, 1994).
436
West & Havell: Weed eradication progress, Raoul Island, NZ
of the plot-based searching methodology). Using Panetta’s
model (Panetta, 2015), each plot can be evaluated to see if
the species has been extirpated from it, and the distribution
of a species among the plots can be evaluated to see if
the species is being contained or is expanding its range.
Then, the relative relationship between extirpation and
containment can be evaluated for each target species to
determine if eradication can be achieved (Panetta, 2015).
METHODS
On-island weed searching
Details of weed searching and removal are given in
West (2002) and Holloran (2006) and here we restate
briefly what the annual plan and actions are: that weeding
plots should be grid-searched on the ground a minimum of
once each year with plots containing the target transformer
species to be searched twice. Within plots, known
infestations are marked (including GPS coordinates) and
specifically searched during grid-searching or between
grid-searches. Finds of immature plants outside of
infestations are recorded as random finds. New infestations
are created when mature, fruiting plants or localised seed
banks are found and, if a new infestation is large, a new
plot may be created. If no target species have been found
at an infestation for the period of the suspected viability
of the propagule bank based upon database records, or the
site has been destroyed by a landslide or volcanic eruption,
the infestation is retired. GPS tracks of grid-searching are
downloaded to Arc-GIS and used to identify any gaps in
search coverage and to document search effort.
Fig. 1 Raoul Island showing places mentioned and the
distribution of weed plots.
weed database. Recording of the number of individuals
in three stage classes – seedlings, adolescents (taller than
30 cm but not yet flowering) and mature (flowering and/
or fruiting) – began in October 1997, so the database now
holds more than 20 years of continuous data.
The number of times each plot was searched per year,
from 1998 to 2016 was extracted from the database. Also,
the number of active, retired and random infestations
in each plot was summarised for the same period. This
information was used to interpret the data on number of
individuals per target species per plot and per year.
Feasibility of eradication
RESULTS
Data on the factors used by Panetta (2015) to determine
feasibility of eradication were compiled from published
information and, where necessary or more appropriate,
from our observations on Raoul Island. The two key
biological factors relevant to extirpation are the length of
the juvenile period (i.e. how quickly can plants produce
more viable propagules?) and seed persistence (i.e. how
long can seed remain viable?). The biological factor that
is most relevant to containment is dispersal modes (i.e. is
spread likely to be short- or long-distance; predictable –
e.g. water or wind – or unpredictable?). Evaluating the data
for these three factors enables identification of eradication
feasibility on a scale from most feasible (a score of 1) to
least feasible (a score of 8 – see Panetta, 2015, p. 232).
On-island weed searching
Evaluation of progress towards eradication
All data on the number of individuals removed per plot
for the nine target weed species from 1 January 1998 to
31 December 2016 were extracted from the Raoul Island
From 1998 to 2011 the mean number of grid searches
per plot exceeded one a year and exceeded two in 2003
(Fig. 2). However, since 2012 the number of plots gridsearched has dropped well below a single search each
year culminating in less than one third of plots being gridsearched in 2016.
Feasibility of eradication
The data for the nine transformer species targeted for
eradication show a wide range of feasibility (Table 2),
from feasible, e.g. African olive (Olea europaea subsp.
cuspidata) and purple guava (Psidium cattleianum) to
much less feasible e.g. Madeira vine (Anredera cordifolia)
and grape. Species with long juvenile phase (> 2 years),
short seed persistence (< 3 years) and short distance or
largely human-mediated propagule dispersal score lower,
and are therefore more feasible to eradicate. Conversely,
Table 3 Percentage of plots occupied by each species in 1997–2000 (from West 2002) and 1998–2016 as well as
the year in which eradication began.
Species
Common name
Olea europaea subsp. cuspidata
Psidium cattleianum
P. guajava
Passiflora edulis
Ricinus communis
Senna septemtrionalis
Caesalpinia decapetala
Anredera cordifolia
Vitis vinifera
African olive
purple guava
yellow guava
black passionfruit
castor oil plant
Brazilian buttercup
Mysore thorn
Madeira vine
grape
Eradication began
1973
1973
1972
1980
1990
1978
1974
1995
1998
% plot occupancy
1997–2000 1998–2016
19.6
13.6
22.4
25.0
11.9
8.6
32.9
36.4
7.7
6.4
72.0
72.1
18.2
20
2.1
2.1
8.4
8.6
437
Island invasives: scaling up to meet the challenge. Ch 2E Other taxa: Plants
species with short juvenile phase (< 2 years), long seed
persistence (> 3 years) and long distance dispersal score
higher. The short juvenile phase means searching has to
be more frequent; the long seed persistence means the
duration of the programme is longer and is extended every
time new seed is added to the soil if a fruiting individual is
not found in time; the long-distance dispersal means that a
greater area must be searched.
Evaluation of progress towards eradication
All transformer species
The number of active and retired infestations for
each species gives a good indication of progress towards
achieving eradication (Fig. 3). Five species – African olive,
yellow guava (Psidium guajava), castor oil plant (Ricinus
communis), purple guava and black passionfruit (Passiflora
edulis) – have considerably more retired than active plots.
Brazilian buttercup (Senna septemtrionalis) and Mysore
thorn (Caesalpinia decapetala) – the two species with
the greatest seed longevity – have proportionally more
active plots. Control of grape began later than the other
species (Table 3), hence the high proportion of active plots
compared to retired. Madeira vine is the only species with
more active than retired plots.
The random infestations give an indication of dispersal
beyond the immediate vicinity of mature plants (Fig. 3)
and the effectiveness of the programme to control weed
reproduction. Black passionfruit and purple guava, both
bird-dispersed, have a relatively high proportion of random
finds. Mysore thorn has the highest proportion of random
finds reflecting not its dispersal ability but its highly cryptic
nature when growing among the tall ground ferns which
grow densely in parts of Denham Bay, and its extensive
original distribution.
African olive
Numbers of this species detected and removed since
1998 are very low (Fig. 4). The last mature individual was
removed in 2008, and two adolescent plants from different
locations in 2010 and 2011. All of these finds were from
within the historic range of this species before eradication
commenced and the percentage of active plots for this
species has decreased (Table 3).
Purple guava
Numbers of this species were low (West 2002)
but began to increase in 2008, increasing an order of
magnitude in 2011, and with very high numbers recorded
in 2016 (Fig. 4) from just a few infestation plots mostly
within the crater around the shores of Blue Lake and Tui
Lake. Most of the purple guava detected in 2015 were from
new detections in the dry crater near Tui Lake, adjacent
Fig. 2 Mean number of grid searches per plot from 1998–
2016.
438
to old infestations where purple guava was last detected
in 2002. More than half the numbers detected in 2016
were from this infestation. Purple guava infestations are
still being found within the known historical range of the
species, but the percentage of plots occupied has increased
slightly (Table 3). Infestations of purple guava and buffers
of up to 100 metres have been intensively searched since
2015, and, subject to resourcing, additional areas within
the crater are likely to be checked. Seedlings of this species
are very cryptic (look very similar to two of the endemic
species) so careful searching is required.
Yellow guava
There is an increased number of yellow guava
“seedling” detections since 2011 but overall the numbers
are quite low (Fig. 4). The last mature individual detected
was in 2008 with no further detections at that site. The
seedlings recorded are generally suckers from roots: those
recorded in 2011 were suckers from just two plants. A
yellow guava shoot was discovered in 2015 in a crack in a
concrete path close to the Hostel. This may have originated
from root suckers from a relic guava root system in adjacent
gardens, as ongoing persistence and lack of other finds
indicates. However, it is also possible that an undetected
mature plant may be present within the range of local birds.
Yellow guava is active in fewer plots within its historical
range (Table 3).
Black passionfruit
The number of black passionfruit being detected and
removed began to increase markedly from 2004 (Fig. 4),
with the biggest number found so far, in 2011, coming
primarily from one infestation where eight mature vines
had been removed the year before. This site is still very
active. To date, despite the increase in numbers detected,
black passionfruit has not materially exceeded its historic
range. Although it now occupies more plots than previously
(Table 3), these are plots within the bounding polygon that
describes the historic range of this species.
Castor oil plant
Numbers of this species are low (Fig. 4), with the last
mature plants removed in 2003 all from one site. With the
exception of three adolescent plants in 2011 and 2012, all
other seedlings and adolescents removed since 2004 have
come from this site. Castor oil plant has not expanded
beyond its historic range and has fewer active plots than
previously (Table 3).
Brazilian buttercup
This species has been the most numerous since the
eradication programme began. Numbers were declining
Fig. 3 Percentage of active (black bars), retired (mid grey
bars) and random (dark grey bars) infestations within the
weeding plots for each transformer species from 1998–
2016.
West & Havell: Weed eradication progress, Raoul Island, NZ
Fig. 4 Number of seedlings (black bars), adolescent (hatched bars) and mature plants (white bars) of African olive, purple
guava, yellow guava, black passionfruit, castor oil plant, Brazillian buttercup, Mysore thorn and grape removed in the
period 1998–2016 from Raoul Island. For Brazilian buttercup the data for the same period are for removal from Raoul
Island and the Meyer Islets: this is the only one of the target species found on these islets that are c. 1 km NE of Raoul
Island.
439
Island invasives: scaling up to meet the challenge. Ch 2E Other taxa: Plants
effectively until 2008 when they began to increase, reaching
a new peak in 2010 – primarily seedlings (Fig. 4). In 2009,
the significant increase in mature and adolescent plants
is due to the discovery of three outlier populations that
extended the range of this species beyond its historic range.
Two of these sites – the westernmost weed plot (see Fig. 1)
and a plot on the cliffs at the southern end of Denham Bay
were found during helicopter surveillance. The other site,
below bluffs at the northern end of Denham Bay was found
during a routine search of a nearby plot at the back of the
bay. Brazilian buttercup is also on North and South Meyer
Islets and is the only one of the target species found off
Raoul Island. This species occupies a marginally greater
percentage of plots than previously (Table 3): the new plots
described above have been virtually cancelled out by the
retirement of some plots due to slips and the 2006 eruption.
Mysore thorn
Mysore thorn numbers have fluctuated through time but
reached a new peak in 2011 as a result of the high number
of mature vines found in 2010 (Fig. 4): the highest number
of mature plants recorded since 2001 (Fig. 5). Mysore
thorn is confined to Denham Bay now that an infestation
at the head of Ravine 6 has been eradicated. Within
Denham Bay, however, this species’ range has increased
slightly, with helicopter surveillance in 2002 and 2009
leading to the detection of two sites on the cliffs above the
bay, including the southernmost site known (see Fig. 1).
However, it is not these newly discovered infestations that
are contributing the higher numbers of all size classes since
2010, it is a number of the historic plots on the flat and
towards the cliffs north of Denham Bay swamp. Mysore
thorn occupies a slightly higher percentage of plots than
previously (Table 3).
Madeira vine
The weight of tubers removed in 1998 was not recorded
but a file note halfway through that year mentions 60 sacks
of tubers had been removed. The amount of tubers removed
is overall less in the past decade than in the previous one
(Fig. 6) as the more accessible plots are controlled. Various
methods for killing the tubers have been trialled and used,
including composting (in black bins using an accelerant),
burning, desiccation followed by burning of the desiccated
tubers, and freezing (the current method). Madeira vine has
not expanded beyond its historic range of two locations
and has the same percentage plot occupancy as previously
(Table 3). Madeira vine has almost been eradicated from
Bell’s Ravine with only small finds in 2015.
Fig. 5 Number of mature plants of Mysore thorn removed
in the period 1998–2016 from Denham Bay, Raoul Island.
440
Grape
Grape vines are now in very low numbers, with no
new sites since 2012 (Fig. 4). Most of the infestation sites
are in Denham Bay and three are in old settlement areas
on the north side of Raoul Island, reflecting past human
occupancy. The percentage plot occupancy for grape has
increased very slightly (Table 3), reflecting a single mature
vine found in 2011 during grid-searching in Denham Bay.
DISCUSSION
Panetta’s model (Panetta, 2015) is a very useful
framework to evaluate eradication feasibility but when
using it, we have been very conscious of the lack of
accurate data on seed longevity in the soil in Raoul Island’s
environmental conditions. We have observed that some of
the transformer species being targeted on Raoul Island have
seedling banks, e.g. African olive and black passionfruit,
and others have the ability to resprout from underground
roots, e.g. yellow guava and grape. It could be useful for
these mechanisms to be added to the model, perhaps as
propagule persistence (replacing seed persistence) given
that resprouts can appear more than three years after any
other stem material has been present above-ground, and
seedlings can remain in a seedling bank for more than three
years until a light gap is created allowing the seedlings to
rapidly grow to into mature plants.
The graphs of species abundance through time (Figs 4
& 5) combined with life history data indicating feasibility
of eradication (Table 2) as well as plot occupancy (Table 3)
and the number of active, retired and random infestations
(Fig. 3) indicate that eradication is very achievable for four
species: African olive, yellow guava, castor oil plant and
grape. Note that species differ in their life history traits
so therefore have different eradication feasibility scores
(Table 2). Grape has the highest score (least feasible) but is
eradicable because the biomass of all grapes was reduced
to essentially zero before the rats were eradicated (West
& Havell, 2011). Dispersal of this species has not been
possible because all resprouts are found and destroyed
before fruits are formed.
With no detections of African olive since 2011 and
estimated seed persistence of 2.4 years, it is theoretically
possible to declare this species eradicated now. However,
given the cryptic nature of this species and seedling
persistence, our preference is to wait until at least 2021
before making this claim (if there are no further detections).
Yellow guava persists as occasional suckers in just three
accessible locations, presumably from relict root systems,
so should be eradicable with annual checks of the locations
although the timeframe is difficult to estimate. Castor
Fig. 6 Quantity (kg) of Madeira vine tubers (aerial and
ground) removed in the period 1998–2016 from Raoul
Island.
West & Havell: Weed eradication progress, Raoul Island, NZ
oil plant has seed longevity of >19 years so with the last
detection of this species in 2012, and assuming no further
detections, it could be deemed to have been eradicated
by 2040. All grapes detected on Raoul Island so far have
been resprouts from persistent root systems. Given the low
feasibility of eradicating this species if it were reproducing
by seed (Table 2) it is vital that fruit are never produced.
All plots containing grape must be searched a minimum of
once a year, ideally twice, given the short juvenile period
possible for resprouts.
The feasibility of eradication for purple guava and
black passionfruit is relatively high (Table 2), however,
both species have been recorded in highest numbers in
recent years (Fig. 4) and have a relatively high proportion
of random finds (for black passionfruit there have been
more random finds than there are active plots). The trend
for both species indicates that the current search effort (Fig.
2) of less than one plot search per year, and the current
area searched, is not sufficient to prevent seed dispersal.
Black passionfruit is known to produce fruit within one
year on Raoul Island, and the development of mature
plants and purple guava seedlings may have occurred in
the dry crater in the three years between 2012 and 2015.
At the moment, both species have still been found within
their historic range but, since they are bird dispersed, it is
quite possible that these two species could be spreading to
areas outside the current extent of searched plots (Fig. 1).
All of these results indicate that grid searching for both
purple guava and black passionfruit must continue until
suppression of reproduction in these species is clearly
demonstrated. Searching needs to be undertaken within
the known range of these species and be extended into
surrounding areas to detect any new infestations from birddispersed seeds. Uowolo and Denslow (2008) suggest the
most effective time for purple guava control is at least three
months after the fruiting season when the majority of seeds
have germinated or died, given the short seed persistence.
Another risk is the potential of large purple guava plants
to sucker and reproduce rapidly after long periods of
quiescence. A weed detection dog is being trained to focus
on black passionfruit, both guava species and grape but is
not yet ready for deployment.
Brazilian buttercup is the most widespread of the
target species (Table 3). Range extensions of this species,
discovered in 2009, plus its occurrence on the Meyer Islets
indicates this species has rare long-distance dispersal,
possibly by birds (although human dispersal can’t be ruled
out). In order for this species to be eradicated, detection
methods need strong focus (Holloran, 2006). Any
opportunities for helicopter surveillance should be taken,
particularly when these coincide with the flowering period.
Although Mysore thorn is confined to Denham Bay,
too many individuals are being missed in plot searches
allowing considerable seed set e.g. the very high numbers
of seedlings recorded in 2011 (Fig. 4). Seed germination
in 2011 could also have been aided by two cyclones that
affected Raoul Island in February (Atu) and March (Bune),
with the latter resulting in widespread treefalls and stripping
of foliage from trees. Given the rapid growth rate of Mysore
thorn (Table 2) and the high proportion of random finds, the
plots in Denham Bay need to be searched a minimum of
twice each year and possibly with a closer spacing between
observers than in the past (we suggest a minimum of 2 m).
The short juvenile period plus the long seed persistence
time make this species less feasible for eradication.
However, because long-distance dispersal is very rare,
this species is eradicable if seed banks and fruiting can be
eliminated (as demonstrated by the eradication in Ravine
6). Of all the transformer species, Mysore thorn poses the
greatest threat to ecosystem recovery, as shown by historic
photographs and reports of Mysore thorn smothering the
Kermadec pohutukawa canopy in Denham Bay (West,
1996). Landing seabirds can get entangled in vegetation:
vines provide greater opportunities for entanglement and
thorny vines (like Mysore thorn) less opportunity for safe
escape (Arcilla, et al., 2015).
Madeira vine is the most difficult species to control on
Raoul Island (West, 2002). It was last detected in its original
location in Bell’s Ravine in 2015 but because it can grow
from tiny aerial tubers and subterranean tubers, it may still
occasionally resprout in that location. However, at the main
location east of Fishing Rock, this species grows on steep,
unstable cliffs so tubers can be removed only from the most
accessible sites and places that can be reached by abseiling
safely. Herbicide is still used to knock back foliage when
necessary to gain access to the herbicide-resistant tubers so
they can be removed. Until a control method is developed
that can kill tubers on the inaccessible cliffs, this species
can only be contained rather than eradicated. Management
so far has successfully contained Madeira vine.
It has been stated frequently in the literature that
eradications are unlikely to succeed if the area occupied
is large (Panetta, 2015). Howell (2012) identified that the
only successful eradications of environmental weeds in
New Zealand were those where the initial extent was <
1 ha, noting that there were other eradications of similar
extent that were unsuccessful. On Raoul Island, four
species currently have distributions of < 1 ha: African
olive, yellow guava, castor oil plant and grape although
the area to be grid-searched in the plots within which they
occur ranges from 6 ha (castor oil plant) to 80 ha (grape).
The area to be searched for the more abundant species –
purple guava, black passionfruit, Brazilian buttercup and
Mysore thorn ranges from 60 ha (Mysore thorn) to 550 ha
(Brazilian buttercup).
However, Panetta’s (2015) model of extirpation and
containment indicates that African olive, yellow guava,
castor oil plant and grape are all currently being extirpated
(the rate of extirpation of managed infestations exceeds
the rate of establishment of new ones) and contained, and
could be eradicated with the current level of resourcing.
As indicated above, breakthroughs in methodology are
needed for extirpation of Madeira vine to become a reality.
For the other four species, there have been more new
infestations leading to greater numbers and, for Brazilian
buttercup, range extension in recent years. The current
level of resourcing is not sufficient to enable eradication
as, based upon GPS track logs, not all known and potential
areas are able to be searched within the time it takes for
each species to fruit. There is also insufficient resourcing to
analyse records of infestation within the Raoul Island weed
database, and therefore plan the work more effectively.
There are many factors that have led to this situation.
Rats (Rattus norvegicus and R. exulans) are no longer
eating flowers, seeds and seedlings of plants. This effect
can be seen in the results for several species, e.g. Mysore
thorn, black passionfruit and yellow guava. No access to
the crater was permitted for two years after the eruption
in 2006 resulting in mature plants of purple guava and
Brazilian buttercup, with dispersal of the former and seed
added to the seed bank for the latter. Cyclones have been
frequent in the past decade, resulting in large areas of
windfallen trees within weed control sites that stimulates
germination (good for reducing the seed bank) but also
impedes access and slows down the rate of grid-searching,
making weed removal harder. A formal process for retiring
plots is not in place and this may have resulted in a lack of
focus on areas that need more regular checking.
Health and safety requirements and biosecurity
management are the factors that have most influenced
the drop in the mean number of plots grid-searched (Fig.
2). Staff effort has been directed towards biosecurity
management as the number of visitors to the island
441
Island invasives: scaling up to meet the challenge. Ch 2E Other taxa: Plants
(including organised tour groups) has increased over the
years. Health and safety requirements have increased
considerably since 2006: permission must be sought to
enter the crater to search the plots (the granting authority is
the Operations Manager based on advice from the Institute
of Geological and Nuclear Sciences which monitors
seismic activity on the island); weed plots that must not
be searched after heavy rain or during seismically active
periods and those that require climbing gear have been
identified. Staff numbers have been reduced since 2015
when deployment of volunteers for six-month periods was
replaced by seconded staff for three-month periods.
These factors, combined with the lack of assistance from
rats, reduced staff resourcing, plus the increased number
of cyclones, has led to the increase in weed abundance,
particularly for purple guava, black passionfruit, Brazilian
buttercup and Mysore thorn. Health and safety standards
should never be compromised but need to be compensated
for by increased resourcing for the eradication of
transformer weeds to be successful. The current budget for
the programme is $555,000, reduced from $566,000 in the
previous year. This needs to be increased to $850,000 to
provide sufficient staff time to check all plots once a year
plus an additional check of all plots with known Mysore
thorn, grape and black passionfruit infestations (two checks
a year, minimum). Madeira vine plots should be checked
at the current rate of one day per week. As Panetta (2015)
states “Despite the best of intentions and the highest level
of professionalism, an eradication effort will not succeed
if it is not adequately resourced”. Increasing the budget is
a wise move given that Cacho, et al., (2007) showed that
total cost of weed eradication is high when low search
effort was involved, but falls rapidly with increasing search
effort because a more intense search effort would reduce
the number of reproductive plants. This is currently the
sticking point for four of the nine transformer species in the
Raoul Island weed eradication programme. Whereas the
bulk of the budget should be spent on Raoul Island (staff,
infrastructure, materials), the off-island support resourcing
is also vital and must be set at the optimal level. Remote
island-based programmes require huge logistical support
from dedicated teams both on and off the island.
The preceding discussion describes a number of factors
that need to be considered when budgeting for success in
this programme and the results described are all influenced
by these. However, myrtle rust, whose impact is yet to
be felt by Raoul Island ecosystems, is the latest agent of
change that will need to be considered. Monitoring plots
have been established to provide early indications of how
this disease will affect Kermadec pohutukawa and what the
flow-on effects will be. It is possible that the dynamics of
the transformer weeds in this eradication programme will
change, just as they have following eradication of rats and
cyclones.
ACKNOWLEDGEMENTS
Thanks to staff on Raoul Island, and those on the
mainland associated with the weed eradication programme
for providing information for this paper. Thanks to Geoff
Woodhouse, Paul Rennie and Clayson Howell for their
comments on an earlier draft of this paper. A special
thanks to the staff who undertake the arduous task of weed
eradication on the island. Thanks, also, to the anonymous
reviewers who improved this paper and to Dick Veitch for
improving the figures.
REFERENCES
Arcilla, N., Chang-Yong, C., Kiyoaki, O. and Lepczyk C.A. (2015).
‘Invasive species and Pacific island bird conservation: a selective
review of recent research featuring case studies of Swinhoe’s storm
petrel and the Okinawa and Guam rail’. Journal of Ornithology 156
(Suppl. 1): S199–207.
442
Broome, K.G. (2009). ‘Beyond Kapiti – a decade of invasive rodent
eradications from New Zealand islands’. Biodiversity 10: 7–17.
Cacho, O.J., Hester, S.M. and Spring, D. (2007). ‘Applying search theory
to determine the feasibility of eradicating an invasive population in
natural environments’. Australian Journal of Agricultural and Resource
Economics 51: 425– 433.
Cuneo, P., Offord, C.A. and Leishman, M.R. (2010). ‘Seed ecology of the
invasive woody plant African olive (Olea europaea subsp. cuspidata):
implications for management and restoration’. Australian Journal of
Botany 52: 342–348.
Ewart, E.J. (1908). ‘Vitality of seeds’. Proceedings of the Royal Society
of Victoria 2: 1–203.
Harden, G.J., Fox, M.D. and Fox, B.J. (2004). ‘Monitoring and assessment
of restoration of a rainforest remnant at Wingham Brush, NSW’. Austral
Ecology 29: 489–507.
Haywood, J.D. (1994). ‘Seed viability of selected tree, shrub, and vine
species stored in the field’. New Forests 8: 143–154.
Holloran, P. (2006) ‘Measuring performance of invasive plant eradication
efforts in New Zealand’. New Zealand Plant Protection 59: 1–7.
Howell, C.J. (2012). ‘Progress toward environmental weed eradication
in New Zealand’. Invasive Plant Science and Management 5: 249–258.
CABI, (2017a). Passiflora edulis (passionfruit). In: Invasive Species
Compendium. Wallingford, UK: CAB International. <www.cabi.org/
isc>.
CABI, (2017b). Psidium guajava (guava). In: Invasive Species
Compendium. Wallingford, UK: CAB International. <www.cabi.org/
isc>.
Kammili, A. and Jatothu, J. (2015). ‘Differential longevity of castor
(Ricinus communis L.) germplasm conserved under uncontrolled storage
conditions across extended periods’. Journal of Crop Improvement 29:
706–719.
Owen, S-J. (1998) Department of Conservation Strategic Plan for
Managing Invasive Weeds. Department of Conservation, Wellington.
Panetta, F.D. (2015). ‘Weed eradication feasibility: lessons of the 21st
century’. European Weed Research Society 55: 226–238.
Parkes, J.P. (1984). ‘Feral goats on Raoul Island II: Diet and notes on the
flora’. New Zealand Journal of Ecology 7: 95–101.
Pyšek, P., Richardson, D.M., Rejmánek, M., Webster, G.L., Williamson,
M. and Kirschner, J. (2004). ‘Alien plants in checklists and floras:
towards better communication between taxonomists and ecologists’.
Taxon 53: 131–143.
Sykes, W.R. and West, C.J. (1996). ‘New records and other information
on the vascular flora of the Kermadec Islands’. New Zealand Journal of
Botany 34: 447–462.
Sykes, W.R., West, C.J., Beever, J.E. and Fife, A.J. (2000). Kermadec
Islands Flora - Special Edition: A Compilation of Modern Material
about the Flora of the Kermadec Islands. Lincoln, Canterbury, N.Z.:
Manaaki Whenua Press.
Uowolo, A.L. and Denslow, J.S. (2008). ‘Characteristics of the Psidium
cattleianum (Myrtaceae) seed bank in Hawaiian lowland wet forests’.
Pacific Science 62: 129–135.
Veitch, C.R., Gaskin, C., Baird, K. and Ismar, S.M.H. (2011). ‘Changes
in bird numbers on Raoul Island, Kermadec Islands, New Zealand,
following the eradication of goats, rats and cats’. In: C.R. Veitch,
M.N. Clout and D.R. Towns (eds.) Island invasives: eradication and
management, pp. 372–376. Occassional Paper SSC no. 42. Gland,
Switzerland: IUCN and Auckland, New Zealand: CBB.
Veitch, C.R., Miskelly, C.M., Harper, G.A., Taylor, G.A. and Tennyson,
A.J.D. (2004). ‘Birds of the Kermadec Islands, south-west Pacific’.
Notornis 51: 61–90.
West, C.J. (1996). Assessment of the weed control programme on Raoul
Island, Kermadec Group. Science & Research Series no 98. Department
of Conservation, Wellington.
West, C.J. (2002). ‘Eradication of alien plants on Raoul Island, Kermadec
Islands, New Zealand’. In: C.R. Veitch and M.N. Clout (eds.) Turning
the tide: the eradication of invasive species, pp. 365–373. Occasional
Paper SSC no. 28. IUCN SSC Invasive Species Specialist Group.
IUCN, Gland, Switzerland and Cambridge, UK.
West, C.J. (2011). ‘Consideration of rat impacts on weeds prior to rat
and cat eradication on Raoul Island, Kermadecs’. In: C.R. Veitch,
M.N. Clout and D.R. Towns (eds.) Island invasives: eradication and
management, pp. 244–247. Occassional Paper SSC no. 42. Gland,
Switzerland: IUCN and Auckland, New Zealand: CBB.
West, C.J. and Havell, D. (2011). ‘Plant responses following eradication
of goats and rats from Raoul Island, Kermadecs’. In: C.R. Veitch,
M.N. Clout and D.R. Towns (eds.) Island invasives: eradication and
management, p. 535. Occassional Paper SSC no. 42. Gland, Switzerland:
IUCN and Auckland, New Zealand: CBB.
West, C.J. and Havell, D. (2013). Raoul Island Review – Phase 2 Part
1: Review of Biodiversity programme. Unpublished internal report,
DOCDM-1157983. Department of Conservation.
West, C.J. and Thompson, A.M. (2013). ‘Small, dynamic and recently
settled: responding to the impacts of plant invasions in the New Zealand
(Aotearoa) archipelago’. In: L.C. Foxcroft, P. Pyšek, D.M. Richardson
and P. Genovesi (eds.) Plant Invasions in Protected Areas: Patterns,
Problems and Challenges, 7: 285–311. Invading Nature – Springer
Series in Invasion Ecology.
L. Ballantyne, D. Baum, C.W. Bean, J. Long and S. Whitaker
Ballantyne, L.; D. Baum, C.W. Bean, J. Long and S. Whitaker. Successful eradication of signal crayfish
Pacifastacus leniusculus using a non-specific biocide in a small isolated water body in Scotland
Successful eradication of signal crayfish (Pacifastacus leniusculus)
using a non-specific biocide in a small isolated water body in Scotland
L. Ballantyne1, D. Baum1, C.W. Bean2, J. Long3 and S. Whitaker4
Lochaber Fisheries Trust, Torlundy, Fort William, PH33 6SW, UK. 2Scottish Natural Heritage, Caspian House,
Mariner Court, Clydebank Business Park, Clydebank, G81 2NR, UK. 3Scottish Environment Protection Agency,
Strathallan House, Stirling, FK9 4TZ, UK. 4Scottish Natural Heritage, Silvan House, 231 Corstorphine Road,
Edinburgh, EH12 7AT, UK. <stan.whitaker@nature.scot>.
1
Abstract The North American signal crayfish (Pacifastacus leniusculus) has been present in Scotland since at least
1995 and the species is now known to be present in a number of catchments. Once established, few opportunities for
containment exist and eradication can often be impossible to achieve. However, in small, isolated water bodies, the
application of a non-crayfish-specific biocide has provided the opportunity to remove this species permanently. In July
2011, signal crayfish were discovered in a flooded quarry pond at Ballachulish in the Scottish Highlands. This is an
isolated site located ~100 km from the nearest known population and it is likely that the population was established
as the result of a deliberate release of these animals 10 years previously. Experience gained from using the eradication
technique at other sites in the UK led to the site being treated with a natural pyrethrum biocide (Pyblast®) in June 2012.
Post treatment monitoring from 2012–2017 indicates that eradication has been successful. Monitoring of native species
affected by the biocide suggests that both invertebrates and amphibians quickly recolonised the quarry pond. Eradication
of crayfish using biocide is only feasible in water bodies where the entire population of crayfish can be exposed to a lethal
dose and the impact on non-target species can be accepted. The technique is not appropriate for large, connected water
bodies, although it may be possible to treat short stretches of canals where biocide exposure can be controlled and isolated
populations of crayfish can be effectively treated.
Keywords: invasive species, natural pyrethrum, ponds
INTRODUCTION
Invasive species are the second largest cause of
biodiversity loss globally through species extinction and
habitat destruction (EEA, 2012). Their impact can be
dramatic and often irreversible, so it is important that
their spread is contained and that eradication is achieved
wherever practicable. As a function of their isolation,
islands may offer the best hope of locally eradicating an
invasive species. Conventionally we think of islands as
areas of land which are surrounded by water. However, for
obligate aquatic species the reverse may be true, and it is
the land which can form an effective barrier to invasion.
In freshwater ecosystems, invasive non-native species can
pose a major threat to native species through competition,
predation and transmission of diseases (EEA, 2012) and
their control in these ‘freshwater islands’, is therefore of
particular importance.
Newly introduced species can establish rapidly and it
is important to detect their presence, and take action, as
early as possible. This is often not possible because, unlike
terrestrial habitats, freshwaters are not easily surveyed
(Boon & Bean, 2010). This means that invasive species
in these habitats may not be detected until they have
become fully established, often making it more expensive
to remove them (Simberloff, et al., 2013).
Signal crayfish (Pacifastacus leniusculus) have been
introduced to over 20 European countries since the 1960s.
After escaping from farms in the 1970s they are now
widespread across parts of England and Wales (Bean, et
al., 2004). The species was first discovered in Scotland in
1995 (Maitland, 1996). In just over 10 years it had been
illegally introduced into at least eight river catchments
(Gladman, et al., 2009). Signal crayfish are omnivores
and, through increased grazing pressure and predation,
they can reduce the diversity of aquatic invertebrates
and significantly alter food webs (Holdich, et al., 2014).
As well as direct predation of eggs and young fish, they
compete with Atlantic salmon (Salmo salar ) and trout
(S. trutta ) for food and space and can mobilise sediment,
causing silting of spawning beds (Gladman, et al., 2012;
Bean & Yeomans, 2016).
Controlling signal crayfish has proved difficult and, in
most situations, impossible to achieve. Several approaches
have been attempted, ranging from the physical removal
of animals using techniques such as trapping and
electrofishing, to the construction of barriers to prevent
their spread (Bean & Yeomans, 2016). Of these, trapping
is often perceived as being the easiest and most effective
option. In reality, however, the removal of crayfish by
trapping has proved ineffective at eradicating signal
crayfish because it does not remove the entire population
(Freeman, et al., 2010). Where trapping has been allowed to
take place on a commercial basis, either as a management
tool or for the establishment of legal fisheries, it has been
associated with the detection of an increased number of
illegal introductions (e.g. Alonso, et al., 2000; DiéguezUribeondo, 2006; Arce & Alonso, 2011; Bohman, et al.,
2011).
The use of biocides to control or eradicate crayfish
populations is a relatively recent development. Early
attempts to eradicate signal crayfish using chlorinated lime
(Kozak & Policar, 2003) were not successful. However,
later trials using natural pyrethrum (as Pyblast®) (Peay,
et al., 2006) showed more promise in trials in Scottish
freshwaters without being totally effective. O’Reilly
(2015) provides a comprehensive review of the toxicity of
Pyblast® and other organophosphates for signal crayfish
control.
There is no single biocide available that is selective
for signal crayfish only. This means that any attempted
eradication using a biocide treatment would be expected
to kill some, or all, of the non-target invertebrate and
vertebrate fauna in the area being treated.
Signal crayfish were first detected in north-west
Scotland in an artificial waterbody, a flooded slate quarry,
near Ballachulish, in 2011. This species is thought to have
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 443–446. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
443
Ballantyne, et al.: Eradication of signal crayfish
been present within the pond for approximately 12 years
prior to its discovery there (P. Madden, pers. comm.).
The nearest signal crayfish population to that discovered
at Ballachulish is located over 100 km south in the River
Kelvin near Glasgow. This reinforced the initial view that
this species was introduced to the Ballachulish quarry pond
by people; in addition, it has footpaths and a recreational
area adjacent so is readily accessed by the public. The pond,
and therefore the crayfish population, was isolated with no
source of natural re-infestation. However the proximity
of the pond to local rivers, and the potential impact that
this species may have on species of conservation and
recreational value, such as Atlantic salmon and trout, made
it essential that the signal crayfish population was removed
as soon as possible.
NN08435835). The affected waterbody has a surface area of
18,776 m2 and a volume of 46,000 m3. Whilst it is relatively
shallow over much of its surface area (approximately 0–5
m deep), a smaller area of deeper water, extending to a
maximum depth of 13 m, is present.
Survey prior to any management action revealed that
signal crayfish were restricted to the larger of the two
ponds. The large pond also hosted a number of invertebrate
and vertebrate species. Vertebrates found during the survey
included fish (trout and European eel (Anguilla anguilla))
and amphibian species such as common toad (Bufo bufo)
and palmate newt (Lissotriton helveticus).
METHODS
Ballachulish quarry pond (Ordnance Survey Great
Britain National Grid Reference NN08525824) is located
immediately west of the town of Ballachulish on the west
coast of Scotland (Fig. 1). The area contains a large pond
and a smaller waterbody located 25 m to the north (NGR
It was deemed acceptable, given the absence of any
conservation priority species, that some mortality of native
fauna would occur as a result of biocide application. The
risk of inadvertently transferring juvenile crayfish in the
act of translocating rescued animals to new locations meant
that no attempt was made to rescue non-target species prior
to the treatment taking place.
Fig. 1 The location of Ballachulish and the quarry pond
relative to western Scotland.
Bathymetric transects of the pond were obtained by the
use of a plumb-line at 100 sample points (Fig. 2). These
were used to divide the pond into compartments of equal
volume. A total volume of 620 l of Pyblast® was applied to
the surface of the pond by boat-mounted sprayers (Fig. 3)
to achieve a target dose rate of at least 0.3 mg/l, on 12 June
2012. Water pumps and a boat with an outboard motor
were used to ensure thorough mixing throughout the entire
water column. In addition, backpack sprayers treated a 1 m
band around the edge of the pond and the shallow margins
of the pond to prevent signal crayfish leaving the water
(Fig. 4). The following day, deep water sections of the
pond were re-treated by spraying Pyblast® down 6 m-long
rigid hoses, increasing the dose rate in the deepest areas
of the pond to at least 0.4 mg/l. Mixing was achieved, as
far as possible, using an outboard engine and shore-based
pumps.
Study site
The effectiveness of the treatment was monitored
by placing 13 sentinel cages, each containing 10 signal
crayfish, of mixed sex, into the pond at different positions
and depths and monitoring their mortality once the biocide
had been applied. Bioassays using the freshwater shrimp
(Gammarus pulex) as a test organism, were conducted
according to the methodology described by Peay, et al.
(2006). These were run on the pond water to monitor
its toxicity at the point of treatment and to monitor the
breakdown of the Pyblast® over subsequent days and
weeks. Natural pyrethroids break down quickly when
exposed to sunlight and their toxicity should reduce
Fig. 2 Ballachulish quarry pond bathymetry.
444
Fig. 3 Pyblast® being applied to the surface of the pond
from boat-mounted sprayers.
Island invasives: scaling up to meet the challenge. Ch 2F Other taxa: Aquatic
Bioassay monitoring indicated that after one month
the concentration of Pyblast® in the pond was below the
lethal limit for G. pulex and it was judged safe to re-open
the pond to the public. Fig. 6 shows the speed at which
the reduction in toxicity of water samples taken from the
surface and 5 m depth in the pond took place following
Pyblast® treatment. These data show that biocide toxicity
in deeper waters took longer to drop below lethal levels
than those near the surface, but confirmed that toxicity
levels dropped to levels non-lethal to signal crayfish in all
areas within 20 days post-treatment.
Fig. 4 Using a backpack sprayer to deliver biocide to the
quarry pond edge.
The amphibian surveys found larval stages of common
toad and palmate newt in late June 2012, which strongly
suggested they had survived the Pyblast® treatment. There
was no difference in size or development stage between
tadpoles from the treated pond and a nearby untreated
pond. All amphibian larvae behaved normally and showed
no physical abnormalities (see O’Brien, et al., 2013). A
low level of fish mortality was observed and this included
one brown trout plus a very small number of European eels
and three-spined sticklebacks (Gasterosteus aculeatus).
DISCUSSION
Fig. 5 Dead signal crayfish in the margins of the pond
following Pyblast® treatment.
rapidly. During this eradication exercise, toxicity levels,
sufficient to kill Gammarus, persisted for 34 days.
The effectiveness of the signal crayfish removal
attempt was monitored through baited Fladen-style traps
set in the pond for a total of 195 trap nights in August/
September each year for five years post-treatment (975
trap nights in total). Traps were set in a range of habitats
and depths throughout the site to maximise the potential
of capture. The ability of invertebrates to recover very
quickly, in as little as 24 days, was already known from
other studies (e.g. Peay, et al., 2006), therefore recovery
of the pond ecosystem was assessed through amphibian
surveys carried out using sweep netting and kick-sampling
in late June, August, September and October 2012. Larval
common toad and palmate newts were measured, aged and
their general behaviour assessed to determine whether it
deviated from that normally expected in undisturbed sites.
Post-treatment monitoring demonstrated that the
application of Pyblast® at a target dose rate of 0.3 mg/l
was successful in removing signal crayfish from the pond.
Monitoring also showed that a number of non-target
species survived the treatment, or were able to recolonise
quickly (O’Brien, et al., 2013). The pond is artificial and
the presence of signal crayfish would have significantly
altered its ecology, meaning that there is no recent, or precrayfish, baseline against which to measure ecosystem
recovery. However, five years after the biocide treatment
an abundant invertebrate and amphibian fauna is present
within the pond, and no lasting chemical effect of the
treatment is visible.
The risk of signal crayfish being spread to new
waterbodies within the local area by natural or
anthropogenic means has been reduced as a result of
this successful eradication. There are now no known
populations in the north-west Highlands which pose a threat
of re-introduction to this site. This project has shown that
full eradication is achievable in small, isolated waterbodies
where the entire signal crayfish population can be exposed
to a natural pyrethroid biocide, and the impact on nontarget species is deemed an acceptable risk.
RESULTS
During, and immediately after the Pyblast® application,
signal crayfish held in sentinel cages were checked
intermittently to assess mortality levels and the efficacy
of treatment. By the end of the first day (12 June 2011)
most of the signal crayfish were dead (Fig. 5), however,
those in deep water sections (as determined from the use
of sentinel crayfish) were still active. Effort was focused
on increasing the concentration of Pyblast® in these areas
and by the third morning (14 June 2011) all signal crayfish,
even in the deep sections, were dead. The annual posttreatment monitoring found no signal crayfish in the pond
for five years after the treatment and in August 2017 the
eradication was declared successful.
Fig. 6 Graph showing the reduction in toxicity of water
samples taken from the surface and 5 m depth in the
pond following Pyblast® treatment on 12t June 2012.
Toxicity was estimated through a bioassay exposing
Gammarus pulex to diluted samples and comparing to
previous reference data collected on their mortality rates.
445
Ballantyne, et al.: Eradication of signal crayfish
A partnership approach to dealing with signal
crayfish in this location was a significant component of
its success. Buy-in from public agencies and the local
fisheries management sector provided the financial and
physical resources required to provide adequate materials
to carry out this work effectively. Ongoing promotion of
a national biosecurity campaign aims to prevent future
reintroductions. At an operational level, careful monitoring
of sentinel signal crayfish and having a sufficient
contingency of Pyblast® to supplement concentrations in
the deepest areas of the pond proved crucial. The quarry
pond at Ballachulish is the largest water body in the UK
to date where signal crayfish have been eradicated using
a natural pyrethroid. The main limitations to the wider
application of this method to large waterbodies are the
financial cost of the biocide (in 2012, Pyblast® cost over
£50 per litre), the manpower required, the collateral damage
to native biota and connectivity to outflowing rivers and
streams. This trial accounted for biocide costs of >£30k
alone, and with additional costs in terms of staff time and
equipment hire (pumps, etc.) the total estimated figure was
£73.1k. Additional costs associated with post-treatment
monitoring are not included within this total. In Sweden
and Norway, less expensive synthetic pyrethroids have
been used (Sandodden & Johnsen, 2010), but these have
the disadvantage of being more toxic and persistent in the
freshwater environment. O’Reilly (2015) showed, using
laboratory-based acute toxicity tests, that signal crayfish
were most sensitive to Detamethrin, a synthetic pyrethroid,
used in the aquaculture industry, and that juvenile signal
crayfish were significantly more sensitive to Pyblast® than
adult conspecifics at concentrations far lower than those
used in this study (57.95 μg/l versus 0.3 mg/l). It may be
possible, therefore, to use alternative biocide approaches
in some situations, or lower the costs of treatment in
populations which are detected at an earlier stage in their
establishment. Recent advances in the detection of invasive
species by environmental DNA may allow for earlier, and
cheaper, identification of new populations through the
expansion of surveillance networks to include a larger
number of waterbodies. Environmental DNA assays have
already been developed for signal crayfish (Larson, et al.,
2017; Harper, et al., 2018) and a wide range of other biota
(e.g. Ficetola, et al., 2008) which will allow for cheaper, and
possibly more reliable, pre-and post-treatment monitoring
of signal crayfish and other species to take place in future
years.
ACKNOWLEDGEMENTS
This work would not have been possible without the
financial support provided by Highland Council and
Scottish Natural Heritage. Technical and logistical support
was provided by, Fishery Trusts, Forestry Commission and
the Scottish Environment Protection Agency. We gratefully
acknowledge the guidance provided by Dr Stephanie Peay
during this work.
REFERENCES
Alonso, F., Temiño, C. and Diéguez-Uribeondo, J. (2000). ‘Status of the
white-clawed crayfish, Austropotamobius pallipes (Lereboullet, 1858)
in Spain: Legislation and conservation’. Bulletin Francais de la Peche
et de la Pisciculture 356: 31–54.
Arce, J.A. and Alonso, F. (2011). ‘Factors related to the presence of
the Austropotamobius pallipes (Lereboullet, 1858) species complex
in calcareous mountain rivers in central Spain’. Knowledge and
Management of Aquatic Ecosystems 401: 25.
Bean, C.W., Maitland, P.S. and Collen, P. (2004). ‘Crayfish in Scotland:
A review of current status and legislative control’. Freshwater Crayfish
15: 220–228.
Bean, C.W. and Yeomans, W.E. (2016). ‘North American Signal Crayfish’.
In: M.J. Gaywood, P.J. Boon, D.B.A Thompson and I.M Strachan (eds.)
The Species Action Framework Handbook, pp. 314–324. Perth, UK:
Scottish Natural Heritage.
446
Bohman, P., Degerman, E., Edsman, L. and Sers, B. (2011). ‘Exponential
increase of signal crayfish in running waters in Sweden – due to illegal
introductions?’ Knowledge and Management of Aquatic Ecosystems
401: 23.
Boon, P.J. and Bean, C.W. (2010). ‘Freshwater Non-Natives – Prevention,
Control or Eradication?’ In: J.M. Baxter and C.A. Galbraith (eds.)
Species Management: Challenges and Solutions for the 21st Century,
pp. 229–246. London, UK: Her Majesty’s Stationery Office.
Diéguez-Uribeondo, J. (2006). ‘The dispersion of the Aphanomyces
astaci-carrier Pacifastacus leniusculus by humans represents the main
cause of disappearance of the indigenous crayfish Austropotamobius
pallipes in Navarra’. Bulletin Francais de la Peche et de la Pisciculture
380–381: 1303–1312.
EEA (2012). The Impacts of Alien Species in Europe. Copenhagen:
European Environment Agency.
Ficetola, G.F., Miaud, C., Pompanon, F. and Taberlet, P. (2008). ‘Species
detection using environmental DNA from water samples’. Biology
Letters 4: 423–425.
Freeman, M.A., Turnbull, J.F., Yeomans, W.E. and Bean, C.W. (2010).
‘Prospects for management strategies of invasive crayfish populations
with an emphasis on biological control’. Aquatic Conservation: Marine
and Freshwater Ecosystems 20: 211–223.
Gladman, Z., Adams, C., Bean, C., Sinclair, C. and Yeomans, W. (2009).
‘Signal Crayfish in Scotland’. In: J. Brickland, D.M. Holdich and E.M.
Imhoff (eds.) Crayfish Conservation in the British Isles. Proceedings
of a Conference held on 25th March 2009 in Leeds, UK, pp. 43–48.
Bristol, UK: Environment Agency.
Gladman, Z.F., Adams, C.E., Bean, C.W., Long, J. and Yeomans, W.E.
(2012). ‘Investigating the threat of non-native North American signal
crayfish (Pacifastacus leniusculus) to salmonid redds’. Aquatic
Conservation: Marine and Freshwater Ecosystems 22: 134–137.
Harper, K.J., Anucha, P.N., Turnbull, J.F., Bean, C.W. and Leaver, M.J.
(2018) Searching for a signal: Environmental DNA (eDNA) for the
detection of invasive signal crayfish, Pacifasticus leniusculus (Dana,
1852). Management of Biological Invasions 9: 137-148.
Holdich, D.M., James, J., Jackson, C. and Peay, S. (2014). ‘The North
American signal crayfish, with particular reference to its success as an
invasive species in Great Britain’. Ethology, Ecology and Evolution 26:
232–262.
Kozak, P. and Policar, T. (2003). ‘Practical Elimination of Signal Crayfish,
Pacifastacus leniusculus (Dana), from a Pond’. In: D.M. Holdich
and P.J. Sibley (eds.) Management and Conservation of Crayfish:
Proceedings of a conference held on 7 November 2002, pp. 200–208.
Bristol, UK: Environment Agency.
Larson, E.R., Renshaw, M.A., Gantz, C.A., Umek, J., Chandra, S., Lodge,
D.M. and Egan, S.P. (2017). ‘Environmental DNA (eDNA) detects the
invasive crayfishes Orconectes rusticus and Pacifastacus leniusculus in
large lakes of North America’. Hydrobiologia 800: 173–185.
Maitland, P.S. (1996). ‘The North American signal crayfish, Pacifastacus
leniuculus (Dana), established in the wild in Scotland’. Aquatic
Conservation: Marine and Freshwater Ecosystems 6: 107–110.
O’Brien, C.D., Hall, J.E., O’Brien, C.T., Baum, D. and Ballantyne,
L. (2013). Impact of a natural pyrethrin biocide on two amphibians,
common toad Bufo bufo and palmate newt Lissotriton helveticus, in
Highland, UK. Conservation Evidence 10: 70–72.
O’Reilly, S. (2015). ‘Assessing the Toxicity of Biocides on the North
American Signal Crayfish Pacifastacus leniusculus (Dana) to Aid
Eradication’. MSc thesis. Glasgow: University of Glasgow.
Peay, S., Hiley, P.D., Collen, P. and Martin, I. (2006). ‘Biocide treatment
of ponds in Scotland to eradicate signal crayfish’. Bulletin Francais de
la Peche et de la Pisciculture 380–381: 1363–1379.
Sandodden, R. and Johnsen, S.I. (2010). ‘Eradication of introduced signal
crayfish Pacifastacus leniusculus using the pharmaceutical BETAMAX
VET’. Aquatic Invasions 5: 75–81.
Simberloff, D., Martin, J.-L., Genovesi, P., Maris, V., Wardle, D.A.,
Aronson, J., Courchamp, F., Galil, B., Garcia-Berthou, E., Pascal,
M., Pysek, P., Sousa, R., Tabacchi, E. and Vila, M. (2013). ‘Impacts
of biological invasions: What's what and the way forward’. Trends in
Ecology and Evolution 28: 58–66.
H. Bardal
Bardal, H. Small- and large-scale eradication of invasive fish and fish parasites in freshwater systems in Norway
Small- and large-scale eradication of invasive fish and fish parasites
in freshwater systems in Norway
H. Bardal
Norwegian Veterinary Institute, Pb. 750 Sentrum, NO-0106 Oslo, Norway. <helge.bardal@vetinst.no>.
Abstract In July 2016, the European Union adopted a list of invasive alien species of concern, and at present there
are two freshwater fish species on the list. Member states are obliged to prevent further spread and to perform rapid
eradication when problem species are discovered at new sites, but continental EU member states have limited experience
with eradication of fish. Eradications are more likely to succeed if the invasive species is confined to insular habitats.
Freshwater invasives can be regarded as island invasives, since their habitats have boundaries against shorelines, saline
waters, waterfalls and dams, and these boundaries make eradications possible. CFT Legumine® containing rotenone
is the only legal piscicide in the EU, and Norway has used CFT Legumine® in eradication efforts for many years.
Species that have been introduced outside their native range and have been successfully eradicated include minnow
(Phoxinus phoxinus), roach (Rutilus rutilus), pike (Esox lucius), common whitefish (Coregonus lavaretus), and the
salmon parasite Gyrodactylus salaris. This manuscript summarises the eradication efforts of invasive fish species and
fish parasite species during the last two decades in Norway, covering eradications from such diverse habitats as small
ponds, lakes, marshlands, small streams and large rivers. An estimated £100 million has been spent in the Gyrodactylus
salaris eradication programme. Costs of invasive fish eradications are given, ranging from less than £10,000 to more than
£200,000. There are no known invasive fish eradication failures in Norway in the last 20 years. A summary of the efforts
in Norway can be an aid for planning control and eradication measures of invasive fish species in other countries.
Keywords: CFT Legumine, fish control, Gyrodactylus salaris, IAS, rotenone
INTRODUCTION
Alien invasive fish species are a global problem
(Gozlan, et al., 2010). Eradication is more likely to succeed
if the invasive species is confined to insular habitats.
Freshwater invasives can be regarded as island invasives,
since their habitats have boundaries against shorelines,
saline waters, waterfalls and dams, and these boundaries
make eradications possible.
The EU requires member states to rapidly implement
measures, including eradications, against invasive alien
species. In July 2016 the EU adopted a list of invasive
alien species of European Union concern that requires
control or eradication (<http://data.europa.eu/eli/reg_
impl/2016/1141/oj>). This list includes two fish species,
topmouth gudgeon (Pseudorasboras parva) and Amur
sleeper (Perccottus glenii), and new species can be added.
Transfer of knowledge is therefore essential as many EU
countries have little experience with such operations. In
Europe, successful eradications against invasive freshwater
fish have been done in Spain (Fernandez-Delgado, 2009),
England (Britton, et al., 2010) and Norway.
The Norwegian Veterinary Institute has extensive
knowledge of fish eradication through their work on behalf
of the Norwegian Environment Agency. A simplified
way to look at historic immigration routes for freshwater
species to Norway is that all indigenous freshwater fish
species can be found in south-eastern Norway, while
the rest of the country has very few indigenous species,
making most south-eastern species, e.g. all cyprinids and
pike (Esox lucius), domestic exotic in other parts of the
country (Huitfeldt-Kaas,1918). Exotic invasive fish are
North-American salmonids, imported for aquaculture
and improvements of wild stocks, and cyprinids from the
European continent (Hesthagen & Sandlund, 2016). One
of the most severe threats to an indigenous fish species
was the introduction of the salmon fluke Gyrodactylus
salaris, which in a worst-case scenario could lead to local
extinction of Atlantic salmon (Salmo salar) populations.
Norwegian authorities have committed to eradicate the
salmon fluke from Norwegian rivers, and the Norwegian
Veterinary Institute is in charge of the project planning
and eradication campaigns. The experiences from these
campaigns against G. salaris are used in other operations
against invasive freshwater fish species.
The piscicide rotenone has been used for fish control
and eradication for more than 70 years (McClay, 2000).
Rotenone is a natural product isolated from roots of
tropical plants in the pea family Leguminosae, and it is
highly toxic to fish (Ling, 2003). The rotenone product
used in Europe is CFT Legumine® (CFT L). It is the
only piscicide currently under assessment of the Biocidal
Product Regulation (BPR, Regulation (EU) 528/2012), and
thus the only piscicide legal for use in Europe. The effect
of rotenone on non-target organisms has been extensively
studied (Ling, 2003; Vinson, et al., 2010; Finlayson, et al.,
2010a; Dalu, et al., 2015) and, even if some invertebrate
taxa are very sensitive, the general findings from Norway
are that most taxa recolonise treated areas within a year
(Fjellheim, 2004; Kjærstad, et al., 2015).
Two different solutions of CFT L have been used in
the described treatments. The first CFT L formula used
contained 2.5% rotenone and the synergist piperonylbutoxid
(PBO). Since PBO did not have the desired synergic effect
(Finlayson, et al., 2010a), the manufacturer made a change
in the product in 2012. The new product omitted PBO
and increased rotenone content to 3.3%. As of 2013, all
treatments described have used the 3.3% solution.
The objective of this manuscript is to present all
rotenone treatments against invasive freshwater fish
species in Norway the last 20 years, and a short summary of
the still ongoing eradication campaign against the salmon
fluke. None of the invasive fish species eradications are
previously published. Only lake volumes are described
in detail, but the treatment area also includes adjacent
streams, pools and marshlands, to ensure that no fish
survives in temporary locations. The work of treating these
surroundings varies depending on the site, but the amount of
CFT L used in these areas is small compared to the amount
used in lakes. A map is included for the geographical
location of the invasive fish species eradications (Fig. 1).
Costs are included to give an idea about the cost of invasive
fish eradications (Table 1). The following descriptions can
be an aid for planning control and eradication measures of
invasive fish species in the EU and for other stakeholders.
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 447–451. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
447
Island invasives: scaling up to meet the challenge. Ch 2F Other taxa: Aquatic
Fig. 1 Geographical location of invasive fish eradications.
1) Hardangervidda, 2) Sør-Fron, 3) Lake Ålmotjønna, 4)
Lake Alsettjønna, 5) Lake Lille Mortetjern, 6) Telemark
Canal, 7) Lake Vikerauntjønna, 8) Lake Klokkartjønna, 9)
Gäddede (in Sweden) 10) Bymarka.
OVERVIEW OF ERADICATION OF INVASIVE
FISH AND FISH PARASITES IN FRESHWATER
SYSTEMS IN NORWAY
Salmon fluke (Gyrodactylus salaris) in rivers and lakes
in Norway
The salmon fluke G. salaris is a freshwater salmon
ectoparasite indigenous to the Baltic region, and an
exotic invasive species in Norway that was first detected
in the 1970s. G. salaris is one of the most severe threats
against Norwegian Atlantic salmon (Anon., 2016), It has
been introduced via fish transports from Sweden, and
distributed in Norwegian rivers through stocking from
salmon hatcheries. Norwegian Atlantic salmon populations
are highly susceptible to the parasite, with up to 95 %
mortality for salmon fry and parr (Johnsen & Jensen,
1991). Rainbow trout (Oncorhynchus mykiss) and Arctic
char (Salvelinus alpinus) can also host the salmon fluke.
It has been found in 50 rivers in Norway, and rotenone
treatments aim to eradicate the Atlantic salmon from
infected river systems since the parasite cannot survive
without its host. The salinity of the fjords acts as a barrier for
dispersion of the salmon fluke, providing defined borders
for the treatment area (Soleng & Bakke, 1997). Extensive
operations to preserve and re-establish local strains of
Atlantic salmon and sea trout (Salmo trutta) are performed
before, during and after eradications (O’Reilly & Doyle,
2007). Forty-three rivers throughout Norway have been
treated and the salmon fluke has so far been successfully
eradicated from 31 rivers. Meanwhile, 12 rivers are still
under post-treatment surveillance awaiting confirmation
of eradication. The rivers differ in size but eradication
campaigns have included 42 km long rivers in rugged
terrain (Sandodden, et al., 2018) and 10 rivers in the same
fjord system in the Vefsna region, where River Vefsna had
a discharge of 200 m3/s on the day of treatment (Stensli
& Bardal, 2014). Also, in the Vefsna region, the salmon
fluke infested Arctic char in three lakes. In total these lakes
covered more than 18 km2, two of them 65 m deep. The
Vefsna region was one of the largest rotenone treatments
ever performed, for both lake and riverine systems (Stensli
& Bardal, 2014). Confirmation of eradication was attained
in the rivers of the Vefsna region in 2017. The lakes in the
same region still await eradication confirmation because of
a different procedure for eradication confirmation.
Minnow (Phoxinus phoxinus) in Hardangervidda
National Park
The minnow is indigenous to south-eastern Norway,
but has been introduced to most parts of the country. It
is believed that minnow has been spread through the use
of live bait and also accidentally released, mistaken as
brown trout (Salmo trutta) fry. Minnow have been present
at Hardangervidda for decades. Minnows can multiply to
high numbers and food competition has had a negative
impact on local fish stocks and birdlife. Minnows are
not present in western watercourses at Hardangervidda
National Park, a high altitude, tree-less plateau in
southwest Norway, but were found up to the water divide
in several places. The risk of further dispersion across
Table 1 Year of treatment, location and target species, with approximate volume of lakes, litres of CFT L used
and approximate cost of treatment.
Year
Location
Target species
1999
2005
2008
2008
2009
2012
2013
2014
2014
2015
2016
2016
Hardangervidda
Sør-Fron
Lake Ålmotjønna
Lake Alsettjønna
Lake Lille Mortetjern
Telemark Canal
Hardangervidda
Telemark Canal
Lake Vikerauntjønna
Lake Klokkartjønna
Gäddede
Bymarka
Minnow
Rainbow trout
Roach
Common whitefish
Roach
Pike
Minnow
Pike
Roach
Lake trout
Several species
Roach
Volume
(1,000 m3)
5.6
120
84
2.9
640
188
675
2,500
CFT L volume
used (l)a
137
7.5
180
90
4.5
100
805
22
293
670
8
4,000
a) CFT L used also includes CFT L in streams, pools and marshlands surrounding the lakes.
448
Cost (£1,000)
30–50
<10
10
10
10
30–50
>100
10
10–30
30–50
10
>200
Bardal: Eradication of invasive fish and fish parasites
the water divide was regarded as imminent. A successful
treatment was conducted in 1999–2000 in the area around
Stigstuv (Tønset & Bakkeli, 2000), which is set at the
east-west water divide at Hardangervidda National Park.
Fish barriers were built to create a buffer zone towards
the water divide. Minnows were found in the buffer zone
again 10 years later, most likely because the barriers were
not working properly. In flooding periods the water level
downstream from the barrier could rise and thus level out
the height difference. The barriers were adjusted in 2013
prior to a new treatment of the Stigstuv area. The treatment
area comprised 40 ponds/small lakes, streams and marshes
within an area of 2 km2. Total water coverage of the ponds/
small lakes was 140,000 m2, with depths of up to 4 m and
average depths of 0.5–1 m. The treatment was performed
during four days in August 2013 by 16 people. Target dose
was 1 ppm CFT L, and a total of 225 l of CFT L was used.
In addition, Lake Hætjørna, another site on
Hardangervidda, also was treated due to minnow invasion.
Two new barriers had been built, one in the inlet and one in
the outlet, making the lake a buffer zone without minnows.
Twelve people treated the lake, ponds and marshes in the
surrounding area in two days, just prior to the treatment at
Stigstuv. Lake Hætjørna covers an area of 0.2 km2, with
an estimated mean depth of 2.5 m and maximum depth 8
m. The target dose was 1 ppm CFT L, and a total of 580 l
of CFT L was used. No minnows were caught in hoop net
surveys in 2014 and 2016. An environmental DNA survey
found no traces of minnows at either site in 2016 (Fossøy,
et al., 2017).
Rainbow trout (Oncorhynchus mykiss) in Sør-Fron
Sør-Fron is a municipality in Oppland County. Rainbow
trout were found in four artificial ponds at a farm in SørFron. Rainbow trout are indigenous to North-America, and
have been exported worldwide for angling and fish farming.
In Sør-Fron it had been released in these ponds, and netting
in October 2000 confirmed their existence. Rainbow
trout can host the salmon fluke and thus act as a vector
for the parasite. There was risk of escape from the ponds
during large floods in the nearby River Lågen. The ponds
were treated with rotenone in October 2005. The volumes
ranged from 350 to 3,750 m3, and two people completed
the job in one day. The target dose was 1 ppm CFT L, and
a total of 7.5 l CFT L was used. On arrival, all ponds were
covered with ice. Ice cover was broken before adding CFT
L, and the dosage was increased slightly to compensate for
the low water temperatures. Prior to treatment the number
of fish was reduced through netting. There has been no
programme for eradication confirmation, but there have
been no later reports of rainbow trout in the ponds.
Roach (Rutilus rutilus) in Lake Ålmotjønna
Lake Ålmotjønna is situated in Rissa municipality in
Sør-Trøndelag County. The roach is indigenous to southeastern Norway, but alien to the Trøndelag region. It is
believed that roach were released by anglers. Roach were
discovered in Lake Ålmotjønna in the summer of 2007.
The purpose of the treatment was to prevent further spread
downstream to the large Lake Storvatnet, which could
lead to a permanent foothold for roach in the region. The
volume of Lake Ålmotjønna was 120,000 m3, and average
depth was 5 m. A rotenone treatment was conducted in
August 2008. Two people worked for one day. Target dose
was 1.5 ppm CFT L, and a total of 180 l CFT L was used.
Only two dead roach were found post-treatment. Fish scale
analysis revealed that the roach had been introduced at
least five years prior to treatment, and apparently had not
reproduced. There has been no programme for eradication
confirmation, but there have been no later reports of roach
in the lake.
Common whitefish (Coregonus lavaretus) in Lake
Alsettjønna
Lake Alsettjønna is situated in Selbu municipality
in Sør-Trøndelag County. The common whitefish
is indigenous in south-eastern Norway and eastern
watersheds, but alien to the region. Common whitefish
was released in Lake Alsettjønna around 1875 as part of a
wedding gift. The purpose of the treatment was to prevent
further spread to the larger Lake Selbusjøen, which could
lead to a permanent foothold for common whitefish in
the region. Common whitefish is a food competitor of
the indigenous brown trout and Arctic char and can be
more effective in exploitation of food sources. It also has
a large capacity for propagation. Increased rainfall in the
catchment could, in the future, make the stream from Lake
Alsettjønna habitable for common whitefish in flooding
periods, leading it to spread to Lake Selbusjøen. A rotenone
treatment was conducted in Lake Alsettjønna in August
2008. The volume was 84,000 m3, and average depth was 4
m. Two people worked for one day. Target dose was 1 ppm
CFT L, and a total of 90 l CFT L was used. There has been
no programme for eradication confirmation, but there have
been no later reports of common whitefish in the lake.
Roach (Rutilus rutilus) in Lake Lille Mortetjern
Lake Lille Mortetjern is situated in Nittedal municipality
in Akershus County. The roach is indigenous to southeastern Norway, where Lake Lille Mortetjern is situated,
but fish had not been recorded in this lake before, making it
ideal for amphibians. Roach are present in a neighbouring
lake in walking distance, so suspicion is that it has been
carried from there and released into Lille Mortetjern.
Roach were discovered here in 2007. The lake is known for
its rich population of amphibians. The endangered smooth
newt (Lissotriton vulgaris), great crested newt (Triturus
cristatus), and moor frog (Rana arvalis) can be found
here. Since the discovery of roach in 2007, the population
of amphibian larvae dwindled to a minimum due to roach
predation (Kooij & Redford, 2012), and lack of recruitment
threatened the long-term survival of the amphibians. For the
first time in Norway, a rotenone treatment was conducted
to benefit endangered amphibians. The lake is small,
only 2,880 m3, and rotenone treatment was performed in
September 2009. Two people completed the treatment in
one day. The target dose was 1.5 ppm CFT L, and a total of
4.5 l of CFT L was used. No mortality of amphibians was
observed during treatment. The following spring, newts
and frogs reproduced in high numbers (Kooij & Redford,
2012). Eradication of introduced fish in amphibian habitats
can be done effectively with rotenone with apparently
few negative effects on the amphibian population. It is
recommended that treatments be carried out in the autumn
when most adult amphibians and metamorphosed larvae
have left their breeding habitat and water temperatures are
still high enough for rapid rotenone degradation. No roach
were detected in biodiversity surveys after the treatment.
Pike (Esox lucius) in the Telemark Canal
The Telemark Canal connects the coast of Telemark
County with the inland by means of eight sluice stations
on a stretch of 105 km. The pike is indigenous to southeastern Norway but is alien to the Telemark region. Pike
were released in the lower parts of the watercourse about
200 years ago. Pike are a voracious predator with the
potential to severely decimate indigenous populations of
fish. Over the last century, pike have spread upstream.
Further upstream are large lakes with populations of
large brown trout and river pearl mussel (Margaritafera
margaritafera) that could be severely affected by invasive
pike. Pike were found between Kjeldal and Hogga sluice,
Hogga being the critical last sluice before entering the large
449
Island invasives: scaling up to meet the challenge. Ch 2F Other taxa: Aquatic
lake system. This led to permission for rotenone treatment
between Hogga and Kjeldal sluice, and the building of an
electric fish barrier in the side canal leading up to Kjeldal
sluice to prevent pike from being sluiced upstream together
with boat traffic. The goal was to stop the pike at Kjeldal
sluice, creating a pike free buffer zone up to Hogga sluice.
A rotenone treatment was carried out between Hogga and
Kjeldal sluice in October 2011, a stretch of about 1 km,
to eradicate pike and restore the buffer zone. The river
segment between the sluices was drained, and five people
treated the remaining pools in one day. The target dose was
1 ppm CFT L, and 100 l of CFT L was used. An electric fish
barrier was established in 2012, at the side canal leading up
to Kjeldal sluice, to stop further spread.
The Norwegian Veterinary Institute conducted a new
treatment at Kjeldal sluice in April 2014, this time only
in the side sluice canal downstream of the area treated
in 2011. The electric fish barrier in the side canal had
been shut down during the winter season in 2013 due to
maintenance work. Therefore, it was necessary to prevent
the pike that had passed the non-functional electric barrier
during winter from entering the previously treated area
upstream of Kjeldal sluice before the boat sluices were
opened for the season start. The side canal was 220 m
long, 3 m deep and 17 m wide. One person did the job in
one day, and a total of 22 l of CFT L was used. The target
dose was increased to 1.5 ppm CFT L to compensate for
water leaking through the sluice gates. Netting has been
conducted in the rotenone-treated areas over several years,
and no pike have been found.
Roach in Lake Vikerauntjønna
Lake Vikerauntjønna is situated in Trondheim
municipality in Sør-Trøndelag County. The roach is
indigenous to south-eastern Norway, but alien to this
region. It is believed that roach had been released, and the
source was other lakes with an alien population of roach in
the same municipality, in Bymarka. A dense population of
roach was detected in Lake Vikerauntjønna in 2013. It is
located only 250 m from Trondheim municipality’s main
potable water source, Lake Jonsvatnet. There was a concern
that roach could adversely affect water quality. The two
lakes belong to separate catchments, but the risk of further
spread to Lake Jonsvatnet was considered to be high due to
the small distance between the lakes. Rotenone treatment
was considered as the only measure that would eradicate
the roach. Lake Vikerauntjønna covers an area of 0.04 km2
with a water volume of 188,000 m3 and maximum depth
of 17 m. Treatment was conducted in September 2014
by five people in one day. The target dose was 1.5 ppm
CFT L, and a total of 293 l of CFT L was used. Fish scale
analysis revealed that the roach had been introduced for the
first time around 2007, and possibly new introductions in
following years too. No roach were detected in biodiversity
surveys after the treatment.
Lake trout (Salvelinus namaycush) in Lake
Klokkartjønna
Lake Klokkartjønna is situated in Blåfjella-Skjækerfjella
National Park in Nord-Trøndelag County. The lake trout
is indigenous to North-America but has been imported
to Scandinavia for fish farming and angling purposes.
The first records of release in Norway are from the 1970s
(Hesthagen & Sandlund, 2016). In Lake Klokkartjønna
the introduction could have come from source populations
in Sweden, since lake trout are more common across the
border, but no one knows for sure. A first finding of lake
trout in Lake Klokkartjønna was recorded in the autumn
of 2010. Lake trout are considered to be a threat to natural
habitats, ecosystems, and indigenous fish populations.
The risk of spread downstream to adjacent lakes was
450
considered high, and permission for rotenone treatment
was granted. Lake Klokkartjønna covers an area of 0.14
km2 with an estimated volume of 675,000 m3. Eradication
was performed in July 2015, and eight people participated
over two days. The target dose was 1 ppm CFT L, and a
total of 670 l of CFT L was used. No lake trout have been
found through post-treatment netting.
Several species at hydroelectric power plant in
Gäddede
Gäddede hydroelectric power plant is situated in
Sweden in Strømsund municipality in Jämtland County
close to the Norwegian border. The power plant separates
two lakes with different fish species due to a natural fish
barrier. The upstream lake contained only indigenous
brown trout and Arctic char whilst the lower lake also
hosted pike, common whitefish, burbot (Lota lota), and
rainbow trout. It was unwanted for any of these fish species
to be spread upwards in the waterway. It is not possible
for fish to pass upstream through the power plant, but a
planned maintenance shutdown in 2016 could enable fish
to pass the turbines and later rise up above the dam into
the upper lake. As a precautionary measure, permission
for rotenone treatment in the stagnant ponds on both sides
of hydro power turbines was given. The volumes of the
ponds ranged from 150 to 2,500 m3, and eradication was
performed in June 2016. Two people did the job in one
day. A high dose of 3 ppm CFT L was used to compensate
for fresh water leaking into the isolated ponds. A total of
8 l CFT L was used. There has been no programme for
eradication confirmation. Dead fish were found during
treatment in the ponds.
Roach in seven lakes in Bymarka
Bymarka is on the outskirts of Trondheim city in SørTrøndelag County. The roach is an invasive species in the
region and was released in the 1880s to three small lakes in
the same watercourse. From the 1960s to the 1980s roach
were spread to three neighbouring lakes, and were found
in another four lakes from 1998 to 2013. It was suspected
that the roach population in Bymarka was the source of
spread. When roach were found in Lake Vikerauntjønna,
close to the Trondheim municipality’s main potable water
source, plans for treatment of the lakes in Bymarka were
put forward. The main reasons were a concern for the
roach to adversely affect potable water quality, a wish to
permanently eradicate this blacklisted species from the
region, and to contribute to conservation of natural fish
stocks and biodiversity. Rotenone treatment was considered
to be the only measure that could eradicate roach from
these lakes. Several of the lakes have a dam, and an attempt
to eradicate roach through dewatering in 2004 failed. One
lake was 17,000 m3 and 10 m deep, while the six other
lakes ranged from 412,000 to 615,000 m3, with maximum
depths of 10–17 m. In September 2016, treatment was
performed by a crew of 14 people for four days. A total
of 4,000 l of CFT L was used and, as before, the target
dose was 1.5 ppm CFT L. Populations of invasive pike in
the lakes were eradicated simultaneously. No roach were
detected in a biodiversity survey after the treatment.
DISCUSSION
Rotenone treatments are not without controversy,
but most times invasive fish eradications are welcomed.
The general public’s acceptance of rotenone treatments
in Norway might be a result of the absence of failure,
thus strengthening the understanding for rotenone as a
necessary and effective tool in the fight against invasive
freshwater fish. The description of rotenone treatments
does not include method, but relevant method can be
Bardal: Eradication of invasive fish and fish parasites
found in Sandodden, et al. (2018). A standard operating
procedure for the use of rotenone in fish management is
given by Finlayson, et al. (2010b). For rotenone analysis,
an on-site determination of rotenone has been developed by
the Norwegian Veterinary Institute (Sandvik, et al., 2018).
The Norwegian Environment Agency is in the process
of writing an action plan, which will identify and prioritise
measures against invasive freshwater fish. This will lift
the coordination of possible eradication measures from
county level to national level, making top prioritised
measures easier to identify. Forthcoming eradication
projects are mostly for domestic invasive pike. The G.
salaris eradication campaign will continue, and is now at
an intermediate planning stage with the next eradication, at
the earliest, in 2022.
Costs of treatments
Costs are not easy to describe uniformly. Eradication
projects have had different levels of participation from the
County Governor, and work hours are usually the main
expense in these smaller eradications. The cost consists
of the Norwegian Veterinary Institute’s planning and
preparations and expenses with treatments, including hired
crew. The cost does not include pre- and post-treatment
biodiversity surveys, cost of CFT L, and County Manager
expenses. Eradication of G. salaris is not included in the
table, but the cost of the eradication campaign so far is
estimated to be about £100 million.
Eradication confirmation
The G. salaris eradication campaign includes a
surveillance programme for eradication confirmation, but
no parallel surveillance programme exists for invasive
freshwater fish. Eradication confirmations are based on the
absence of new detections by biodiversity surveys, local
netting and angling. Successful restocking of indigenous
fish also indicates the absence of the introduced species.
Net trapping and environmental DNA surveys are other
possible ways to document the outcome of a treatment but
there is, at present, no national set of rules for eradication
confirmation. However, there are no examples from
Norway, during the past 20 years, of failed rotenone-based
eradication attempts against invasive freshwater fish.
This may be because all eradications are assigned for
planning and execution to a national competence group
for rotenone treatment, which gives continuity-based
experience and knowledge. Secondly, smaller lentic
waters are less complicated treatments due to longer
time for adequate mixing of rotenone and thus ensuring
lethal concentration in all parts of the lake, which should
leave the target fish no opportunities to accidently avoid
lethal exposure. Large-scale lotic waters systems are also
possible to succeed in, proven by the G. salaris eradication
campaign.
ACKNOWLEDGEMENTS
I wish to thank all my colleagues who have participated
in the described treatments, and the Norwegian
Environment Agency for their role in fighting invasive
species. I appreciated the constructive comments on the
manuscript given by two anonymous referees.
REFERENCES
Anon. (2016). Status for Norske Laksebestander i 2016. (Status of
Norwegian Salmon Stocks in 2016). Rapport fra Vitenskapelig
råd for lakseforvaltning no. 9. Trondheim: Vitenskapelig Råd for
Lakseforvaltning.
Britton, J.R., Davies, G.D. and Brazier, M. (2010). ‘Towards the
successful control of the invasive Pseudorasbora parva in the UK’.
Biological Invasions 12: 125.
Dalu, T., Wasserman, R.J., Jordaan, M., Froneman, W.P. and Weyl, O.L.F.
(2015). ‘An assessment of the effect of rotenone on selected non-target
aquatic fauna’. PLOS ONE 10(11): e0142140.
Fernandez-Delgado, C. (2009). Restoration of a Waterfowl Community
by Eradication of the Invasive Fish Species Cyprinus carpio, with
Rotenone. Aphanius, investigacion peces. Cordoba (Spain): University
of Cordoba.
Fjellheim, A. (2004). Effekt av Rotenonbehandling på Bunndyrsamfunnene
i et Område ved Stigstu, Hardangervidda. (Effect of rotenone treatment
on benthic invertebrates in an area at Stigstu, Hardangervidda). Rapport
no. 122. Bergen: LFI-Unifob, Universitetet i Bergen.
Finlayson, B., Somer, W.L. and Vinson, M.R. (2010a). ‘Rotenone toxicity
to rainbow trout and several mountain stream insects’. North American
Journal of Fisheries Management 30: 102–111.
Finlayson, B., Schnick, R., Skaar, D., Anderson, J., Demong, L., Duffield,
D., Horton, W. and Steinkjer, J. (2010b). Planning and Standard
Operating Procedures for the Use of Rotenone in Fish Management—
Rotenone SOP Manual. Bethesda, Maryland: American Fisheries
Society.
Fossøy, F., Dahle, S., Birkeland Eriksen, L., Hagen Spets, M., Karlsson,
S. and Hesthagen, T. (2017). Bruk av Miljø-DNA for Overvåking av
Fremmede Fiskearter – Utvikling av Artsspesifikke Markører for
Gjedde, Mort og Ørekyt. (Use of environmental DNA for monitoring
invasive species – development of species-specific markers for pike,
roach and minnow). Rapport 1299. Trondheim: Norsk Institutt for
Naturforskning.
Gozlan, R.E., Britton, J.R., Cowx, I. and Copp, G.H. (2010). ‘Current
knowledge on non-native freshwater fish introductions’. Journal of Fish
Biology 76: 751–786.
Kooij, J van der and Redford K., (2012). Rotenonbehandling av
Amfibietjern på Holumskog. Overvåking av Dyrelivet. (Rotenone
treatment of an amphibian tarn at Holumskog. Monitoring of fauna).
Nittedal: Nittedal Kommune.
Hesthagen, H. and Sandlund, O.T. (2016). Tiltaksrettet Kartlegging og
Overvåking av Fremmed Ferskvannsfisk – en Tilstandsvurdering av
Spredningen pr. 2016. (Targeted mapping and surveillance of foreign
freshwater fish – an assessment of the spread per 2016). Rapport 1302.
Trondheim: Norsk Institutt for Naturforskning.
Huitfeldt-Kaas, H. (1918). Ferskvannsfiskenes Utbredelse og Indvandring
i Norge med et Tillæg om Krebsen. (Spread and immigration of
freshwater fishes in Norway with an addition on crayfish). Kristiania:
Centraltrykkeriet.
Johnsen, B. O. and Jensen, A. J. (1991). ‘The Gyrodactylus story in
Norway’. Aquaculture 98: 289–302.
Kjærstad, G., Arnekleiv, J. V. and Speed, J. D. M. (2015). ‘Effects of
three consecutive rotenone treatments on the benthic macroinvertebrate
fauna of the River Ogna, Central Norway’. River Research and
Applications 32: 572–582.
Ling, N. (2003). ‘Rotenone: A Review of its Toxicity for Fisheries
management’., Science for Conservation 211, Wellington, New
Zealand: Department of Conservation.
McClay, W. (2000). ‘Rotenone use in North America (1988–1997)’.
Fisheries 25: 15–21.
O’Reilly, P. and Doyle, R. (2007). ‘Live Gene Banking of Endangered
Populations of Atlantic Salmon’. In: E. Verspoor, L. Stradmeyer and
J.L. Neilsen (eds.) The Atlantic Salmon. Genetics, Conservation and
Management, pp. 425–469. Oxford, UK: Blackwell Publishing.
Sandodden, R., Brazier, M., Sandvik, M., Moen, A., Wist, A.N. and
Adolfsen, P. (2018). ‘Eradication of Gyrodactylus salaris infested
Atlantic salmon (Salmo salar) in the Rauma River, Norway, using
rotenone’. Management of Biological Invasions 9(1): 67–77.
Sandvik, M., Waaler, T., Rundberget, T., Adolfsen, P., Bardal, H. and
Sandodden, R. (2018). ‘Fast and accurate on-site determination
of rotenone in water during fish control treatments using liquid
chromatography’. Management of Biological Invasions 9(1): 59–65.
Soleng, A. and Bakke, T.A. (1997). ‘Salinity tolerance of Gyrodactylus
salaris (Platyhelminthes, Monogenea): Laboratory studies’. Canadian
Journal of Fisheries and Aquatic Sciences 54(8): 1837–1864.
Stensli, J.H. and Bardal, H. (2014). ‘Bekjempelse av Gyrodactylus
salaris i Vefsnaregionen’. (Eradication of Gyrodactylus salaris in
the Vefsna region). Veterinærinstituttets rapportserie 2-2014. Oslo:
Veterinærinstituttet.
Tønset, K. and Bakkeli, G. (2000). Bekjempelse av ørekyte (Phoxinus
phoxinus) i Stigstuvområdet i perioden 19.09–25.09.2000. (Eradication
of minnow (Phoxinus phoxinus) in the Stigstuv area in the period 19.09–
25.09.2000 ') Rapport Desember 2000. Trondheim: Veterinærmedisinsk
Oppdragssenter.
Vinson, M.R., Dinger, E.C. and Vinson, D.K. (2010). ‘Piscicides and
invertebrates: After 70 years, does anyone really know?’ Fisheries 35:
61–71
451
M. Cárdenas-Calle, J. Pérez-Correa, P. Martinez,, I. Keith, F. Rivera, M. Cornejo, G. Torres, F. Villamar, R. Zambrano, A. Cárdenas, M. Triviño, L. Troccoli, G. Bigatti, J. Coronel and E. Mora
Cárdenas-Calle, M.; J. Pérez-Correa, P. Martinez,, I. Keith, F. Rivera, M. Cornejo, G. Torres, F. Villamar, R. Zambrano, A. Cárdenas, M. Triviño, L. Troccoli, G. Bigatti, J. Coronel and E. Mora. First
report of marine alien species in mainland Ecuador: threats of invasion in rocky shores
First report of marine alien species in mainland Ecuador:
threats of invasion in rocky shores
M. Cárdenas-Calle1,2, J. Pérez-Correa1,3, P. Martinez,5, I. Keith4, F. Rivera5, M. Cornejo1, G. Torres1,3, F. Villamar1,
R. Zambrano1, A. Cárdenas1, M. Triviño1, L. Troccoli1,6, G. Bigatti3,7, J. Coronel1,8 and E. Mora1
Bioelite S.A. Cdla. Bosques el Salado Mz 301 solar 2B frente a Ciudad Colón, Guayaquil, Ecuador. <bioelitesa@
gmail.com / maritza.cardenasc@ug.edu.ec>. 2Universidad de Guayaquil, Ciudadela Universitaria “Salvador Allende”,
Av. Delta y Av. Kennedy, Guayaquil, Ecuador. 3Universidad de Especialidades Espíritu Santo, Km. 2.5 vía La Puntilla,
Samborondón, Ecuador. 4Charles Darwin Foundation, Av. Charles Darwin s/n, Puerto Ayora, Galápagos, Ecuador.
5
Institto NAZCA de Investigaciones Marinas, Ángel Rojas y Juan León Mera, conjunto Montepiedra II casa 3, San Juan
Alto, Cumbaya, Quito, Ecuador. 6Universidad del Oriente, Av. Gran Mariscal. Edificio Rectorado, Cumaná, Estado
Sucre, Venezuela. 7LARBIM (IBIOMAR-CONICET), Bv. Almirante Brown 2915, Puerto Madryn, Chubut, Argentina.
8
Universidad Agraria del Ecuador, Av. 25 de Julio y Pio Jaramillo, Guayaquil, Ecuador.
1
Abstract Invasive species are of significant concern, especially in mega-diverse countries, because they cause negative
effects such as loss of native biodiversity, ecological alterations, disease spread, and impacts on economic development and
human health. In mainland Ecuador, information on invasive invertebrates in marine ecosystems is scarce. The objective
of this study was to describe and locate the invasive species present in the rocky shores of the intertidal and subtidal
zones along 10 areas (83 sites) covering most of the Ecuadorian coast during 2015–2016. Benthic macroinvertebrates
communities were measured over quadrats located randomly on a 50 m transect positioned parallel to the coast in the
intertidal and subtidal zone, covering an area of 1,860 km2. Six invasive species were recorded: Arthropoda (Amphibalanus
amphitrite), Cnidaria (Pennaria disticha, Carijoa riisei), Bryozoa (Bugula neritina), Rhodophyta (Asparagopsis
taxiformis) and Chlorophyta (Caulerpa racemosa). The areas with highest abundance of invasive species were in Jama
(not a protected area), Marine and Costal Wildlife Reserve Puntilla of Santa Elena and Santa Clara Island Wildlife Refuge
(protected areas). The most abundant species was Carijoa riisei with a relative abundance of up to 80%. It was the most
aggressive of the invasive species registered in the subtidal zone, mainly in northern centre of the Ecuadorian coast. C.
riisei is growing on native coral (Pocillopora spp.) and on sessile macroinvertebrate communities (Pinctada mazatlanica,
Muricea appresa and Aplysina sp.) that are being affected by its invasion. This study must be taken into account by local
and regional government authorities to create public policy programmes of monitoring for surveillance and control of
invasive species. These programmes should focus on integration of socio-economic and ecological effects. They should
be complemented by experimental design and analysis of environmental variables to provide technical information for a
baseline of bio-invasion analysis along the Ecuadorian coast and Galápagos, to avoid the expansion of invasive species
negatively affecting the marine biodiversity of mega-diverse countries such as Ecuador and other countries of South
America.
Keywords: Carijoa riisei, continental Ecuador, Galapagos Islands, invasive species, macroinvertebrates, marine
ecosystems, marine protected areas
INTRODUCTION
Invasive species are a cause of worldwide concern
especially in mega-diverse countries because they can
cause loss of native diversity, ecological alterations,
increases in pests, diseases (Prenter, et al., 2004), impacts
on benthic communities, impacts to the water column
(Darrigran & Damborenea, 2011). Additionally, they can
affect economic development and human health (Lowe,
et al., 2000; Pimentel, et al., 2005). Many species are
transported accidentally through anthropogenic means
breaking geographic barriers that once restricted their
range of expansion (Schüttler & Karez, 2008); they invade
new areas, where they can settle, reproduce, spread and
compete with native species.
Biological invasions, along with climate change, are
key processes that feedback and affect global biodiversity.
Climate change facilitates the dispersal and establishment
of species which aggravates their impacts and makes their
control more difficult, while invasive species can influence
the magnitude of the environmental impacts by altering
the structure and function of ecosystems (Mendoza, et al.,
2014).
At present, there are numerous global and regional
initiatives dedicated to optimising information and
management of invasive alien species, including the
Global Invasive Species Program (GISP), the IUCN-ISSG
Invasive Species Global Information Network on Invasive
Alien Species (GISIN), the Global Invasive Species
programme of The Nature Conservancy (TNC-GISI) and
the Inter-American Invasive Species Information Network
(IABIN-I3N) (Schüttler & Karez, 2008).
On mainland Ecuador, information on invasive
invertebrate species on intertidal rocky shores and subtidal
zones is limited, fragmented and scattered. However,
research on non-native species conducted in the Galapagos
Islands (1,000 km off the coast of mainland Ecuador) has
increased in the last decade, both in the terrestrial and
marine environments (Campbell, et al., 2015). In 2012,
the Charles Darwin Foundation (CDF), in collaboration
with the Galapagos National Park Directory (GNPD), the
Galapagos Biosecurity Agency (ABG), the Ecuadorian
Navy and the Ecuadorian Navy Oceanographic Institute
(INOCAR), initiated the Marine Invasive Species Project
in the Galapagos Marine Reserve (Keith, et al., 2015).
The study of non-native species in Ecuador has mainly
been done in the Galápagos Islands, due to the importance
of this unique ecosystem in the world and the relative lack
of scientific funding on the mainland. In the Galápagos
Marine Reserve (GMR), an initial baseline study produced
a list of seven non-native species in the GMR (Keith, et al.,
2016). The marine invasive species team of the CDF have
continued the research and applied different methodologies
to learn more about non-native species in the GMR and
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
452
up to meet the challenge, pp. 452–457. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Cárdenas-Calle, et al.: Marine alien species in mainland Ecuador
the Eastern Tropical Pacific (ETP) region (I. Keith, pers.
comm.). The objective of this study was to identify invasive
species located in rocky shore habitats of the intertidal and
subtidal zones covering 1,860 km2 of the Ecuadorian coast
during 2015–2016, that could be considered as threats for
Ecuadorian mainland as well as Galapagos vulnerable
ecosystems.
MATERIALS AND METHODS
Study area
Fieldwork was carried out in 10 areas along the
Ecuadorian coast in six protected coastal marine areas
(acronym in Spanish: AMCP) and four non-protected
areas. The study areas ranged from Playa Escondida
(0°49'8.05" N, 80° 0'22.66" W) in the north of Ecuador
in Esmeralda province to Santa Clara Island in the south
of Ecuador (3°11'21.11" S, 80°27'10.21" W) in El Oro
province, covering 1,478 km2 of protected areas and
382 km2 of additional areas on the mainland coast. This
survey included the protected areas (Fig. 1) from the
north of Ecuador in the Galeras San Francisco Marine
Reserve (acronym in Spanish: RMGSF) (Esmeralda
province); Wildlife Refuge and Marine Coastal Pacoche
(Pacoche) and Machalilla National Park (acronym in
Spanish: PNM) (Manabí province); El Pelado Marine
Reserve (acronym in Spanish: REMAPE); and Wildlife
Coastal Marine Reserve Puntilla of Santa Elena (acronym
in Spanish: REMACOPSE) (Santa Elena province) to
Santa Clara Island Wildlife Refuge (Santa Clara) (El Oro
province). The non-protected areas (Fig. 1) were: Jama,
Canoa (Manabí province), Ayampe-La Entrada (between
Manabí and Santa Elena provinces) and Copé (Santa
Elena province). The Ecuadorian coast has an extension
of 2,900 km corresponding to 45% of open coastal and
55% of inner coastal waters (Ayón, 1988). There is a wide
range of geological features along the coast, including
bluffs, barriers and sandplains, estuaries and lagoons, and
engineered shoreline structures (Boothroyd, et al., 1994).
Environmental Ministry (Ministry of Environment, 2014).
The composition and abundance of the macroinvertebrates
present in the rocky shore in the intertidal and subtidal
zones were quantified using a band-transect system parallel
to the coastline.
Data were collected in the intertidal zone following the
standardised protocol from the South American Research
Group on Coastal Ecosystems (SARCE) (SARCE, 2012).
At each site, 10 quadrats (50 × 50 cm each) were randomly
placed and sampled along a 50 m transect positioned
parallel to the waterline in the mid, low and high intertidal
level. A total of thirty replicates was sampled for each site.
The abundance of colonial organisms was estimated by
percent cover and all mobile individuals (>2 cm long) were
counted. Most identification of biota was done in the field,
although occasional problematic specimens were collected
for reference and sent to specialists for identification. For
the subtidal zone, at each site the organisms were separately
estimated in two transect blocks by a diver, one on each side
of 50 m transect line set along a shallow depth (normally
6–8 m). Every transect block encompassed a total reef area
of 50 m × 5 m. The next diver scanned the nearby transect
block by swimming back parallel to the initial transect at a
distance of 5 m from the transect line (Edgard, et al., 2011).
This up and back procedure for two adjacent blocks was
The climate on the coast varies seasonally from dry
season (May to November) to the rainy season (December
to April). The average annual temperature is above 22oC,
with maxima fluctuating between 32–38oC and minima
fluctuating around 15oC (Sonnenholzner, et al., 2013).
Ecuador belongs to the Tropical East Pacific (TEP) region,
with two sub-regions known as Panama Bight Ecoregion
and Guayaquil Ecoregion (Sullivan & Bustamante,
1999; Miloslavich, et al., 2011). The northern half of the
Ecuadorian mainland coast corresponding to the Panama
Bight Ecoregion extends from Azuero Peninsula of
Panamá to Caráquez Bay. It is characterised as a tropical
zone, covered mostly by mangroves and dense rainforest
vegetation (Miloslavich, et al., 2011), with >2,000 mm/yr
of rainfall and without ecologically dry months through
the year (Sonnenholzner, et al., 2013). The southern
Ecuadorian coast, falling within the Guayaquil Ecoregion,
extends from Caráquez Bay to Illescas Peninsula in the
north of Perú and is characterised by a drier climate with
<100 mm/yr of rainfall (Miloslavich, et al., 2011).
Survey
A total of 83 sites were sampled from February
2015 to February 2016 along the four coastal provinces
of Ecuador (Esmeraldas, Manabí, Santa Elena and El
Oro). These sites were established considering aspects
such as representativeness of ecosystems; areas with
greater and lesser anthropogenic intervention, biological
processes (reproduction hotspots, feeding areas, seabirds
and sea turtle nesting sites); and sensitive habitats or
areas of great ecological importance according to the
requirements established in the terms of reference of the
Fig. 1 Study area and location of the sampling sites on the
Ecuadorian coast.
453
Island invasives: scaling up to meet the challenge. Ch 2F Other taxa: Aquatic
repeated along the shallower depth contour, generating a
duplicate transect block data at each site. Sessile organisms
were estimated by percent coverage of different taxa and
grouped in substratum classes (crustaceans, cnidarians,
sponges, ascidians, bryozoans, hydroids) within the
transect lines. The cover was generally recorded by divers
within 10 quadrats (0.5 × 0.5 m), placed sequentially every
5 m along the 50 m transect, and mobile organisms were
counted along each quadrat. Digital photo quadrats were
taken during the field work. We summed counts across all
quadrats to create site totals.
Data analysis
To explain the biological assemblage, an X sites by Y
species matrix of abundances was built to perform a nonmetric multidimensional scaling (n-MDS) and cluster
analysis to visualise the similarity of studied areas. For both
analyses a similarity matrix was generated using the BrayCurtis index on the fourth root transformed data to remove
the weight of the dominant species (Clarke & Gorley,
2006). Further bubbled MDS analyses were performed to
visually establish the differences among the abundances
of Carijoa riisei between zones, using the statistical
package Plymouth Routines in Multivariate Ecological
Research (Primer). In order to determine the difference
of organism abundance by province, a nonparametric
ANOVA was performed, after the assumptions were not
fulfilled, using the Kruskal Wallis test. In addition, to
determine the difference in abundances between protected
and unprotected zones, we applied the ANOSIM test using
PRIMER V6 (Clarke & Gorley, 2006).
RESULTS
A total of six alien invasive species from five phyla were
recorded: Cnidaria (Pennaria disticha, Carijoa riisei),
Bryozoa (Bugula neritina), Arthropoda (Amphibalanus
amphitrite), Rhodophyta (Asparagopsis taxiformis) and
Chlorophyta (Caulerpa racemosa). Assemblages were
numerically dominated by cnidarians. The most abundant
species was Carijoa riisei (Table 1; Fig. 2). Invasive
species were recorded at 24 sites (14 sites in the subtidal
zone and ten in the intertidal zone). In the subtidal zone,
the area with the highest presence of invasive species was
the RMGSF in the north of Ecuador (Esmeralda province)
while in the intertidal zone it was Punta Carnero site in
the REMACOPSE, south-central part of the coast in Santa
Elena province. (Table 1; Fig. 3).
The n-MDS of invertebrate invasive species abundance
showed four groups with major similarity (60%), one group
formed by Jama, REMAPE and RMGSF, the second group
clustered the sites of REMAPE (south-central coast); the
third group formed REMACOPSE, Ayampe, Copé and
Santa Clara (central and south-central coast) and the last
one grouped by REMACOPSE, Pacoche, Santa Clara,
Ayampe and Canoa (Fig. 4).
Amphibalanus amphitrite, Pennaria distincha and
Carijoa riisei were the invasive species with greatest
Distribution maps of species were prepared using
the information collected from the fieldwork. These
maps represent the relative abundance with a percent of
coverage/m2 on every site for inter-tidal and subtidal zones,
with scales ranging between 0.1–50%, indicating a spatial
approximation of alien invasive species location and
coverage. Besides the status, invasiveness of each species
was established using international databases such as: the
IUCN list of 100 most harmful invasive alien species in
the world (Lowe, et al., 2000); Global Invasive Species
Database, ISSG (IUCN/SSC, 2014); and Invasive Species
Compendium (<www.cabi.org/isc>).
Fig. 2 Alien species found in the survey: a) Asparagopsis
taxiformis, b) Amphibalanus amphitrite c) Caulerpa
racemosa, and d) Carijoa riisei growing on the bivalve
Pinctada mazatlanica.
Fig. 3 Relative abundance and distribution of invasive
species along the Ecuadorian coast during 2015–2016.
454
Cárdenas-Calle, et al.: Marine alien species in mainland Ecuador
DISCUSSION
occurrence. Of these three, C. riisei was most abundant
(Fig. 5) in the non-protected area located in the central
coast of Ecuador (Jama) (Table 1). However, it was also
recorded in the north zone of the Galera San Francisco
Marine Reserve (Punta Alta, Piedra de Quingue) and in
the south-central coast at El Pelado Marine Reserve (La
Pared).
This is the first report investigating the presence of
invasive species along the Ecuadorian coast, including
marine protected areas and unprotected areas, covering the
coast from north to south of the country and two ecoregions
in four distinct provinces. There are four species classified
as macroinvertebrate invasive species worldwide, of which
the majority are the cnidarians, mainly the Anthozoa class.
Although the invasive species recorded are not listed in
the 100 world's worst invasive alien species according to
IUCN (Lowe, et al., 2000), two species (Carijoa riisei and
Bugula neritina,) are listed in the Global Invasive Species
Database (ISSG) and four species (Carijoa riisei, Bugula
neritina, Pennaria disticha and Amphibalanus amphitrite)
are registered by the Global Register of Introduced and
Invasive Species (GRIIS).
Statistically, no significant differences were found
between the abundance of invasive species by provinces
(global R=0.08, p>0.001) or by protected and unprotected
zones (global R=-0.06, p>0.001).
The MAP’s that presented the greatest number of
invasive species were REMACOPSE (four species) and
Ayampe (three species), followed by Jama and REMAPE
(less than three species. Galeras San Francisco, Canoa,
Pacoche and Copé recorded low benthic numbers of
invasive species (Table 1).
Canoa
Wildlife Refuge and
Marine Coastal Pacoche
Ayampe – La Entrada
Bajo Copé
SANTA ELENA
El Pelado Marine
Reserve
(REMAPE)
Puntilla de Santa Elena
Marine and Coastal
Wildlife Reserve
(REMACOPSE)
EL ORO
Santa Clara Island
Wildlife Refuge
Pennaria
distincha
Bugula
neritina
MANABÍ
Jama
Carijoa
riisei
ESMERALDAS
Galeras San Francisco
Marine Reserve
Aspargopsis
taxiformis
Area
Caulerpa
racemosa
PROVINCE
Amphibalanus
amphitrite
Table 1 Invasive species recorded by provinces, areas and sites on the Ecuadorian coast, including abundance
(coverage percentage) in the subtidal and intertidal zones.
Punta Alta
-
-
-
-
1.37
-
Piedra de Quingue
-
-
-
-
11.5
-
Vaca Brava 1
Punta Ballena*
Bajo Londres
Cabo Pasado*
Liguiqui*
0.11
3.28
0.96
-
-
-
20.25
44.57
-
-
Los Ahorcados 1
La Entrada*
Seco Manta
Bajo Fer 3
0.03
-
-
-
0.49
4.68
8.86
-
0.12
-
La Pared
Bajo 40
Corales
Guarro
Bajo Ballena
Chocolatera*
Loberia*
Punta Carnero*
Anconcito*
0.11
7.8
16.01
0.5
0.03
0.22
-
0.37
4.82
-
1.12
0.25
-
4.44
-
5.31
0.17
-
-
0.6
1.68
-
-
18.64
3.67
-
-
Sites
Sur*
Norte*
Sitio 2
Sitio 3
* Sites with results of intertidal zones.
455
Island invasives: scaling up to meet the challenge. Ch 2F Other taxa: Aquatic
Fig. 4 Non-metric multidimensional scaling ordination,
showing relative abundance of marine invasive species
registered along the Ecuadorian coast during the period
2015–2016.
Fig. 5 2D bubble MDS configuration showing relative
abundance of Carijoa riisei.
Carijoa riisei showed a greater abundance in the central
zone of the Ecuadorian coast, mainly in Jama. This species
has increased its colonisation in some areas of the El Pelado
Marine Reserve in two years (2013–2015) (Cárdenas-Calle
& Triviño, 2014). The invasion of C. riisei to new sites is
probably caused by marine currents and maritime traffic.
The invasive growth of C. riisei was noted among colonies
of Pocilloporidae corals, Pinctada mazatlanica, Muricea
appresa and Aplysina sp., confirming the imminent threat
of this species to the sessile biota of the marine protected
areas (Martínez, 2013). This species has an extensive
geographic distribution in the Pacific from the Philippines,
Indonesia, Australia, and Thailand, South Atlantic (Silva,
et al., 2011) and Caribbean region (Kahng & Grigg,
2005; Kahng, et al., 2008;) with a variety of reproductive
strategies (Barbosa, et al., 2014) including sexual and
asexual reproduction, growing in different habitats, but
preferring shallow areas.
invasions. For these reasons it is necessary to begin an
alliance between national and international academics and
environmental authorities (Ministry of Environment) in
Ecuador to develop a strategy for surveillance and research
on the ecological effects of invasive species in the coastal
zones. With Carijoa riisei it is necessary to quantify
mortality and replacement of existing coral communities
in Ecuador, because this information is currently unknown,
as is the habitat and biota preferences for colonisation. It is
important to know its distribution, its ecological effects on
native fauna, and its preferences (habitat, substrates, depths
and environmental variables) to allow the establishment of
substantial management actions to avoid its dispersion to
other sensitive areas, such as the Galapagos Islands where
it is still absent.
Carijoa riisei has caused great impacts and damage
to coral areas in Hawaii (Barbosa, et al., 2014) where it
is currently considered a pest and has affected over 70%
of the colonies of black corals Antipathes dichotoma and
A. grandis (Global Invasive Species Database, 2017). It is
considered a common invasive species from Florida (USA)
to Santa Catarina (Brazil), displacing native species. It is
now known to monopolise benthic surfaces under optimal
conditions for its growth, from the intertidal zone to depths
of >100 m (Venkataraman, et al., 2016). C. riisei competes
successfully over black coral and invertebrates (Kahng
& Grigg, 2005) and is dispersed through marine vectors
(Grigg, 2003), and it is reported as a major biofouler in the
Atlantic region (Concepcion, et al., 2010).
The rapid growth of the C. riisei colonies and their
widespread dispersion in coral ecosystems has begun to
generate great concern worldwide for being considered a
threat to the diversity of sessile corals and invertebrates.
For this reason, it is listed in the database of invasive
species of IUCN (Global Invasive Species Database) and
there is evidence of ecological impacts of this species in
some countries of the South Pacific, as in Colombia, where
high mortality of corals and octocoral coating has been
reported on the island of Malpelo (Sánchez, et al., 2011).
Orensanz, et al. (2002) detected more than 40 invasive
species in the Southern Atlantic Ocean, where poor
knowledge of the regional biota makes it difficult to track
456
We found that the greatest abundance of invasive
species was in the Ecuadorian central coast (Manabí),
belonging mainly to the cnidarians. However, the largest
diversity of species was in the south-central coast (Santa
Elena). The presence of these invasive species is possibly
due to the currents, ballast water and encrustations of
invaders on ships. We can speculate that factors such as
marine currents, rise of temperature, increase of maritime
traffic, global warming and invasive breeding strategies
will accelerate the augmentation of invasive alien species
and the loss of diversity of corals, octocorals, sponges
and other marine sessile invertebrates on the Ecuadorian
coast. Four of the six non-native species found on the
mainland of Ecuador (Pennaria disticha, Bugula neritina,
Asparagopsis taxiformis and Caulerpa racemosa) from
Table 1, are already present in the Galapagos Islands
(Danulat & Edgard, 2002; Keith et al, 2016).
This study must be taken into consideration by local
and regional government authorities to create public
policies and programmes to monitor for surveillance and
control of invasive species. These programmes have to
be integrated with socio-economic and ecological effects
and complemented by experimental design and analysis of
environmental variables to provide technical information
and a baseline of bio-invasions along the Ecuadorian
coast and Galápagos. It is important to avoid or limit the
expansion of invasive species that negatively affect the
marine biodiversity of mega-diverse countries such as
Ecuador and other countries of South America.
Cárdenas-Calle, et al.: Marine alien species in mainland Ecuador
ACKNOWLEDGEMENTS
We thank James Mair, Emeritus Professor, Heriot Watt
University for providing comments on an earlier version
of the manuscript. We also appreciated the collaboration
in the laboratory of Verónica Araujo, student from the
Universidad de Guayaquil. We are especially grateful to
the Ministry of Environment of Ecuador (Subsecretaría
de Gestión Marina y Costera), the Inter-American Bank
and the Global Environment Facility (GEF) for financial
support. This work was part of the project "Quantitative
subtidal and intertidal marine biodiversity inventories
in six marine protected coastal areas and four areas of
possible expansion" (CFC-001-2015).
REFERENCES
Ayón, H. (1988). Grandes Rasgos Geomorfológicos de la Costa
Ecuatoriana. Ecuador: PMRC-Proyecto de Manejo de Recursos
Costeros.
Barbosa, T.M., Gomes, P.B., Bergeron, A.S., Santos, A.M., Chagas,
C., Freitas, E.M.S. and Perez, C.D. (2014). ‘Comparisons of sexual
reproduction in Carijoa riisei (Cnidaria, Alcyonacea) in South Atlantic,
Caribbean, and Pacific areas’. Hydrobiologia 734: 201–212.
Boothroyd, J., Ayón, H., Robadue, D., Vásconez, J. and Noboa, R. (1994).
Características de la Línea Costera del Ecuador y Recomendaciones
para su Manejo. Technical Report 2076. Ecuador: PMRC-Proyecto
de Manejo de Recursos Costeros (Atlas). <http://simce.ambiente.gob.
ec/sites/default/files/documentos/anny/Caracter%C3%ADsticas%20
de%20la%20l%C3%ADnea%20costera%20del%20Ecuador%20y%20
recomendaciones%20para%20su%20manejo.pdf>.
Campbell, M.L., Keith I., Hewitt C.L., Dawson T.P. and Collins, K.
(2015). ‘Evolving marine biosecurity in the Galápagos Islands’.
Management of Biological Invasions 6(3): 227–230.
Cárdenas, M. and Triviño, M. (2014). Caracterización Bio-ecológica
de Siete Sitios de Buceo Situados Alrededor del Islote El Pelado en
Ayangue. Guayaquil-Ecuador: Reporte Técnico. Corpei.
Clarke, K.R. and Gorley, R.N. (2006). Primer v6: User Manual / Tutorial.
UK: Plymouth.
Concepcion, G.T., Kahng, S.E., Crepeau, M.W., Franklin, E.C., Coles,
S.L. and Toonen, R.J. (2010). ‘Resolving natural ranges and marine
invasions in a globally distributed octocoral (genus Carijoa)’. Marine
Ecology Progress Series 40: 113–127.
Danulat, E., and Edgar, G.J. (eds.) (2002). Reserva Marina de Galápagos.
Línea Base de la Biodiversidad. Santa Cruz, Galápagos, Ecuador:
Fundación Charles Darwin. Fundación Charles Darwin/Servicio Parque
Nacional Galápagos. <http://www.darwinfo undation.org/files/library/
pdf/RMG-Linea-Base-Bio. pdf>.
Darrigran, G. and C. Damborenea (2011). ‘Ecosystem engineering
impacts of Limnoperna fortunei in South America’. Zoological Science
28: 1–7.
Edgard, G.J., Banks, S., Bessudo, S., Guzman, H.M., Henderson, S.,
Martinez, P., Rivera, F., Soler, G., Ruiz, D. and Zapata, F. (2011).
‘Variation in reef fish and invertebrate communities with level of
protection from fishing across the Eastern Tropical Pacific seascape’.
Global Ecology and Biogeography 20: 730–743.
Global Invasive Species Database. (2017). ‘Species Profile: Carijoa
riisei’.
<http://www.iucngisd.org/gisd/speciesname/carijoa+riisei>.
Accessed 12 March 2017.
Grigg, R.W. (2003). ‘Invasion of a deep-water coral bed by an alien
species, Carijoa riisei’ Coral Reef 22: 121–122.
IUCN. (2017). ‘The IUCN Red list of Threatened Species’. <www.
iucnredlist.org>. Accessed 15 June 2017.
IUCN/SSC. (2014). ‘Invasive Species Specialist Group. Global Invasive
Species Database’. <http://www.iucngisd.org/gisd/>. Accessed 15 June
2017.
Kahng, S.E., Benayahu, Y. and Wagner, D.R.N. (2008). ‘Sexual
reproduction in the invasive octocoral Carijoa riisei in Hawaii’. Bulletin
of Marine Science 82: 1–17.
Kahng, S.E. and Grigg, R.W. (2005). ‘Impact of an alien octocoral,
Carijoa riisei, on black corals in Hawaii’. Coral Reef 24: 556–562.
Keith, I., Dawson, T., Collins, K. and Bank, S. (2015). ‘Especies
Marinas Invasoras en la Reserva Marina de Galápagos: Un Caso para
Investigación Adicional, Mejoramiento del Manejo y Revisión de
Políticas’. In: Informe Galápagos 2013–2014, pp. 83–88. Puerto Ayora,
Galápagos. Ecuador: DPNG, CGREG, FCD and GC.
Keith, I., Dawson, T., Collins, K. and Campbell, M.L. (2016). ‘Marine
invasive species: Establishing pathways, their presence and potential
threats in the Galápagos Marine Reserve’. Pacific Conservation Biology
22: 377–385.
Lowe, S., Browne, M., Boudjelas, S. and De Poorter, M. (2000). ‘100 of
the World’s Worst Invasive Alien Species: A Selection from the Global
Invasive Species Database’. The Invasive Species Specialist Group
(ISSG) a specialist group of the Species Survival Commission (SSC)
of the World Conservation Union (IUCN). <http://www.issg.org/pdf/
publications/worst_100/english_100_worst.pdf>. Accessed 15 June
2017.
Martínez, P. (2013). Potencial Amenaza de la Especie Invasora Carijoa
riisei en Áreas Marinas Protegidas del Ecuador (Potential Threat of the
Invasive Species Carijoa riisei in Ecuadorian Marine Protected Areas).
Unpublished report. Quito, Ecuador: Instituto Nazca de Investigaciones
Marinas.
Mendoza, R., Born-Schmidt, G., March, I.J. and Álvarez, P. (2014).
‘Especies Invasoras Acuáticas y Cambio climático’. In: R. Mendoza
and P. Koleff (eds.) Especies acuáticas invasoras en México, pp. 469495. México: Comisión Nacional para el Conocimiento y Uso de la
Biodiversidad.
Miloslavich, P., Klein, E., Díaz, M.J., Hernández, C., Bigatti, G., Campos,
L., Artigas, F., Castillo, J., Penchaszadeh, P.E., Neill, P.E., Carranza, A.,
Retana, M., Díaz de Astarloa, J.M., Lewis, M., Yorio, P., Piriz, M.L.,
Rodríguez, D., Yoneshigue-Valetin, Y., Gamboa, L. and Martín, A.
(2011). ‘Marine biodiversity in the Atlantic and Pacific coasts of South
America: Knowledge and gaps’. PLOS ONE 6: 1–43.
Ministry of Environment. (2014). Términos de Referencia de la
Consultoría Inventarios Cuantitativos Submareales e Intermareales
de Biodiversidad Marina en Seis Áreas Marino Costeras Protegidas y
cuatro Zonas de Posible Expansión (Reference Terms of the Consultancy
Submarine and Intertidal Quantitative Inventory of Marine Biodiversity
in Six Marine Protected Areas and Four Zones of Possible Expansion).
Guayaquil: Ministry of Environment.
Orensanz, J., Schwindt, E., Pastorino, G., Bortulus, A., Casas, G.,
Darrigran, G., Elías, R., López Gappa, J.J., Obenat, S., Pascual, M.,
Penchaszadeh, P., Piriz, M.L., Scarabino, F., Spivak, E.D. and Vallarino,
E.A. (2002). ‘No longer a pristine confine of the world ocean–a survey
of exotic marine species in the southwestern Atlantic’. Biological
Invasions 4: 115–143.
Pimentel, D., Zúñiga, R. and Morrison, D. (2005). ‘Update on the
environmental and economic costs associated with alien invasive
species in the United States’. Ecological Economics 52: 273–288.
Prenter, J., MacNeil, C., Dick, J.T.Z. and Dunn, A.M. (2004). ‘Roles of
parasites in animal invasions’. Trends in Ecology and Evolution 19:
385–390.
Sánchez, J.A., Gómez, C.E., Escobar, D. and Dueñas, L.F. (2011).
‘Diversidad, abundancia y amenazas de los octocorales de la isla
Malpelo, Pacífico Oriental Tropical, Colombia’. (‘Diversity, abundance
and threats to Malpelo Island´s octocorals, Tropical West Pacific,
Colombia’). Boletín de Investigaciones Marinas y Costeras – INVEMAR
40: 139–154.
SARCE. (2012). Protocolo y Diseño de Muestreo para la Evaluación de
la Biodiversidad Marina en Suramérica (Protocol and Sampling Desing
for the Evaluation of Marine Biodiversity in South America). Caracas,
Venezuela: South American Research Group on Coastal Ecosystems.
https://www.oceanbestpractices.net/handle/11329/412?show=full
Schüttler, E. and Karez, C.S. (eds.) (2008). Especies Exóticas Invasoras
en las Reservas de la Biosfera de América Latina y el Caribe. Un
Informe Técnico para Fomentar el Intercambio de Experiencias
entre las Reservas de Biosfera y Promover el Manejo Efectivo de las
Invasiones Biológicas. (Exotic Invasive Species of Biosphere Reserves
of Latin American and the Caribbean. A Technical Report to Promote
the Effective Management between Biosphere Reserves and to Promote
the Effective Management of Biological Invasions). Montevideo,
Uruguay: UNESCO.
Silva, A.G., Lima, R.P., Gomes, A.N., Fleury, B.G. and Creed, J.C. (2011).
‘Expansion of the invasive corals Tubastraea coccinea and Tubastraea
tagusensis into the Tamoios Ecological Station Marine Protected Area,
Brazil’. Aquatic Invasions 6, supplement 1: S105–S110.
Sonnenholzner, J., Brandt, M., Francisco, V., Hern, A., Luzuriaga, M.,
Guarderas, P. and Navarro, J.C. (2013). ‘Echinoderms of Ecuador’. In:
J.J. Alvarado and F.A. Solís-Marín (eds.) Echinoderm Research and
Diversity in Latin America, pp. 183–233. Berlin Heidelberg: SpringerVerlag.
Sullivan, K.K. and Bustamante, G. (1999). Setting Geographic Priorities
for marine Conservation in Latin América and the Caribbean. Arlington,
Virginia: The Nature Conservancy.
Venkataraman, K., Raghunathan, C., Satyanarayana, C. and Rajkumar,
R. (2016). ‘Invasion of snowflake coral, Carijoa riisei (Duchassaing
and Michelotti, 1860), in Indian seas: Threats to coral reef ecosystems’.
Indian Journal of Geo-Marine Sciences 45: 1403–1408.
457
J.C. Horrill, M.K. Oliver and J. Stubbs Partridge
Horrill, J.C. M.K. Oliver and J. Stubbs Partridge. Lessons on effectiveness and long-term
prevention from broad-scale control of invasive alien species in Scotland’s rivers and lochs
Lessons on effectiveness and long-term prevention from broad-scale
control of invasive alien species in Scotland’s rivers and lochs
J.C. Horrill1, M.K. Oliver1 and J. Stubbs Partridge2
Rivers and Fisheries Trusts of Scotland, 11 Rutland Square, Edinburgh EH1 2AS, UK. <chris@rafts.org.uk>. 2Scottish
Natural Heritage, Great Glen House, Leachkin Road, Inverness IV3 8NW, UK.
1
Abstract Prior to 2008 there were few invasive alien species (IAS) initiatives operating in Scotland on a scale required
for effective control. The establishment of the Biosecurity and Invasive Non-Native Species Programme by the Rivers and
Fisheries Trusts of Scotland was the first attempt to link local efforts with national IAS strategy on scales appropriate to
the effective control of target species. The programme worked with 26 local fisheries trusts to produce biosecurity plans
that covered over 90% of Scotland’s rivers and lochs. The programme implemented a range of prevention measures,
including promoting awareness of invasive species issues and the need for biosecurity among water users. Projects were
established for invasive plants on most major river systems, and for American mink (Neovison vison) in the north of
Scotland. These projects involved public/private partnerships, using a mix of professional staff and volunteers. Interactive
data management systems were developed to manage input from a large number of individuals and to inform an adaptivemanagement approach. These control projects demonstrated that it is feasible to reduce the size and density of target
populations of invasive species across large geographic areas. The key to maintaining the momentum of this control
effort in the future will be to demonstrate sustainable IAS management in the longer term. This challenge led to the
formulation of the Scottish Invasive Species Initiative (SISI) whose overall aim is the development of a long-term, costeffective strategy for IAS management throughout the north of Scotland. SISI will test strategies derived from experience
and information from previous control projects. Important areas that the initiative will seek to address include defining
outcomes, integrating IAS management into other management initiatives, and maintaining partnership interest and
cohesiveness in a challenging funding environment.
Keywords: adaptive management, biosecurity, community-based effort, giant hogweed, Himalayan balsam, IAS,
Japanese knotweed, mink, rivers
INTRODUCTION
Scotland depends on the quality of its iconic natural
environment for economic and recreational wealth. The
high number of protected sites (1,868) and designated
natural features (5,376) reflects the importance placed on
natural heritage.
Watercourses are integral, defining features of Scotland’s
landscapes and culture. Historically, Scottish society relied
on healthy rivers and lochs for food, recreation, transport
and industry. Art, folklore and traditional activities have
long drawn inspiration from them. Today, the economic
reliance extends to whisky-distilling, salmon-farming,
tourism and many new forms of recreation. In 2010,
Scottish residents generated £2.3 billion from their visits
outdoors (SNH, 2011). Recreational freshwater fishing is
estimated to support around 4,300 jobs, contributing £79.9
million to the economy (Marine Scotland, 2017).
Within increasingly fragmented landscapes, water
courses also function as corridors between habitats for
biodiversity. This vital function is compromised by invasive
alien species (IAS) (Also known as Invasive Non-Native
Species (INNS) in the United Kingdom) for which rivers
and lochs are excellent pathways into the broader natural
environment. The margins and shorelines of watercourses
themselves are among the most exposed to the risk of IAS
spread and damage. Climate change, pollution and habitat
disturbance accelerate rates of invasion, with corresponding
costs for socio-economic, human and ecological well being
(Forest Research, 2008; Williams, et al., 2010).
The UK and Scottish Governments have recognised
the IAS threat. The Great Britain Invasive Non-Native
Species Strategy (GBNNSS, 2008) is a policy and
strategic response. The Scottish Environment Protection
Agency (SEPA) addresses the threat through the INNS
supplementary plan to the Scotland and Solway-Tweed
River Basin management plans (SEPA, 2009a; SEPA,
2009b). Scottish Natural Heritage (SNH), a Scottish
Government agency, has included IAS in its Species Action
Framework (Raynor, et al., 2016).
Prior to high level recognition of this sort, the vast
majority of reponses to IAS were small-scale and localised.
Management on larger scales was confined to catchmentbased control of invasive alien plant species (IAPS) on
the River Tweed (Tweed Forum, 2006) and to control
of American mink (Neovison vison) in the Cairngorms
National Park and rural Aberdeenshire (Bryce, et al., 2011).
The scale of the threat, the likely severity of ecological,
social and economic impacts and the prospect of rises
in control and eradication costs have constituted a case
for better, more strategic and systematic approaches to
managing IAS. This paper reports on the results and lessons
learnt between 2008–2017 from the work of 26 member
organisations of the Rivers and Fisheries Trusts of Scotland
(RAFTS) in partnership with government agencies and
universities to address the IAS threat to Scotland’s rivers
and lochs. We will also refer to an ambitious project in
which those lessons are incorporated to manage multiple
IAS cost effectively in the long term over 29,500 km2 of
northern Scotland.
METHODS
Biosecurity planning
At the northern invasion front for high-impact IAS of
the United Kingdom and Europe, Scotland was well placed
to manage the threats strategically at national and local
scales. On the national scale there was an opportunity to
defend the IAS-free region to the north of the front, control
IAS in the lightly infested catchments in northern and
southern Scotland, before addressing the more impacted
areas of central Scotland.
RAFTS and its 26 local Trust members created areaspecific biosecurity plans in three phases between 2008
and 2010 (Fig. 1). All plans used a template designed by
RAFTS in consultation with the Great Britain Non-Native
Species Secretariat, Scottish Government, SNH and
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
458
up to meet the challenge, pp. 458–465. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Horrill, et al.: Alien species in Scotland’s rivers and lochs
SEPA. The template linked key elements of IAS policy
and strategy to local action and acted as a framework for
universal consistency. Plan objectives reflected the three
key elements of the Great Britain INNS Strategy (GBNNSS,
2008): (1) prevention, early detection, and surveillance;
(2) monitoring and rapid response; and (3) mitigation,
control and eradication. Objectives and actions were also
linked to related plans and initiatives such as River Basin
Management Planning (SEPA, 2009a; SEPA, 2009b). This
approach translated the key elements of national policy and
strategy into action across relevant sectors in ways which
emphasised coordination and partnership.
Funding secured for a series of projects from 2009 to
2017 enabled local organisations to coordinate and monitor
the control of invasive alien plant species and American
mink by professionals and volunteers (Table 1). RAFTS
provided the overall coordination, strategic direction and
evaluation of the activities. SNH, SEPA, the University
of Aberdeen (mink) and Queens University Belfast
(plants) provided specific technical support. Principal
target species were the IAPS, giant hogweed (Heracleum
mantegazzianum), Japanese knotweed (Fallopia japonica)
and Himalayan balsam (Impatiens glandulifera) and the
alien invasive mustelid American mink – all recognised
as high-impact species for waterbodies and/or biodiversity
(UKTAG, 2015).
Engagement
Engagement of key stakeholders was critical given
the scale of the work, the need to obtain permissions
for access and the recruitment and maintenance of the
volunteer workforce. Awareness campaigns, mailshots,
presentations, meetings with local environment/community
groups, schools and individuals, newsletters, the websites
and media were means to initiate contact with potential
volunteers. Working in public areas and approaching
landowners for permission also proved effective in
engaging local communities.
Once engaged, stakeholders were kept informed
through websites, newsletters, media and meetings and
later through interactive reporting systems. Participating
organisations and individuals received skills-training for
efficient, effective, legally compliant surveillance and
control of IAS. Formal training courses were tailored to
roles in the control strategy. Volunteers were also offered
informal training if they were unavailable for formal
courses.
Fig. 1 Map of Trust areas.
Table 1 Summary of projects implemented through the RAFTS Biosecurity and Invasive Non-Native (Alien) Species
Programme with duration, description of activities and geographic scope and participating local partner organisations.
Duration
Name of Project
Pan Scotland Invasive 2009–
2016
Non Native Plant
Species Control
Scottish Mink
Initiative (SMI)
2010–
2015
Controlling priority
invasive non-native
riparian plants and
restoring native
biodiversity (CIRB)
2010–
2014
Description
Participating Trusts
A series of projects for the control and
eradication of invasive alien riparian plant
species in northern, southern and central
Scotland. Included biosecurity, awareness
and training of professional staff and
volunteers.
Aiming to eradicate breeding mink from
20,000 km2 (later extended to 28,000 km2)
of north and north-eastern Scotland. The
Initiative also supports awareness and
local capacity building activities as well
as the development of local management
models for future mink control.
Control and eradication of invasive alien
riparian plant species in 12 catchments in
southern Scotland, piloting biosecurity,
awareness activities, training of Trust staff
and volunteers, best practice identification
and dissemination.
Annan; Argyll, Ayrshire, Cromarty
Firth; Deveron, Isla and Bogie;
Don; Dee; Esk River; Forth;
Findhorn, Nairn and Lossie;
Galloway; Lochaber; Kyle of
Sutherland; Tweed; West Sutherland
Cromarty Firth; Deveron, Isla and
Bogie; Dee; Don; Esk; Findhorn,
Nairn and Lossie; Spey, Tay, Ythan
Argyll; Ayrshire; Galloway; Tweed
Forum
459
Island invasives: scaling up to meet the challenge. Ch 2F Other taxa: Aquatic
IAPS densities, distribution and control
Surveys of river and loch catchments identified the
location, extent and abundance of IAPS. The distribution
of IAPS populations were entered into a geo-database
along with estimates of abundance based on the DAFOR
scale (Kent & Coker, 1992). The impact of treatment was
monitored by recording distribution and abundance posttreatment. Treatments varied by species but were primarily
foliar leaf spray (Japanese knotweed and Himalayan
balsam), stem injection (Japanese knotweed) and physical
removal (Himalayan balsam).
Initially the majority of local Trusts took a ‘top down’
approach to control, starting at the upstream extent of IAPS
distribution and working downstream. The rationale was
the reduction of potential reinfestation of treated sites from
upstream populations. Later, working from the lower to the
upper catchment was adopted by some Trusts when treating
whole catchments. This tactic recognised that plants lower
in the catchment developed earlier than those in the upper
regions.
the Dee, Spey and Ythan where control had been ongoing
since 2006 / 2007 (Bryce, et al,. 2011; Lambin, et al.,
2019). The GLMM model accounted for differences
among catchments, which was fitted as a fixed effect and
as an interaction with time of mink control (i.e. the effect
of mink control was allowed to vary by catchment). Nonindependence between multiple records from the same
raft(s) was accounted for by fitting raft as a random effect.
The effectiveness of IAPS treatment was assessed by
the area cleared of infestation (i.e. no regrowth occurred for
a year or more), the percentage decrease in coverage and
the number of sites in a low maintenance state (DAFOR
≤ 1 (Rare) = 1–10% coverage) before and after treatment.
Where coverage was recorded using the DAFOR scale,
the mid value for each category of the index was used.
Use of DAFOR categories, although simpler to record,
encompasses score ranges of 10–25% and therefore more
subtle changes in IAPS coverage may not be apparent with
this index.
RESULTS
Mink control
Stakeholder engagement
Volunteers and paid staff relied mainly on mink rafts
to detect and trap American mink. Originally conceived
by the Game and Wildlife Conservation Trust (GWCT)
(Reynolds, et al., 2004), the mink raft is a floating platform
on which a tunnel covers a clay pad. The raft is anchored to
the bank of a waterway. American mink are predominantly
active within 10 m of waterways (Yamaguchi, et al., 2003),
are naturally attracted to tunnels and leave footprints in the
clay when investigating them. Once a mink is detected, a
live-capture cage-trap is inserted in the tunnel. Captured
mink were despatched humanely. Carcasses were tagged
and sent to Aberdeen University to determine sex, age and
provenance based on genetic profile (Fraser, et al., 2013;
Melero, et al., 2015; Ruiz-Suarez, et al., 2016).
Throughout the reporting period a total of 1,000
volunteers serviced 2,020 surveillance points for mink
control (Fig. 2) and at least 391 volunteers participated in
IAPS control, contributing ≥ 2,587 hours of work. Actual
numbers at any given time varied, being dependent on the
size of area being managed and funding availability. In
2015 there were approximately 800 volunteers participating
in mink and IAPS control. Continual recruitment was
necessary to offset loss of volunteers. Volunteers left
because of a number of reasons. A small but significant
number decided it was not really something they wanted
to do shortly after recruitment. Other reasons were moving
from the area, changed employment and boredom.
Evaluation
Volunteers participating in mink control were from a
broad range of backgrounds. Residents of the area with no
Stakeholder engagement and impacts on IAS
populations were evaluated in 2015 as measures of success.
Data recorded for stakeholder engagement included
contacts, background, and time spent. Assessment of mink
control recorded raft locations and status, raft checks, mink
sightings and captures. The locations and extent of target
IAPS were recorded using geographic positioning systems
and abundance by percentage cover or the DAFOR scale.
From 2012 data recording by volunteers and professional
staff used specifically designed digital tools that not only
managed data but also fed back information to users. The
web- and map-based interactive geo-database for IAPS
management made it easy to acquire survey and monitoring
data and to translate changes in IAPS treatment status and
abundance to maps presented on the website. An online
platform, the MinkApp was developed in collaboration
with Aberdeen University’s dot.rural initiative (http://
www.dotrural.ac.uk/) for the recording, management and
presentation of data derived from American mink control.
The MinkApp used natural-language-generation (NLG) to
inform volunteers by email of mink captures and sightings
in their area.
Trends in mink detections and captures were used to
determine whether large-scale coordinated control efforts
had had an impact on mink populations. The best (least
biased) impact data were derived from the checking
records for mink rafts. Detection rates could be calculated
from the percentage of raft checks where mink footprints
were observed. Further analysis through a generalised
linear mixed effects model (GLMM) was carried out on
long-term mink detection data from three test catchments:
460
Fig. 2 Location of the 2,020 surveillance points (rafts,
tunnels, traps) monitored for American mink between
2006 and 2015.
Horrill, et al.: Alien species in Scotland’s rivers and lochs
connection to the local environment constituted the largest
proportion, followed by two professional groups – estate
workers (game keepers and land managers) and fisheries
personnel (managers, owners, guides and anglers). These
three groups provided 78% of volunteers. The remaining
22% came from conservation organisations, government
agencies and local councils, the tourism and leisure
industry, farmers, fish farmers and University staff.
The degree to which individual volunteers engaged
with control activities varied greatly, with most content
with participating in surveillance e.g. checking mink rafts.
However, a relatively small but significant proportion
of volunteers, in terms of their contribution, received
instruction for skilled activities e.g. humane despatch,
stem injection and foliar spray near watercourses. These
latter tasks required informal training and/or certification
and increased commitment from the volunteer and host
organisation.
There was only one landowner where there was issue
with gaining access to land despite the large geographic
area and the number of landowners involved. Access
permissions were initially given verbally but insurance
requirements meant that written permissions were
increasingly required.
to July 2015) were concentrated along the frontier of the
project area, which was consistent with frontier catchments
receiving an influx of dispersing mink from outside of the
control area and the coast.
Trends in mink captures followed those of the detection
rate, with a decrease from over 280 in 2012, to only 98
mink captured in the 12 months prior to July 2015.
Although mink were captured across the raft network, the
areas with the highest numbers of captures reflected the
optimum habitat for mink and the history of control effort.
In agreement with the mink raft detection data, nearly all
of the captures in 2015 were from lowland or coastal areas,
indicating an overall contraction of the mink population
both in range and population size (see also Lambin, et al.,
2019).
The GLMM analysing how mink detection rate changes
with year of mink control, showed a clear and statistically
significant (P < 0.0001) negative relationship (Fig. 4;
Table 2). Based on the fitted curves, the model predicts that
American mink
Across the entire control area, and considering all raftcheck records in a calendar year, there was a steady decline,
from a positive check rate of around 0.14 in 2011, to a low
of around 0.02 in July 2015 (Fig. 3). The majority of the
86 positive raft checks towards the end of the study period
(from a total of 2,776 recorded in the period July 2014
Fig. 3 Changes in the mink detection rate (number of
positive raft checks / total number of raft checks) per
year of coordinated mink control. Numbers above the
points show the total number of checks from which the
rates are estimated.
Fig. 4 The effect of control on mink detection rates
(abundance) calculated by a generalised linear mixed
effects model (GLMM). The black lines are fitted curves
for the Dee, Spey and Ythan river catchments The grey
lines areas are 95% profile confidence intervals.
Table 2 Summary table for a generalised linear mixed model (GLMM) analysing the relationship between
mink detection rate (per raft check) and the year of mink control (by river catchment). Data are for the
rivers Dee, Spey and Ythan. Observations is the number of raft checks. Groups refers to the number
of rafts.
Observations: 9086
Intercept
Year of control
Catchment (Spey)
Catchment (Ythan)
Year of control: Spey
Year of control: Ythan
Groups: 399
Estimate
-1.33
-0.35
-1.12
-0.55
0.00
0.15
Residual d.f. 9079
S.E.
0.24
0.05
0.44
0.30
0.10
0.06
Variance: 1.57
Z value
-5.67
-7.27
-2.58
-1.85
0.02
2.54
St. dev: 1.25
P value
< 0.0001
< 0.0001
0.01
0.06
0.99
0.01
461
Island invasives: scaling up to meet the challenge. Ch 2F Other taxa: Aquatic
mink abundance will be reduced to ca. 40% of the starting
abundance in four years and further to around 6% of initial
levels after nine years. A large amount of the uncertainty in
the model’s predictions (illustrated by the 95% confidence
intervals [grey lines] in Fig. 5) is attributable to differences
between the catchments, rather than the overall estimate of
the effect of mink control (Table 2). This was particularly
true of one catchment where the mink population remained
high before dropping abruptly after control in the adjacent
catchment.
A small number of rafts influenced trends significantly
with a majority of rafts never detecting any mink
footprints. All information on mink presence came from
36% of rafts (n = 357) checked at least once. In fact, only
6% of checked rafts (a mere 59) accounted for 637 (53%)
of the 1,307 detections. Whilst factors such as duration of
raft placement and checking frequency may influence this
result, the take home message is that a small portion of the
raft network does most of the work in detecting, and vis a
vis removing, mink.
Invasive alien plants
The 10 river Trusts that supplied information surveyed
a minimum of 2,403 km of waterways (Table 3). Their
surveys revealed that IAPS were widespread (extending
over ca. 1,603,821 m2) and had become a serious threat
to riparian biodiversity and activities along Scottish river
corridors.
Japanese knotweed was the most frequently
encountered IAPS. Trusts recorded it in all survey areas
though the extent varied significantly among them (Table
3). Giant hogweed was least prevalent and abundant. Three
Trusts reported it absent and a fourth discovered only one
small stand. But in all other areas infestations averaged >
4,000 m2. In Ayrshire giant hogweed had invaded 188,000
m2 . Himalayan balsam infestations proved to be the most
challenging. This IAPS had reached 699,233 m2 of river
corridor. Stands in two catchments extended over tens of
kilometres.
Table 3 Summary of the area surveyed (in metres) and area recorded as infested by each
IAPS for each of the 10 trusts (reported as m2).
Annan
Argyll
Ayrshire
Cromarty
Dee
FNLT
Galloway
Lochaber
Nith
WSFT
Total
Area surveyed
GH
HB
JK
Total
197,000
195,000
739,000
300,000
170,000
103,500
114,000
42,400
160,450
30,000
2,403,834
20
188,000
27,000
4,176
72,000
4,196
0
41,955
0
337,347
200,000
204,000
128,500
32,938
62,700
75
0
70,430
590
699,233
11,364
9,198
257,000
54,500
41,768
88,500
21,663
43,500
39,718
30
567,241
211,384
9,198
649,000
210,000
78,882
223,200
25,934
43,500
152,103
620
1,603,821
Fig. 5 Schematic of a graduated three-phase strategy for mink control (based on capture data from the River Dee, NE
Scotland). In Stage 1 (years 1–4) mink abundance is at its initial maximum. The box on the right illustrates how the
strategy moves from a saturated raft network in Stage 1, to cover all female capture locations in Stage 2, and only a
subset of these in Stage 3.
462
Horrill, et al.: Alien species in Scotland’s rivers and lochs
Success in clearing areas of infestations was limited
with 16%, 11% and 10% of the original area of infestation
cleared for giant hogweed, Japanese knotweed and
Himalayan balsam, respectively (Table 4). However,
decreases in coverage between 50% and 80% were
common for all three target IAPS.
The greatest decrease in coverage was for Japanese
knotweed, with five areas achieving >85% decrease.
Despite the reduced coverage, shoots from the sub-surface
rhizome prevented sites from being categorised as cleared.
Cover of giant hogweed fell by 53%–75%. However, there
was mixed success in controlling Himalayan balsam (Table
4). Trusts reported that effective control of this IAPS was
problematic as it is easy to miss individual plants hidden
among native vegetation, or in areas of limited access. In
four areas, Himlayan balsam was also anecdotally observed
to quickly colonise sites that had recently been cleared of
giant hogweed or Japanese knotweed. Of note, however,
is that both Nith and Cromarty Trusts, using a targeted
approach and a larger coordinated workforce, decreased
coverage of Himlayan balsam by >82% (as well as clearing
>29,000 m2) across large areas.
Standardised percentage coverage decreased from a
median of 38% (mean 33%) to 5% (mean 14%) in 447
pre-treatment sites following control. The majority of
sites (327; 73%) showed improvement, 103 (23%) were
recorded as having no change, and infestation levels at 17
(4%) had got worse. Around half the sites infested by giant
hogweed and Japanese knotweed, and 38% of those by
Himalyan balsam, were in a low maintenance state after
treatment (Table 5). This was despite the reported increase
of infestations of giant hogweed after the large floods of
the winter of 2013/14.
Costs
The work reported in this paper was undertaken
through the sequential securing of short-term (1–4 year
duration) funding. Consequently funding was cyclical
with periods of higher funding alternating with those of
low or no funding (Fig. 6). Using northern Scotland as an
example, the amount of funding secured for IAS work has
increased in each subsequent funding phase, from £124,000
(1996–2005), £639,000 (2006–2009) to over £1.95 million
provided in the period 2010–2015. The increased funding
reflected the expanding geographic reach (from 5,000
km2 to almost 30,000 km2) and complexity of the work
undertaken. This included the addition of IAPS control in
2009 and biosecurity, awareness, education and capacity
building activities after 2010.
Fig. 6 Funding for IAS work in northern Scotland.
Table 4 Area cleared (no growth detected in post-treatment survey) in m2 and relative percentage decrease
in mean coverage at infested sites for each of the INNPS and each Trust. A dash (-) indicates that no data
were available.
GH
HB
JK
Area cleared % decrease Area cleared % decrease Area cleared % decrease
Annan
Argyll
Ayrshire
Cromarty
Dee
FNL
Galloway
Lochaber
Nith
WSFT
Total
0
53,452
750
0
40*
54,242
53
60
57
0
75
-
0
>6,684
38,000
0
0
29,871
0
>74,555
19
25
82
0
50
94**
90
0
8,070
29,722
7,750
0
2,840
16,268
0
30
64,680
63
88
47
81
65
84
42
99
100
* The Nith group recorded the number of plants treated, rather than area cleared.
** Percentage calculated as change in the number of plants treated between initial (maximum) levels and final treatment
in 2014.
Table 5 Number and percentage of the total number of sites that were in a low maintenance state
before and after treatment for each target species.
Total no. of sites
Giant hogweed
Japanese knotweed
Himalayan balsam
468
598
293
Before treatment
No. of sites % of sites
82
41
40
17
7
14
After treatment
No. of sites % of sites
243
295
111
63
88
100
463
Island invasives: scaling up to meet the challenge. Ch 2F Other taxa: Aquatic
DISCUSSION
Findings and lessons for future work
The control strategies and measures for both IAPS
and American mink have had a demonstrable, although
variable, impact in suppressing target populations in terms
of coverage and population density over large geographic
areas (see also Bryce, et al., 2011; Melero, et al., 2015;
Oliver, et al., 2016). The variation in results suggests
there is room for improvement in strategy and local
implementation.
The use of an evidence-based approach, derived from
evaluation of activities and research associated with
the project, provided the central core of the adaptivemanagement strategy. The findings were utilised to improve
control strategies (e.g. concentration of surveillance in
lowland areas and along migration routes for American
mink, control methods for IAPS, engagement and retention
of volunteers, and implementation of management efforts
at an appropriate geographic scale in defensible areas
for all IAS). An example of the latter is that the GLMM
analysis highlighted the importance of taking a coordinated
multi-catchment approach to mink control as the number
of mink in a catchment depends on control both within that
catchment and in neighbouring catchments.
Working over such a large geographic scale, including
urban areas, with limited secured funding was made possible
by the use of a large trained volunteer workforce supported
by professional staff. Staff were either employed by the
project or from local organisations. The latter arrangement
allowed the building of capacity for volunteer management
and IAPS control within the organisations. Although
this approach helped to build longer term management
sustainability, it sometimes resulted in competing priorities
between the project and the organisation. Employing
dedicated project staff avoided this conflict but did not
effectively address long-term sustainability, as employment
ended with the cessation of project funding.
The use of large volunteer networks rather than
increased numbers of staff reduced employment costs, a
significant cost. However it did not reduce liability risk
for the organisation(s) that supported the network. To
mitigate risk as the project developed, RAFTS increasingly
used written rather than verbal permission for volunteer
participation and access agreements. The information
and training given to volunteers increased, particularly
regarding health and safety. Organisation policies and
public liability insurance was also regularly reviewed
in light of volunteer numbers and their work. Changes
in project management structure required revision of
all agreements. One outcome of these changes was that
significant numbers of volunteers expressed concern and
dissatisfaction with perceived increased bureaucracy, with
a small number withdrawing their participation.
Management over such a large area required the building
and maintenance of coordinated partnerships with defined
roles for individual partners at both local and national level
(Table 6). At the local level, non-government/non-profit
organisations (Trusts) provided the hub of the partnerships
and collaboration. The Trusts have close ties to sectors
of the local communities, particularly landowners. At
the national level RAFTS was the main contact point for
government agencies and universities, and coordinated the
work of the local organisations. Partnership arrangements
were not pre-determined but rather developed over the
course of the work and in response to the varied demands
of the management strategies employed. Partnerships and
collaboration involved over 70 organisations, including
the Scottish Government, state agencies, local authorities,
universities, >50 local non-government organisations and
businesses and over 800 volunteers at any one time.
Coordination was generally effective but there were
instances of inconsistency of approach and in data collection
among local organisations (Arts, et al., 2013). Although
consistency of data collection improved with the advent
of the on-line reporting systems, ensuring consistency of
approach and data collection among large numbers of local
organisations remained a significant challenge.
Common interest formed the basis for collaboration.
Differing characteristics of communities (individuals and
community organisations) within and among geographic
areas of Scotland meant approaches to engagement varied.
The diverse composition of the volunteer base demonstrated
that IAS control, particularly of American mink, provided
a common base for a wide range of community groups,
some of which had a history of conflicting interest (e.g.
gamekeepers and bird conservationists). Motivational
factors included professional or commercial interest and
a concern for the local environment – as expressed by
residents who made up a large proportion of the volunteers.
Taking action and demonstrating results were
important factors in retaining participating volunteers and
organisations. Demotivating factors included the breaks
in project activities caused by short term funding cycles
and perceived increased bureaucracy. The use of online reporting systems provided a means to disseminate
progress and results through a limited functionality for data
interrogation (mink) (Beirne & Lambin, 2013) and a map
interface for IAPS. These reporting mechanisms became
part of an overall volunteer and organisational recruitment
and retention strategy that combined a variety of awareness
activities with training and legal empowerment. Successful
control also influenced volunteer retention with the lack of
detection of IAS leading to boredom. Maintaining interest
and motivation remains a critical long-term challenge for
future management (Beirne & Lambin, 2013).
Despite repeated efforts to obtain long term funding,
IAS control in Scotland has relied on short term, or project
specific, funding. The resultant funding cycles occur as
one project has to finish before funding for the next stage
can be secured. Start-stop cycles result in a loss of staff,
volunteers, equipment and, as a consequence, momentum,
capacity and credibility (see also Lambin, et al., 2019).
Furthermore, overall costs increase as start up costs
(staff and volunteer recruitment, training, control) exceed
recurrent costs of established projects.
Funders’ regulations also influence the work that can
be undertaken. The majority of short-term funders require
Table 6 Contributions by participating institutions.
Level of collaboration
Strategy
Management
Implementation
Evaluation
464
Partnership organisations
RAFTS, GB Non Native Species Secretariat, Scottish Environment Protection Agency,
Scottish Natural Heritage, national park authorities (Cairngorms and Loch Lomond),
RAFTS and 18 member Trusts
18 local trusts, other non-government organisations e.g. (Scottish Wildlife Trust,
Royal Society for the Protection of Birds), local authorities (Highland, Moray, Rural
Aberdeenshire, Angus, North Tayside, Argyll and Bute, Ayrshire, Dumfries and Galloway.
RAFTS, University of Aberdeen, Queens University Belfast.
Horrill, et al.: Alien species in Scotland’s rivers and lochs
tangible benefits for their support. These benefits are more
easily expressed in terms of IAS reductions than prevention
(biosecurity), where no occurrence or a ‘negative’ result
defines success. Regulations have also prevented funding
being used for rapid-response, another key element of
successful IAS management. Funding for IAS management
should recognise that ‘negative’ results indicate success
both in prevention and control, have flexibility to allow
for rapid response and changes in approach required by
adaptive-management and be available for work over
appropriate geographic- and time-scales.
Although there is still no long-term funding of IAS
control in Scotland, project funding has been secured
for the Scottish Invasive Species Intiative (SISI) (2018–
2022). SISI aims to develop a long-term, cost-effective
management system for multiple IAS across 29,500 km2 of
northern Scotland. The project builds on the experiences of
its predecessors and tests more focused strategies for IAS
management.
One such approach to mink control derives from the
variation in the relative contribution of individual rafts to
overall detection rates, coupled with the analysis from the
GLMM. The model predicts abundance will be more than
halved following four years of control and reduced to < 10
% after ten years. Accordingly, capture data will be used to
reduce raft coverage in three stages over the same timeframe
(Fig. 5). If patterns of mink dispersal and settlement are
influenced by habitat quality, despite the species’s mobility
and generalist habits, reductions would track capture rates
for females. This assumes that populations under control
pressure will reoccupy optimum habitat preferentially, and
that concentrations of female mink will indicate where that
is. Reactive redeployment may be required in response
to localised increases in mink activity. If successful, the
strategy will use the best available evidence and scientific
understanding to substantially reduce costs.
Protecting non-invaded areas through awareness
targeted to user groups (e.g anglers and boaters) and the
use of biosecurity stations and individual biosecurity kits
is a key component of the project. Habitat restoration using
resilient native communities will be tested as a means to
reduce reinvasion risk of areas cleared of IAS.
Emphasis is placed on strengthening the capacity of
local organisations, so IAS management becomes part of
normal working practices. SISI will also develop means
to maintain volunteer participation over the timeframes
required to manage IAS. Evidenced-based adaptive
management is central to the strategic approach of SISI
and the project will develop interactive and map-based
data-recording systems.
SISI faces some significant challenges in balancing
costs with outcomes, particularly in regard to reducing
introductions and spread over such large geographic areas
and defining what reduction in IAS can be sustained.
Effective coordination, and quality assurance, of the work
undertaken by multiple local organisations is not to be
underestimated.
Despite the challenges, it is envisaged that by the end
of SISI the more focused control will have suppressed
target populations to levels where that suppression can be
affordably maintained by motivated local organisations
and their volunteer networks (Fig. 6, from 2022–2026).
However, post-project IAS management in Scotland will
still require additional funding to that provided by local
organisations and at present it is not clear how that will be
provided.
ACKNOWLEDGEMENTS
Funders of this work were: Scottish Natural Heritage;
People’s Trust for Endangered Species; Tubney Charitable
Trust; Cairngorms National Park Authority; the Scottish
Government and the European Community Cairngorms,
Highland, Moray, Rural Aberdeenshire and Rural Tayside
Local Action Groups LEADER 2007–2013 Programme:
EU Interreg Iva: SEPA Water and Environment Fund, Local
Trusts and various local funders. The authors would also
like to thank Aberdeen University and Queens University
Belfast for their technical support.
REFERENCES
Arts, K., Webster, G., Sharma, N., Melero, Y., Mellish, C.S., Lambin, X.
and van der Wal, R. (2013). ‘Capturing mink and data: Interacting with
a small and dispersed environmental initiative over the introduction
of digital innovation’. Framework for Responsible Research and
Innovation in ICT FRRIICT. <http://responsible-innovation.org.uk/
resource-detail/1059/>. Accessed August 16, 2015.
Beirne, C. and Lambin, X. (2013). ‘Understanding the determinants of
volunteer retention through capture-recapture analysis: Answering
social science questions using a wildlife ecology toolkit’. Conservation
Letters 6: 391–401.
Bryce, R., Oliver, M.K., Davies, L., Gray, H., Urquhart, J. and Lambin,
X. (2011). ‘Turning back the tide of American mink invasion at an
unprecedented scale through community participation and adaptive
management’. Biological Conservation 144: 575–583.
Forest Research. (2008). A Survey and Appraisal of Rhododendron
Invasion in Argyll and Bute. Roslin, Midlothian, UK:Forest Management
Division, Northern Research Station.
Fraser, E.J., Macdonald, D.W., Oliver, M.K., Piertney, S.B. and Lambin,
X. (2013). ‘Using population genetic structure of an invasive mammal
to target control efforts – an example of the American mink in
Scotland’. Biological Conservation 167: 35–42.
GBNNSS. (2008). The Great Britain Invasive Non-native Species
Strategy. Sand Hutton, York, UK: The Great Britain Non-native Species
Secretariat, Animal and Plant Health Agency.
Kent, M. and Coker, P. (1992). Vegetation Description and Analysis: A
Practical Approach. New York: John Wiley and Sons..
Lambin, X., Horrill, J.C. and Raynor, R. (2019). ‘Achieving large scale,
long-term invasive American mink control in Northern Scotland
despite short-term funding’. In: C.R. Veitch, M.N. Clout, A.R. Martin,
J.C. Russell and C.J. West (eds.) Island invasives: scaling up to meet
the challenge, pp. 651–657. Occasional Paper SSC no. 62. Gland,
Switzerland: IUCN.
Marine Scotland. (2017). An Analysis of the Value of Wild Fisheries in
Scotland.Edinburgh, UK: Marine Scotland.
Melero, Y., Robinson, E. and Lambin, X. (2015). ‘Density-and agedependent reproduction partially compensates culling efforts of invasive
non-native American mink’. Biological Invasions 17: 1–13.
Oliver, M.K., Piertney, S.B., Zalewski, A., Melero, Y. and Lambin,
X. (2016). ‘The compensatory potential of increased immigration
following intensive American mink population control is diluted by
male-biased dispersal’. Biological Invasions 18: 3047–3061.
Raynor, R., Horrill, C., Taylor, M. and Taylor, K. (2016). ‘American
Mink. Version 1.0.’. In: M.J. Gaywood, P.J. Boon, D.B.A Thompson
and I. Strachan (eds.) The Species Action Framework Handbook, p. 27.
Perth, Scotland: Scottish Natural Heritage.
Reynolds, J.C., Short, M.J. and Leigh, R.J. (2004). ‘Development of
population control strategies for mink Mustela vison, using floating
rafts as monitors and trap sites’. Biological Conservation 120: 533–543.
Ruiz-Suarez, Melero, Y., Giela, A., Henriquez-Hernandez, L.A., Sharp,
E., Boada, L.D., Taylor, M.J., Camacho, M., Lambin, X., Luzardo,
O.P. and Hartley, G. (2016). ‘Rate of exposure of a sentinel species,
invasive American mink (Neovison vison) in Scotland, to anticoagulant
rodenticides’. Science of the Total Environment 569–570: 1013–1021.
SEPA. (2009a). Scotland River Basin Management Plan 2009–2015.
Scottish Environment Protection Agency. <www.sepa.org.uk>.
Accessed 13 August 2017.
SEPA. (2009b). Solway Tweed River Basin Management Plan 2009–
2015. Scottish Environment Protection Agency. <www.sepa.org.uk>.
Accessed 13 August 2017.
SNH. (2011). Scottish Recreation Survey, Annual Summary Report. SNH
Commissioned Report 535. <www.snh.gov.uk>. Accessed 13 August
2017.
Tweed Forum. (2006). The Tweed Invasives Project. Melrose, UK: Tweed
Forum.<http://www.tweedforum.org/publications/tweed-invasives>.
Accessed 13 August 2017.
UKTAG. (2015). ‘Revised classification of aquatic alien species according
to their level of impact’. Aquatic alien species and the WFD: proposed
amendments to the impact classification in the UKTAG alien species
guidance., UK: Environment Agency. <www.wfduk.uk>. Accessed 15
August 2017.
Williams, F., Eschen, R., Harris, A., Djeddour, D., Pratt, C., Shaw, R.S.,
Varia, S., Lamontagne-Godwin, J., Thomas, S.E. and Murphy, S.T.
(2010). The Economic Cost of Invasive Non-Native Species on Great
Britain. CAB/001/09. UK: Centre for Agriculture and BioSciences
International. <www.nonnativespecies.org>. Accessed 14 August 2017.
Yamaguchi, N., Rushton, S. P. and Macdonald, D. W. (2003). ‘Habitat
preferences of feral American mink in the Upper Thames’. Journal of
Mammalogy 84: 1356–1373.
465
466
Chapter 3: Strategy
With Sections: A Biosecurity
B Collaboration
C Outcomes
D Scaling up
467
J.R. Balchin, D.G. Duncan, G.E. Key and N. Stevens
Balchin, J.R.; D.G. Duncan, G.E. Key and N. Stevens. Biosecurity on St Helena Island – a socially inclusive model for protecting small island nations from invasive species
Biosecurity on St Helena Island – a socially inclusive model for
protecting small island nations from invasive species
J.R. Balchin1, D.G. Duncan1, G.E. Key2 and N. Stevens1
Agriculture and Natural Resources Division, Scotland, St Helena Island STHL1ZZ. 2GB Non-Native Species
Secretariat, Animal and Plant Health Agency, Sand Hutton, York YO411LZ, UK. <gillian.key@apha.gov.uk>.
1
Abstract St. Helena Island, 122 km2 (47 sq. miles) is a UK Overseas Territory in the South Atlantic. It is a remote volcanic
island situated in the sub-tropics 1,127 km (700 miles) from Ascension Island and 2,736 km (1,700 miles) from South
Africa. Its resident population of ca. 4,500 is serviced by a single supply ship which visits up to 25 times a year. Isolation has
acted historically as a natural barrier to pest arrival and border control has followed the conventional practice of protecting
agricultural interests through restrictions on fresh produce, plant materials, livestock and pets. The benefits of isolation
were compromised in 2016 when the first airport opened. Private jets arrive now from Africa, Europe and South America,
and commercial flights started at the end of 2017. A programme of biosecurity capacity building and strengthening
was established in anticipation of this air traffic. St Helena authorities introduced a national biosecurity framework and
associated policy (entitled Biosecurity St Helena), the latter constructed through multi-sectoral consultation, and key
stakeholders participated throughout in policy development. Biosecurity St Helena applies international standards set
by the International Plant Protection Convention across the biosecurity continuum. As is typical in small island nations,
human and financial resources are limited, so that the biosecurity strategy addresses mainly higher risks. Compliance is
heavily reliant on public awareness. Active communication engages all community sectors in biosecurity work through
education, information, advocacy and feedback. Authorities use key performance indicators to measure the effectiveness
of this approach. Biosecurity St Helena is a model of actively socialised biosecurity for other small island nations.
Keywords: biosecurity strategy, capacity building, community engagement, inclusive planning, small island nation
INTRODUCTION
St Helena Island, a United Kingdom Overseas Territory
in the South Atlantic Ocean, is a volcanic island with an
area of 122 km2 (47 square miles) and total population of
4,534 (St Helena Government, 2016). St Helena is remote
and isolated, lying 1,127 km (700 miles) southeast of
Ascension Island and 2,736 km (1,700 miles) from South
Africa, with a sub-tropical, maritime climate. A total of 502
endemic species are currently known, comprising around
one third of the total endemic biodiversity of the UK
Overseas Territories and making a significant contribution
to global biodiversity (Churchyard, et al., 2014). The
economy is based mainly on agriculture, fishing, a small
but growing volume of tourism and income from offshore
employment.
The Government’s Agricultural and Natural Resources
Division (ANRD) reviewed biosecurity practices in in 2013
(Key, 2013) and concluded that capacity was inadequate to
address the new biosecurity pressures associated with air
access. Lack of biosecurity-specific legislation or overall
operational framework severely compromised post-border
controls and enforcement. There were no fumigation or
other specialist facilities for local treatment of contaminated
goods. Tellingly, the common interests of different sectors,
particularly agriculture, public health and the environment
were not harmonised for biosecurity purposes.
Until 2016 the only regular access to St Helena was via
the Royal Mail Ship (RMS) St Helena, calling around 25
times a year in passage from Ascension Island and South
Africa. Most commodities are imported. More than 69% of
the island’s annual requirements for agriculture and food
are sourced from South Africa, including almost all fruit,
and significant quantities of vegetables.
The St Helena Government’s programme to upgrade
biosecurity in 2013 departed from the existing emphasis
on managing agricultural and animal imports at the border.
It moved biosecurity to a risk-based approach across the
broader continuum of invasive pest organisms in marine
and terrestrial environments. Interception measures preborder, at-border and post-border were to be more closely
integrated. Resource limitations in the small-island context
argued for greater investments in pre-border controls and
post-border surveillance.
In 2010 the UK Government announced its intention
to build an airport on St Helena, conditional upon the St
Helena Government’s commitment to internal investment
and increased tourism. Air access and expansion of the
tourism sector augmented biosecurity risks to St Helena.
To meet the challenge, the St Helena Government launched
a programme in 2013 to upgrade biosecurity arrangements.
The 2009 South Atlantic Invasive Species Strategy had
already defined dedicated biosecurity capacity for St
Helena as a strategic priority for the prevention of invasive
species and unwanted organisms in the region (Shine &
Stringer, 2010).
St Helena has a limited range of existing cosmopolitan
pests and is very vulnerable to new introductions harmful
to the economy, community health, environment and the
new investments in tourism development (Pryce, 2015).
Until now, biosecurity has relied heavily on its isolation as
an oceanic island and limited modes of entry to minimise
exposure to new pest threats.
PREPARING FOR NEW BIOSECURITY
MEASURES
Approaches to building the new biosecurity
framework
The ANRD led a new policy team comprising agency
representatives of Environment, Customs and Public
Health; the first time this multi-sector team had been
brought to the same table. Their purpose was to establish
the architecture of the new biosecurity system through an
overarching policy statement. The policy team recognised
that understanding of biosecurity issues was essential
for community buy-in and compliance. Accordingly,
the team developed the new biosecurity policy in full
consultation with all sectors in the community, from
farmers to politicians, coupled with close participation
throughout the reform process. Stakeholders were given
multiple opportunities to discuss new ideas and to object
to them if warranted. The policy team intended biosecurity
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
468
up to meet the challenge, pp. 468–472. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Balchin, et al.: Biosecurity on St Helena Island
awareness in the community, as a whole, to benefit from
these approaches.
Consultations, commencing with twelve focal
groups of stakeholders in 2013, explored attitudes to the
current biosecurity procedures. Despite some criticisms,
all 54 stakeholders that were consulted supported the
current arrangements and the need to strengthen them in
anticipation of air access.
BIOSECURITY ON THE GROUND
Participants in a subsequent workshop agreed on the
vision for biosecurity policy and then defined strategic
objectives and expected outcomes. The broader public
were invited to consider the resulting policy statement.
The St Helena Government endorsed the policy after it
had been revised to incorporate feedback. Now entitled
Biosecurity St Helena, the policy was launched officially
in the biosecurity facility at the seaport in November 2014
(St Helena Government, 2014).
At the border
Today, ANRD is the agency lead for biosecurity. It
holds the authority to approve import licences and has
the principal duty to launch responses to incursions. The
community relies on ANRD for government leadership
in matters of compliance and enforcement of the new
biosecurity legislation.
Biosecurity St Helena defined
The policy vision and principles govern the new
biosecurity arrangements and are supported by the
island’s legal and institutional structures. The policy is the
blueprint for “an effective biosecurity system of shared
responsibility that protects the sustainable future of our
island environment, allowing a vibrant economy, safe
movement of people and goods, and enhanced livelihoods
and health” (St Helena Government, 2014, page 3).
Overarching outcomes are:
● Effective management of biosecurity risks to St
Helena’s environment, agriculture, amenities, public
health and well-being, including safety;
● Effective governance of St Helena’s biosecurity
system through shared responsibility and roles
Biosecurity St Helena recognises that a zero-risk
approach is not practical and works to reduce the risk to
an acceptably low level. The policy endorses a white list
and licencing approach, whereby all high-risk goods are
prohibited except those for which import health standards
have been developed. Import health standards specify
the conditions under which goods can be imported and
the treatments required in response to pest organisms
intercepted or simply suspected pre-border, at the border
or post-border.
Six crucial principles guide biosecurity work:
1. Leadership for effectiveness
biosecurity apparatus
throughout
the
2. Clear communication of stakeholder roles,
responsibilities, and the ‘what, why and how’ of
biosecurity investments.
3. Shared responsibility across all sectors and interests
for mutual benefit
4. Risk-based responsiveness to the probability of
border challenges, potential harm and changes in the
nature of threats.
5. Evidence-based decision-making supported by
quality systematic research
6. Co-operation between sectors to minimise the
probability of new incursions and manage existing
ones.
A multi-sector plan was developed alongside the
biosecurity policy to put the new structures in place. The
policy team supervised developments for the first year, and
thereafter improvements were mainstreamed into the work
plans of ANRD, Environment, Customs and Public Health,
taking effect in the 2016/2017 financial year.
The St Helena Government recruited two full-time
biosecurity officers in 2015, the first in the island’s
history. They work closely at the border with Customs
whose warrants they also hold. Customs and Immigration
officials received the same biosecurity training to ensure
harmonised border security.
Import health standards apply for a range of
commodities, and the island’s main traders assisted the
development of these standards. Inspection procedures now
align with international standards set by the International
Plant Protection Convention (IPPC) for phytosanitary
(plant health) risks and the World Organisation for Animal
Health (OIE) for zoosanitary (animal health) imports.
Inspection practices are codified for consistency and
transparency. Import Health Standards, application forms
and general guidance are now available on-line at <http://
www.sainthelena.gov.sh/st-helena-biosecurity-service/>.
The officers employ a dog trained to detect honey,
bananas and citrus, to protect St Helena’s disease-free bees
and bananas. Likewise, citrus (commonly intercepted on
incoming visitors) may introduce newly emerging diseases
such as huanglongbing citrus greening. Dog handling at
the border is governed by a Standard Operating Procedure
written together with Customs who run their own detection
dogs.
The full-time team has extended biosecurity operations
on the wharf, beyond the former pre-occupation with
fresh produce. Customs help with passenger and cargo
profiling so that higher risks can be ranked for quarantine
inspections. Profiling relies on interception data for visitors
and surveys of imported cargo arriving by sea, and will be
refined as data accumulates for both visitors and freight
arriving by air.
Personal goods in shipping containers and vehicles
shipped in break-bulk were predicted to be high risk
freight. Between January 2016 and March 2017, 99
(40.4%) of 245 imported vehicles (mostly cars) were
found to be contaminated with soil. Inspectors intercepted
75 live spiders in 16 (6.5%) of the vehicles. Over the same
period, 23 live spiders were intercepted in four (11.8%) of
the 34 incoming containers of personal goods. The spiders
belonged to seven species known from the UK and South
Africa. They were found mainly in the space behind vehicle
wing mirrors, on the windscreen wipers, and behind the
rear-mounted spare wheel on SUVs. Most spiders in
shipping containers were discovered immediately inside
the doors.
Soil samples collected from vehicles (typically from
rear wheel-arches) were weighed, then placed in seed
trays for up to two months to check for seed germination.
A mixture of grasses and small dicotyledons germinated
successfully from nine (9.1%) of the 99 samples but none
survived long enough to identify species.
Building and operating St Helena airport
Construction of St Helena airport commenced in
January 2012. Three new biosecurity pressures had to be
managed
469
Island invasives: scaling up to meet the challenge. Ch 3A Strategy: Biosecurity
A second supply vessel now visited every six weeks or
so until October 2015. The ship departed from a new port
of origin and was the first vessel able to moor alongside
the island at a specially constructed wharf. The normal
supply vessel RMS St Helena barged freight ashore from
an anchorage in the bay.
The new vessel discharged large quantities of
construction materials, including river and dune sand.
Several hundred off-shore workers arrived (mainly from
Africa and Thailand), for whom biosecurity awareness was
low to zero.
ANRD negotiated quarantine agreements with the South
African construction company. Consignments of sand were
fumigated in Namibia and inspected on arrival in the port
area. The team inspected break-bulk consignments before
disembarkation from the vessel. Compliance improved to
a good standard after some initial teething troubles. Only
two pests were intercepted during the construction phase
- flattened giant dung beetles (Pachylomera femoralis
Coleoptera: Scarabaeidae) on open metal gantries; and
ice plant, (Galenia papulosa Aizoaceae) in river sand.
Construction staff were quick to report biosecurity issues
and responded appropriately.
Border and biosecurity officials meet all inbound
flights. Airport biosecurity is guided by a Standard
Operating Procedure refined through preliminary test-runs
with flights and arriving passengers. An x-ray scanner
screens all in-bound baggage. Fresh produce is examined
in a small, sole-purpose biosecurity room in the airport’s
cargo compound.
The Public Health Committee obliges ‘disinsection’ of
all inbound flights, recognising known risks of introducing
aerial insect vectors such as mosquitoes (Gratz, et al.,
2000). Eighteen private jet and three medevac flights had
been treated by March 2016. Commercial flights had not
yet commenced.
Post-border surveillance
The 2013 review of biosecurity (Key, 2013) revealed
serious weaknesses in post-border surveillance for
pest species by-passing earlier lines of defence. Today,
monitoring and surveillance behind the island’s borders
are structured to detect and eradicate pest intruders before
they can establish. Biosecurity staff direct their attention
to surveillance at the airport construction site and all other
ports of entry; targeted surveillance for introduced tephritid
fruit flies; and readiness to respond to pest detections.
Surveillance at the construction site has mapped every
location at which shipping containers were landed or opened
with the participation of the construction company. At each
location, the biosecurity team installed a monitoring point
comprising a covered breeze block in which crumpled
newspaper and a sticky trap attract and contain unwanted
invertebrates. A monitoring protocol, identification guide
and reference collection assist surveillance. As construction
wound down in 2016, monitoring was migrated to new
sites around the two seaports and the airport. Each station
will include mosquito traps in the future.
Surveillance operates pheromone-baited sticky traps
for five species of economically harmful tephritid fruit flies
at ten pivotal fruit-growing sites across the island.
The biosecurity team have engaged relevant
stakeholders in the preparation of nine response plans for
incursion emergencies. The plans address terrestrial and
marine risks from a range of phytosanitary, zoosanitary and
invasive non-native species. They were refined through a
simulation exercise, and further exercises are planned for
the future.
470
Engaging the community
The principle of responsibility shared universally by the
St Helena community and visitors is central to biosecurity
arrangements. But policy consultations with stakeholders
revealed poor understanding of what biosecurity is and
what the biosecurity team does. In response, a multi-sector
communication strategy targeted key audiences with
biosecurity messaging. The strategy adopted Border Security
(a popular TV programme on Australian border security
services) as its brand but switched later to Biosecurity St
Helena to align messages with the new biosecurity policy.
Biosecurity St Helena branded pens, shopping bags and
mugs were a popular means of reinforcing the messages.
Outreach comprises a programme of press releases, articles
in the local print and radio media, activities with local
primary and secondary schools, and visits by groups to
observe biosecurity inspections at the wharf. Councillors,
government officials and airport officials were among the
first groups invited. The outreach programme continues as
a core element in the biosecurity team’s work plan. New
stakeholders involved in air access readily embraced the
messaging, which focused on collective responsibility for
protecting the island for the future.
Site visits were very productive; feedback was positive
from visitors who were not previously aware of the wharf
facility or only generally familiar with the biosecurity
team’s functions.
The public are actively encouraged to be vigilant for
new invasive non-native species. In March 2015, a public
awareness campaign comprising press announcements and
leaflets invited the public to report unusual tracks, signs,
weeds or invertebrates. Reporters are rewarded with a gift
of branded promotional goods.
Sustaining external support for Biosecurity St Helena
is a priority. Biosecurity reaches well back into the supply
chains through visits to overseas agents and suppliers who
are expected to comply with stringent, time-consuming
or costly quarantine requirements often for commodity
quantities small relative to their normal trade volumes.
Face-to-face contact with suppliers and South African
Cape Inspection Service aims to translate their goodwill
into co-operation, especially for frequently imported
high-risk goods such as South African, produce and plant
propagation materials.
Measuring success
A comprehensive database records imports and
interceptions. Another holds baseline data for all taxa of
native and introduced species, together with a reference
collection of pest species known on the island. ANRD uses
these data to measure biosecurity outcomes and assess
threats based on empirical evidence.
Even so, establishing meaningful indicators to measure
biosecurity effectiveness is a challenge. The number of
interceptions is a commonly used metric, but one open to
confounding interpretation: does an increase in the number
of interceptions indicate (i) a decrease in effectiveness (i.e.
more introductions arriving owing to poorer pre-border
measures) or (ii) an increase (more interceptions owing to
better inspection practices)?
To resolve this ambiguity, ANRD uses five key
performance indicators based on the notion of tolerance
thresholds for interceptions. Once a threshold is exceeded,
the biosecurity team investigates likely causes and applies
appropriate remedies.
The indicators relate to passenger, fresh produce and
cargo pathways arriving by sea (Table 1). Table 2 shows
the biosecurity performance results for the 2016 calendar
year. None of the thresholds was exceeded in any indicator.
Balchin, et al.: Biosecurity on St Helena Island
Table 1 Tolerance thresholds used as measures of biosecurity performance for three main risk-pathways on St Helena
Island.
Pathway
Percentage of passengers arriving without a
quarantine-risk item in their baggage
Threshold
No more than five in every
100 passengers arrive with
prohibited goods such as honey,
fruit, nuts
1
Percentage of fresh-produce lots inspected which No more than five in every 100
do not conceal a quarantine pest
lots inspected have a quarantine
pest (dead or alive)
Number of quarantine pests detected at the border No more than three quarantine
as a percentage of the total number of imported
pests detected for every 100
shipping containers and uncontainerised vehicle
units
of any type
Number of quarantine pests detected post-border No more than two quarantine
as a percentage of the total number of imported
pests detected post-border for
shipping containers and uncontainerised vehicle
every 100 units
of any type
Number of animals breaching border biosecurity No more than five in 100
requirements as a percentage of total animal
animal imports fail to satisfy
imports
requirements
Breaches include identity issues, disease, or
incorrect paperwork
Notes
Includes passengers and crew
on RMS St Helena and private
yachts, but excludes dayvisitors on cruise ships
A “lot” is defined as the total amount of any one type of produce which are clearly from the same source.
1
Table 3 lists commodity types by passengers’ reasons
for visiting. The ‘other’ class of passenger was most
likely to arrive with prohibited goods. This class includes
construction workers possessing few or no English
language skills and therefore less likely to have understood
the biosecurity arrival information provided on the ship.
Fresh produce was most frequently seized, typically apples,
pears, oranges and other citrus picked off the breakfast
table in the ship before disembarkation.
strong biosecurity system through active participation
is particularly important. Biosecurity St Helena is a
relatively short and succinct blueprint which could have
been constructed quite quickly. Instead, the St Helena
Government chose purposefully to pursue a process of
intensive consultation which extended preparation over
a period of nearly a year. Thus, the drafting process was
considered as important for social acceptance as the
resulting document. Local priorities and concerns are now
reflected in the language and layout of the plan.
DISCUSSION
Most importantly, Biosecurity St Helena demonstrates
the benefits of political will to integrate agricultural
and environmental interests for biosecurity purposes.
Limitations on human resources common to small island
states have been largely overcome on St Helena by close
co-operation between biosecurity and customs officials.
Biosecurity compliance and enforcement can be
challenging in small, isolated communities if stakeholders
are not willingly engaged through knowledge of need
and benefit. Socialising the processes of building a
Table 2 Results for five key performance indicators for the 2016 calendar year in which the RMS St Helena,
186 yachts and eight cruise ships visited.
Indicator
Threshold
2016
Percentage of passengers arriving
without a quarantine-risk item in their
baggage
95%
98%
Percentage of fresh-produce lots
inspected which do not carry a
quarantine pest
95%
Number of quarantine pests detected at
the border as a percentage of the total
number of units imported
3
Number of quarantine pests detected
post-border as a percentage of the total
number of units imported
Number of animals that breach
border biosecurity requirements as a
percentage of total animal imports
2
5%
Data
3,930 passengers arrived
469 items confiscated from 60 passengers, of which
76% were fresh produce, 1% honey, and 24% other
items
97.4% 62 phytosanitary import licences issued”
366,085 kg fresh produce and 16,050 kg seed
potatoes imported
536 lots inspected
1.1
1,023 containers and 250 vehicles imported
14 interceptions, of which
4 were tephritid larvae,
8 Lepidoptera larvae and
2 other taxa
0.1
1 interception: a chafer beetle
0%
42 animal import licences issued
471
Island invasives: scaling up to meet the challenge. Ch 3A Strategy: Biosecurity
Table 3 Goods seized from passengers (n = 60) by
purpose of visit in 2016. Some passengers imported
more than one type of risk item.
Purpose of visit
Returning resident
Tourist
Government worker
Other
Type of items seized
Honey
Fresh
Other
produce produce
0
12
7
0
5
3
1
4
3
1
18
10
Extending sea-port biosecurity vigilance to high-risk
shipping containers and vehicles revealed their prominence
as vectors for harmful hitch-hikers such as spiders. This
had not been known before.
Two main weaknesses remain in the island’s
biosecurity framework. First, new biosecurity legislation
has been delayed by other priorities in the Attorney
General’s Office. In the meantime, existing statutes and
regulations are neither harmonised nor aligned with
international biosecurity expectations, so that Biosecurity
St Helena lacks explicit legal mandates for compliance and
enforcement. Warnings must substitute for fines and other
legal sanctions, a shortfall which is disadvantageous to the
new system.
Second, import risks are not yet assessed systematically
or comprehensively. The new biosecurity policy requires
all produce or risk material not on the white list (i.e. not
subject to agreed import health standards) to be submitted
for import risk assessment (IRA) but, in common with
many small island nations, St Helena lacks the domestic
technical expertise to apply the international guidelines on
pest risk analysis (IPPC, 2017). Biosecurity officers cannot
refer to specialist networks for advice on risk likelihood
and impact, assessments of which are required at each level
of the IRA process. They are often too busy to attempt
these formal assessments themselves. Yet, under pressure
of requests to import new commodities, St Helena’s
biosecurity officers regularly have to make such decisions.
Pragmatic guidelines for IRA are being applied in the
interim. Risk evaluation for familiar commodities can
rely on levels of confidence acquired through practical
experience and knowledge of their points of origin. For
example, fresh produce from South Africa, vehicles from
Ascension Island, UK or South Africa, and selected plant
propagating materials from the UK or South Africa are
relatively well known and already have import conditions
defined for them. These conditions must be revised if the
risk profile alters through, for example, a change in pathway
or reports of a new pest or disease in the country of origin.
For commodities of these sorts, the biosecurity team assess
new risks using simple web-based resources such as the
CABI Invasive Species Compendium (<https://www.
cabi.org/isc/>) and CABI Crop Protection Compendium
(<https://www.cabi.org/cpc/>).
Biosecurity St Helena does not have risk assessment
measures in place for unknown pathways or commodities
such as novel plant or animal species imported for
propagation or breeding. These are highly concerning and
challenging to address.
The key factor for success in socialising biosecurity is
considered to be the amount of time and effort committed
to listening, talking and responding to the community,
from farmer to government official, and utilising a range
of communication media. No attempt was made to directly
tackle the few more resistant individuals with arguments.
472
It was found that time and peer pressure were in most cases
sufficient to bring them round, and compliance was high.
In conclusion, St Helena Island faces increasing
pressures from invasive species and is typical of small
island nations in having too few resources to cope. Despite
this, it has risen to the challenge and has in place a model for
autonomous biosecurity by a small island nation. What has
made this possible – and what compensates so significantly
for chronic resource stresses – is the decision to engage
business and local communities in developing Biosecurity
St Helena and sustaining it day-by-day. Harmonising of
public services has been highly effective. Recognising that
a solely official approach to Biosecurity St Helena would
lack necessary resilience and buy-in, the socialising of
biosecurity is what makes Biosecurity St Helena a model
for other small island nations.
ACKNOWLEDGEMENTS
The authors would like to acknowledge the St Helena
community for being part of protecting the island’s future.
We are also grateful to the anonymous reviewers of this
paper for their invaluable comments and suggestions for
improving the text.
REFERENCES
Churchyard, T., Eaton, M., Hall, J., Millett, J., Farr, A., Cuthbert, R. and
Stringer, C. (2014). The UK’s Wildlife Overseas: A Stocktake of Nature
in our Overseas Territories. Sandy, UK: RSPB.
Gratz, N.G., Steffen, R. and Cocksedge, W. (2000). ‘Why aircraft
disinsection?’ Bulletin of the World Health Organization 78 (8): 995–
1004.
IPPC (2017). ‘International Standards for Phytosanitary Measures, ISPM
No. 11’. Pest Risk Analysis for Quarantine Pests, Including Analysis
of Environmental Risks and Living Modified Organisms. Produced by
the Secretariat of the International Plant Protection Convention Adopted
2013. <https://www.ippc.int/static/media/files/publication/en/2017/05/
ISPM_11_2013_En_2017-05-25_PostCPM12_InkAm.pdf >.
Key, G.E. (2013). Biosecurity Status Report. St Helena: Agriculture and
Natural Resources Division, St Helena Government.
Pryce, D. (2015). ‘St Helena Pest and Biocontrol Species List’. <http://
www.sainthelena.gov.sh/integrated-pest-management/ >.
Shine, C. and Stringer, C. (2010) South Atlantic Invasive Species Strategy
and Action Plan. Based on the Contributions by Participants to a
Regional Meeting, Ascension Island, 14–19 May 2009. Unpublished
report.
St Helena Government (2014). Biosecurity St Helena. A Policy for All
Members of the St Helena Community. St Helena: Agriculture and
Natural Resources Division, St Helena Government.
St Helena Government (2016). St Helena 2016 Population & Housing
Census. Summary Report. St Helena: St Helena Statistics Office, St
Helena Government.
C.L. Boser, P. Power, A. Little, J. Matos, G.R. Howald, J.M. Randall and S.A. Morrison
Boser, C.L.; P. Power, A. Little, J. Matos, G.R. Howald, J.M. Randall and S.A. Morrison. Proactive
planning and compliance for a high-priority invasive species rapid response programme
Proactive planning and compliance for a high-priority invasive species
rapid response programme
C.L. Boser1, P. Power2, A. Little3, J. Matos1, G.R. Howald4, J.M. Randall1 and S.A. Morrison1
The Nature Conservancy, 201 Mission St. 4th floor, San Francisco, CA 94105. <cboser@tnc.org>. 2National Park
Service, 1901 Spinnaker Dr., Ventura, CA 93001. 3U.S. Fish and Wildlife Service, 1901 Spinnaker Dr., Ventura, CA
93001. 4Island Conservation, 2100 Delaware Ave, Santa Cruz, CA 95060.
1
Abstract The California Channel Islands (USA) are home to numerous endemic taxa and provide refugia for many
Californian coastal species. Over recent decades, land managers have made considerable progress in eliminating alien
invasive mammals and non-native invasive plants and ants. The California Islands Biosecurity Group aims to protect that
investment, by coordinating and finding economies of scale in efforts to prevent invasive species introductions, detect
incursions, and to respond rapidly and successfully to incursions that do occur. The highest priority is the prevention or
early detection and elimination of Rattus introductions to any of the rat-free islands or islets in the archipelago. The rapidity
of a response to a suspected incursion of rats could be hindered not only by the time required to detect it and mobilise
the field effort, but also by the time required to secure regulatory permissions to implement the pre-planned actions. To
address these issues, we are proactively obtaining necessary permits and developing strategy that could allow response
action within 36 hours of a known or suspected introduction of Rattus. Our process includes developing a flowchart to aid
managers in determining the appropriate management response, which considers seasonality, location, and potential nontarget taxa at the incursion site. Completing the necessary compliance process proactively, and maintaining a ready stock
of the materials needed to equip a management response, reduces the time between detecting an incursion and being able
to respond to it and thereby increases our ability to protect the conservation value of – and investment in – these islands.
Keywords: biosecurity, California Channel Islands, eradication, planning, rapid response, rat introduction
INTRODUCTION
The extraordinary ecological, economic, and cultural
damage invasive alien species can cause on islands, and
the high costs and challenges of controlling or eradicating
these species, has led many island managers to place
greater attention on efforts to prevent biological invasions
in the first place (Ruiz & Carlton, 2003; Broome, 2007;
Rout, et al., 2011; Bassett, et al., 2016). Biosecurity
programmes are designed to proactively prevent invasive
species from arriving on an island, detect introductions
quickly, and prevent non-native species from establishing
on islands via rapid response actions (Broome, 2007;
Russell, et al., 2008a; Russell, et al., 2008b). The ability
to rapidly detect and respond to incursions is important for
a variety of reasons, including that small populations are
less expensive and more tractable to eradicate than large
ones; some invasive species may not be eradicable at all
with existing methods once even a small population has
become established. The question of “how rapid does rapid
need to be?” is to some degree determined by the biology
of the invader. The urgency of detecting and responding to
an invasive plant versus an insect or a rodent may vary, for
example, because their population growth and patterns of
spread may be different. Understanding those differences is
important for allocating scarce management resources, as
costs of surveillance are likely to increase if finer temporal
scale data are needed to provide adequate detection. Costs
of interventions may also increase as infestations increase
in area and become more established.
This paper focuses on one critical window of time
in a biosecurity strategy: the period between detecting a
potential incursion and responding to it with management
actions. Specifically, we provide a biosecurity programme
case study from an archipelago of islands in the United
States which has identified priority invaders and invasion
scenarios, proactively planned responses, and sought
environmental compliance permits in advance so that
planned actions can be carried out quickly if an incursion
is detected. Risks of rat invasion on islands are often
high, because rats are common stowaways on large boats.
If they successfully invade and establish populations
on some islands, then they may not be eradicable with
existing technologies for a variety of reasons, including
the potential that available treatment options may pose
unacceptable risks to populations of non-target native
species. Meanwhile, animals at very low abundance can
be difficult to detect (Morrison, et al., 2007; Russell, et
al., 2008a). Further compounding that challenge in the
context of biosecurity is that some species (e.g. rats) may
be unusually mobile and wide-ranging and exhibit other
unexpected behaviours in novel environments, and when
their numbers are low (Russell, et al., 2008a). Thus, if a rat
is detected or suspected (e.g. if there was a shipwreck of
a vessel known to be infested), there may be little time to
respond in a localised area with relatively high confidence
that the animal remains within the project area. A variety
of factors can limit response time, ranging from technical
(e.g. determining the specific methods which would be
most effective under the particular circumstances of the
incident), to operational (e.g. getting necessary materials
to the incident location), to administrative (e.g. who would
make decisions within the institutions with jurisdictions
over the proposed response). Environmental review
and compliance processes are time consuming for land
managers and, combined with required public comment
periods in the national permitting process, limit the ability
to respond rapidly. Here, we describe how the California
Channel Islands Biosecurity Working Group has taken
steps to improve the ability of conservation managers to
respond quickly to a rat incursion and improve their chances
of eliminating it. The most important of these steps are the
advance completion of required environmental review and
permitting, and the staging of materials necessary for a
response to a potential rat incursion.
SITE DESCRIPTION
The Channel Islands encompass eight islands ranging
in size from 250 ha to 25,000 ha, as well as numerous
islets around them, all of which are within 100 km of the
southern California mainland (USA). Five of the islands
are included in Channel Islands National Park (Park). The
US Navy owns one of the islands in the Park (San Miguel);
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 473–477. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
473
Island invasives: scaling up to meet the challenge. Ch 3A Strategy: Biosecurity
The Nature Conservancy owns 76% of the largest Park
island (Santa Cruz); and the National Park Service owns
the remainder. Four of the five Park islands are rat-free,
as is an islet separated from the fifth island by just 700 m.
The islands provide important nesting habitat for seabirds,
including five with either state or federal designations, and
one federally threatened shorebird (McChesney & Tershey,
1998; Howald, et al., 2005; CDFW, 2017a). They are home
to 14 federally threatened plant species and 23 endemic
animal species. The three largest islands in the Park are
home to the island fox (Urocyon littoralis), which was
recently removed from the endangered species list (USFWS
2016). The islands have been the focus of significant
conservation and restoration efforts over recent decades to
remove the most destructive introduced invasive species,
including pigs, goats, sheep, horses, donkeys, rabbits,
rats, cats, and ongoing efforts are underway to eradicate
Argentine ants (Linepithema humile) and 32 weed species
from Santa Cruz Island (Morrison, 2007, Morrison, 2011;
Cory & Knapp, 2014; McEachern, et al. 2016; Boser, et
al., 2017).
In order to protect the investments made and the
significant biological and cultural values of these islands
– and the broader archipelago in which they sit – the
California Islands Biosecurity Working Group (Group)
was established in 2012 (Boser, et al., 2014). The Group
is composed of biologists and managers from the federal
agencies, non-profit institutions, and partners that own
and or have management responsibility or investment in
the island resources. The Group meets quarterly in person
or by phone to share updates on obstacles, technological
and logistical advances to improve biosecurity, and to
suggest improvements to current biosecurity practices
and education programmes for professionals, visitors and
resident populations. A central organising principle of the
Group is that even though there are discrete islands with
various jurisdictions across the archipelago, a coordinated
and collaborative approach to biosecurity enhances
effectiveness and achieves a myriad of efficiencies (Boser,
et al., 2014; Matos, et al., in press). To accomplish an
economy of scale on biosecurity priorities and to ensure
that group objectives are met, the landowning entities
jointly fund a full-time position to lead the implementation
of group objectives. Staff biologists for the U.S. Navy, the
National Park Service and The Nature Conservancy form
sub-working groups to deliver results to jointly relevant
projects including camera traps to detect invasive species,
checks of boats and planes departing for the islands, and
development of educational materials and messaging. A top
priority of the Group is preventing the establishment of rats
on rat-free islands, because of the well documented threats
rats pose to island ecosystems, including on the California
Channel Islands (Atkinson, 1985; Campbell & Atkinson,
2002; Jouventin, et al., 2003; Howald, et al., 2005; Towns,
et al., 2006; Jones, et al., 2008; Banks & Hughes, 2012).
However, given current available technologies, significant
advance preparation would be required if we were to mount
a timely and effective response to a rat incursion.
THE URGENCY OF RESPONSE
A comprehensive biosecurity programme will include
measures to prevent incursion, as well as plans to respond
to an incursion should one occur. The rapidity of the
response action needs to be tailored to the species of
concern. For rats, field work conducted by Russell, et al.
(2008a) assessing the behaviour of collared individual rats
introduced to a rat-free island indicated that most stayed
within the introduction area for 2–3 days, but after that time
could rapidly move away from that site. We interpreted their
data to indicate that if a pregnant rat or a small population
of rats were introduced to an island, there would be a very
474
narrow window of time to take action in a localised area,
and expect with some certainty that the rat(s) would be near
the introduction point (Russell, et al., 2008a; Russell, et al.,
2008b). For incursions on large islands (e.g. an island like
250,000 ha Santa Cruz Island), once rats have dispersed
from the introduction point, we may consider them not
possible to remove with current technologies, or consider
that impacts to non-target species on an island-wide scale
may be too great. Only a few islands larger than 10,000 ha
have successfully completed a rat eradication (Howald, et
al., 2007), the largest effort being recently conducted on
South Georgia Island at approximately 390,000 ha (pers.
comm, T. Martin 2017), and each of these used broadcast
rodenticides. While rodenticides are currently the most
effective tool to eradicate rat infestations (Kaudeinen
& Rampaud, 1986; Tershy, et al., 1997; Howald, et al.,
2007), they can have substantial non-target impacts on
native species (Kaukeinen & Rampaud, 1986; Brown,
et al., 1988; Eason & Spurr, 1995; Eason, et al., 2002;
Howald, et al., 2010). This underscores the importance of
having confidence that a suspected rat infestation remains
contained within an area small enough so that risks to nontarget native species are acceptably low if rodenticides
are utilized. Quick deployment and response following an
introduction increases that confidence. Efforts to collect
data if rodenticides are used would include genetic analysis
of target carcasses collected during post-treatment actions,
collection and toxicology analyses of non-target carcasses
discovered for at least one year, and a comparison of nontarget population data pre- and post-treatment for at least
five years.
One programme planning goal for the Group is to
have the ability to react to an incursion within 36 hours.
In the United States, however, environmental review and
compliance documents permitting action that may impact
non-target species, such as rodenticide use in conservation
areas with sensitive or endangered species, can take
months to years to complete. Given the importance of both
the environmental review process and the need for a rapid
response after an incursion, we sought to undertake the
planning and permitting processes in advance, so that if an
incursion does occur we are prepared to react appropriately
and quickly.
ENVIRONMENTAL REVIEW AND COMPLIANCE
The United States’ 1969 National Environmental
Policy Act (NEPA) established a framework for protecting
the environment from ill-considered actions by ensuring
that federal agencies of the United States integrate
environmental values into their decision-making processes
before taking federal action. The Channel Islands National
Park and the U.S. Navy islands are federal properties and
therefore an environmental review must be completed
before actions that could impact their environments, such
as rodenticide use, could be taken. A NEPA document
contains assessments of alternative actions (referred to as
Alternatives) which could be implemented to achieve a
stated objective that are created in consultation with subject
matter experts. Further, the analysis chapter reviews the
impacts of alternatives on the natural and cultural resources
within the action area. A “preferred alternative” is selected
by the lead agency after the document is released for
public review and comment, revised if necessary, and
one alternative is selected for implementation by the lead
agency.
Prior to undertaking the environmental review process
and structuring an environmental compliance document,
we reviewed other scenarios that require proactive
planning such as emergency scenarios where human life
or property is at stake. In California, the Department of
Boser, et al.: Planning invasive species rapid response
Fish and Wildlife has published a California State Oil Spill
Contingency plan (CDFW, 2017b) that outlines and permits
actions required for rapid responses to oil spills. Similar to
a rat introduction on an island, the location of the oil spill is
not known during the planning process, so planning efforts
must include and address a variety of contingencies that
may be utilised depending on the timing and location of the
spill. In rat response planning and compliance documents,
we similarly need to evaluate locational- and seasonallydependent scenarios. This includes a thorough assessment
of how proposed actions may impact specific biological
and cultural resources on the five islands included in the
action area.
Although this proactive document must plan for the
introduction event in an unknown location and time, it
must nonetheless provide enough detail that the proposed
response actions can be thoroughly assessed by the public
and subject matter experts. Due to the ambiguity around
the time or location of an introduction, and the biological or
cultural resources that may be present at the site at any given
time, the alternatives in a proactive compliance document
must be structured differently than is typical in NEPA
documents. The preferred alternative must encompass all
feasible response actions, from the most minimal actions
such as deploying remotely triggered cameras, to setting
rodent traps, and/or broadcasting rodenticide. Proactively
permitting each of these actions would allow managers to
appropriately scale their response and utilise the tools that
are appropriate under the specific circumstances at that
time and place. The creation and use of a flowchart which
directs decision-makers to recommended actions based on
the known resources in a proposed project area and season
allows the managers of the incident to quickly identify
a recommended response. If agreed to by all consulting
agencies prior to the emergency, the flowchart could be
used to rapidly recommend response actions so they can
be approved and quickly enacted. A contact list, similar to
those used in incident response plans, must be created prior
to an incident to maximise the likelihood of rapid action.
Depending on the proposed action and the resources
in the affected area, additional federal and state laws may
apply. In the California Channel Islands, the protected
status of resident bald eagles (Haliaeetus leucocephalus)
and numerous protected migratory birds require managers
to adhere to regulations in the Bald and Golden Eagle
Protection Act (1940) and the Migratory Bird Treaty Act
(1918). The status of Channel Islands National Park as
proposed wilderness under the Wilderness Act (1964)
requires agency staff to complete an assessment of impacts
to “wilderness character” in a process structured similarly
to a NEPA review. The California Environmental Quality
Act (1970) is similar in scope to NEPA, and a California
Environmental Quality Act compliance document is
required if the project is conducted on state lands or if a
project uses state money. The federal Endangered Species
Act (1973) and the California Endangered Species Act
(1970) lists endangered and threatened species and
additional permitting may be required if these species
are present in the affected area. The sheer scope of the
assessments and review that must occur to adhere to
federal and state laws designed to protect natural resources
illustrates the need to develop a functional tool to
proactively gain consensus on the need to take emergency
actions to protect the resources these laws were designed
to protect.
CONSIDERATIONS IN PRE-STAGING
MATERIALS
The compliance document must describe and
accommodate assumptions about how tools needed
to respond to a rat incursion could be staged for rapid
deployment. For instance, if a potential action described in
the environmental review document calls for the use of a
conservation rodenticide, we could consider a brodifacoumbased conservation pellet designed for use on islands by
Bell Laboratory which has been shown to be effective at
eradicating rats from islands (Kaudeinen & Rampaud,
1986; Tershy, et al., 1997; Howald, et al., 2007). This
specialised bait must be ordered months before it could be
used, because it is only manufactured every few months.
The rodenticide loses palatability after one year, so the bait
must be properly disposed of and reordered annually. The
type of bait packaging we might use would depend on the
planned staging and deployment method approved in the
compliance document, whether it be broadcast deployment
and thus must be loaded into a hopper on the mainland and
slung out to the island preloaded or packaged for transport
by boat and loaded into a hopper on the affected island. For
implementation to go smoothly, contingency contracts for
services need to be in place prior to any incursion. A rapid
deployment of rodenticide to an incursion site is dependent
on all technical and logistical parts of an operation being
pre-approved and permitted.
ASSESSMENT OF NON-TARGET IMPACTS
The permitting documents required in the United States
include a description of projected impacts on non-target
species, services such as transportation, and systems such
as air and water quality. These anticipated impacts can
range from “none” to “major” (the latter typically defined
as population-level impacts). Monitoring programmes for
bald eagles, peregrine falcons (Falco peregrinus), island
foxes, island scrub-jays (Aphelocoma insularis), island
spotted skunks (Spilogale gracilis amphiala), and seabirds
are implemented annually on these islands, providing
us with data on the distribution and abundance of these
species on each island. Based on this detailed information,
we can annually estimate how many individuals and
what percentage of the population we could expect to be
impacted by response actions and then use this information
to recommend appropriate minimisation measures. The
estimates can assist managers in assessing the risk of
taking a specific action rather than taking no action at a
site. For instance, the 2016 fox density estimate on Santa
Cruz Island is approximately 10 foxes per square kilometre
(unpublished data, A. Dillon, Colorado State University).
If an area of 60 ha is treated, with the possibility of an
additional 40 ha within a 200 m buffer zone, the impacted
area may be as large as one square kilometre and
approximately 10 foxes may be impacted by broadcasting
bait. The island’s total fox population is estimated to be
2,100 foxes (unpublished data, A. Dillon, Colorado State
University) and thus we could expect that action to impact
0.5% of the total population. However, we could require
minimisation measures that could include fox trapping in
the affected area immediately before treatment to remove
as many individuals as possible for translocation to lower
density areas of the island. Island foxes are easily trapped
using box traps, and we expect we could remove as many
as 5–8 foxes from the affected area with just one night of
trapping. Similar calculations using known home range
data and population estimates for raptors could be used
to determine worst-case scenarios if rodenticide is to be
broadcast and also if bait stations are to be used. These
estimates would also include risk of transient birds entering
the treatment zone, possibly in response to availability of
contaminated carcasses. The output of that analysis may
assist managers in deciding on the best tool to use after an
incursion at a specific site. We are likely to assume 100%
mortality of the native mouse population in the project area
and a buffer zone if broadcast rodenticide is used. These
calculations and considerations must be built into an action
flowchart which could be used by managers and federal
475
Island invasives: scaling up to meet the challenge. Ch 3A Strategy: Biosecurity
agencies to assess the ecological cost of a broadcast baiting
response action relative to the likelihood and ecological
cost of a rat population establishing on the island.
CONCLUDING REMARKS
The benefits of proactive planning and rapid response
to biodiversity protection and ecosystem function are clearcut. Substantial evidence exists which suggests that if rats
establish themselves on new California Channel Islands,
they would have population-level effects on seabirds
(Howald, et al., 2005) and potentially many of the listed
plant species (Corry & McEachern, 2009). The approach
we used to increase the ability to undertake rapid response
– via proactive environmental review and compliance and
having at the ready the necessary materials to respond –
could be followed for other islands in the USA and, with
appropriate modifications, for islands in other nations.
Although we have made substantial progress in our
ability to respond rapidly to a rat detection, we still face a
significant limitation in our ability to detect new incursions
in a timely manner. This is due in large part to the size,
ruggedness, and inaccessibility of the Channel Islands.
Although we have a camera array at sites of suspected
higher risk of incursion, the frequency at which we can
retrieve data from these cameras, and process the images
they capture, represents an important weakness in our
current programme. We are hopeful that with emerging
technologies – in particular mobile and networking
technologies – we will be able to retrieve these data in real
time (Pimm, et al., 2015). Advances in machine learning
and image recognition (Lillesand, et al., 2014) also can
be applied to speed up processing of images and flag
suspicious images promptly to managers.
Rapidly developing technologies may play another
role as we adaptively improve our planning documents.
Specifically, we are considering the effect of emerging
molecular methods of rat control on how we might assess
risk and uncertainty, especially with regards to evaluation
of non-target impacts. For example, even though rats, if
established, would have negative impacts on many native
species of the islands, the most successful tool currently
used to eliminate rats on islands, rodenticide, has impacts
to a broad array of non-target taxa (Kaukeinen & Rampaud,
1986; Brown, et al., 1998; Eason & Spurr, 1995; Eason, et
al., 2002; Howald, et al., 2010). Alternative technologies
such as gene drives that produce “daughterless” offspring
may be available for use on invasive mammals in the
relatively near term (Regalado, 2017). Such technology
has already been developed for some species of insects
(Gantz, et al., 2015). While there remains uncertainty
about whether such technologies could accomplish
eradication objectives, or would be suitable for use in
low-density populations, substantial progress has been
made towards advancing the technology in the past two
years (Committee on Gene Drive Research in Non-Human
Organisms, 2016; Regalado, 2017). Clearly, there are
numerous and complex ecological, ethical, philosophical,
and policy issues associated with field application of these
technologies and identifying and resolving those issues
will present new planning and permitting considerations,
constraints and timelines. However, given the pace of
developing technologies, and the known concerns with
existing technologies, biologists should weigh current
ecological costs of action against the likelihood of future
technologies becoming available and the possibility
that their use would ultimately provide better ecological
outcomes. For example, a rapid response using existing
technologies might be advisable – even in the face of nontarget impacts – if catastrophic and irreversible damage to
native species, such as extinction or a severe reduction in
genetic diversity, were deemed likely to result before new
476
rat-control technologies could be available for use, even
under optimistic scenarios.
The proactive planning and permitting approach we
outline can be applied broadly to conservation challenges
that require the ability to respond rapidly to foreseen
episodic or biologically threatening emergencies. We
recognise that in island ecosystems, that experience high
extinction rates and frequent state changes, managers must
be nimble and quickly direct management actions for
the preservation of biodiversity. The model we outline to
proactively invest in planning, reviewing, and permitting
essential biosecurity response measures, will improve our
ability to protect the native biota of islands in the USA, if
not worldwide.
ACKNOWLEDGEMENTS
We thank the members of the California Islands
Biosecurity Working Group for their input into this
project and their dedicated work to promoting a culture of
biosecurity on the California Islands.
REFERENCES
Atkinson, I.A.E. (1985). ‘The Spread of Commensal Species of Rattus to
Oceanic Islands and their Effects on Island Avifaunas’. In: P.J. Moors
(ed.) Conservation of Island Birds: Case Studies for the Management
of Threatened Island Species, pp. 35–81. Cambridge: International
Council for Bird Preservation.
Banks, P.B. and Hughes, N.K. (2012). ‘A review of the evidence for
potential impacts of black rats (Rattus rattus) on wildlife and humans in
Australia’. Wildlife Research 39: 78–88.
Bassett, I.E., Cook, J., Buchanan, F. and Russell, J.C. (2016). ‘Treasure
islands: biosecurity in the Hauraki Gulf Marine Park’. New Zealand
Journal of Ecology 40: 250–266.
Boser, C.L., Cory C., Faulkner K.R., Randall J.M., Knapp J.J. and
Morrison S.A. (2014). ‘Strategies for biosecurity on a near-shore island
in California’. Monographs of the Western North American Naturalist
7: 412–420.
Boser, C.L. Hanna, C., Holway, D.A., Faulkner, K.R., Naughton, I.,
Merrill, K., Randall, J.M., Cory, C. Choe, D.-H. and Morrison, S.A.
(2017). ‘Protocols for Argentine ant eradication in conservation areas’.
Journal of Applied Entomology 141: 540–550.
Broome, K. (2007). ‘Island Biosecurity as a Pest Management Tactic in
New Zealand’. In: G.W. Witmer, W.C. Pitt, and K.A. Fagerstone (eds.)
Managing Vertebrate Invasive Species: Proceedings of an International
Symposium, pp. 104–106. Fort Collins, Colorado: National Wildlife
Research Center.
Brown, R.A., Hardy, A.R., Greig-Smith P.W. and Edwards, P.J. (1988).
‘Assessing the impact of rodenticides on the environment’. EPPO
Bulletin 18: 283–292.
Campbell, D.J. and Atkinson, I.A.E. (2002). ‘Depression of tree
recruitment by the Pacific rat (Rattus exulans Peale) on New Zealand’s
northern offshore islands’. Biological Conservation 107: 19–35.
[CDFW] California Department of Fish and Wildlife. (2017a). Natural
Diversity Database. Special Animals List. Periodic publication.
[CDFW] California Department of Fish and Wildlife. (2017b). California
State Oil Spill Contingency Plan. California Department of Fish and
Wildlife Office of Spill Prevention and Response.
Committee on Gene Drive Research in Non-Human Organisms. (2016).
‘Recommendations for Responsible Conduct; Board on Life Sciences;
Division on Earth and Life Studies’; National Academies of Sciences.
Washington, DC: National Academies Press.
Corry, P.M. and McEachern, A.K. (2009). ‘Patterns in Post-grazing
Vegetation Changes Among Species and Environments. San Miguel
and Santa Barbara Islands’. In: C.C. Damiani and D.K. Garcelon (eds.)
Proceedings of the 7th California Islands Symposium pp. 201–214.
Arcata, California: Institute for Wildlife Studies.
Cory, C. and Knapp J.J. (2014). ‘A program to eradicate twenty-four
nonnative invasive plant species from Santa Cruz Island’. Monographs
of the Western North American Naturalist 7: 455–464.
Eason, C.T. and Spurr E.B. (1995). ‘Review of the toxicity and impacts
of brodifacoum on non-target wildlife in New Zealand’. New Zealand
Journal of Zoology 22: 371–379.
Boser, et al.: Planning invasive species rapid response
Eason, C.T., Murphy E.C., Wright G.R.G. and Spurr, E.B. (2002).
‘Assessment of risks of brodifacoum to non-target birds and mammals
in New Zealand’. Ecotoxicology, 11: 35–48.
Russell, J.C., Beaven, B.M., MacKay, J.W.B., Towns, D.R. and Clout,
M.N. (2008a). ‘Testing island biosecurity for invasive rats’. Wildlife
Restoration 35: 215–221.
Gantz, V.M., Jasinskiene, N., Tatarenkova, O., Fazekas, A. Macias, V.M.,
Bier, E. and James, A.A. (2015). ‘Highly efficient Cas9-mediated
gene drive for population modification of the malaria vector mosquito
Anopheles stephensi’. Proceedings of the National Academy of Sciences
112: E6736–E6743.
Russell, J.C., Towns, D.R. and Clout, M.N. (2008b). ‘Review of Rat
Invasion Biology: Implications for Island Biosecurity’. Science
for Conservation 286. Wellington, New Zealand: Department of
Conservation.
Howald, G.R., Faulkner K.R., Tershy B., Keitt B., Gellerman H., Creel
E.M., Grinnel M., Ortega S. and Croll D.A. (2005). ‘Eradication of
black rat from Anacapa Island: biological and social considerations’. In:
D. Garcelon and K Schwemm (eds.) Proceedings of the Sixth California
Islands Symposium, pp. 299–312. Arcata, California: Institute for
Wildlife Studies.
Howald, G., Donlan, C.J., Galván, J.P., Russell, J.C., Parkes, J.,
Samaniego, A., Wang, Y., Veitch, D., Genovesi, P., Pascal, M., Saunders,
A. and Tershy, B. (2007). ‘Invasive rodent eradication on islands’.
Conservation Biology 21: 1258–1268.
Howald, G., Donlan, C.J., Faulkner, K.R., Ortega, S., Gellerman, H.,
Croll, D.A. and Tershy, B.R. (2010). ‘Eradication of black rats Rattus
rattus from Anacapa Island’. Oryx 44: 30–40.
Towns, D.R., Atkinson, I.A.E. and Daugherty, C.H. (2006). ‘Have
the harmful effects of introduced rats on islands been exaggerated?’
Biological Invasions 8: 863–891.
Tershy, B. R., Croll, D.A. and Howald, G.R. (1997). ‘Options for
Removing Introduced Black Rats (Rattus rattus) from Anacapa
Island, Channel Islands National Park’. Unpublished Report, Island
Conservation and Ecology Group, Santa Cruz, CA.
[USFWS] US Fish and Wildlife Service. (2016). ‘Endangered and
threatened wildlife and plants; Removing the San Miguel Island fox,
Santa Rosa Island fox, and the Santa Cruz island fox from the Federal
list of endangered and threatened wildlife, and reclassifying the Santa
Catalina Island fox from endangered to threatened’. Federal Register
81(156): 53315–53333.
Jones, H.P., Tershy, B.R., Zavaleta, E.S., Croll, D.A., Keitt, B.S.,
Finkelstein, M.E. and Howald, G.R. (2008). ‘Severity of the effects of
invasive rats on seabirds: a global review’. Conservation Biology 22:
16–26.
Jouventin, P., Bried, J. and Micol, T. (2003). ‘Insular bird populations can
be saved from rats: a long-term experimental study of white-chinned
petrels Procellaria aequinoctialis on Île de la Possession (Crozet
archipelago)’. Polar Biology 26: 371–378.
Kaudeinen, D.E. and Rampaud, M. (1986). ‘A review of brodifacoum
efficacy in the U.S. and worldwide’. In: T. Salmon (ed.) Proceedings of
the Twelfth Vertebrate Pest Conference, pp. 16–50. Davis: University
of California.
Lillesand, T., Kiefer, R.W. and Chipman, J., (2014). Remote sensing and
image interpretation. USA: John Wiley & Sons.
Matos, J., Little, A. Broome, K., Kennedy, E., Méndez Sánchez, F,
Latofski-Robles, M., Irvine, R., Gill, C., Espinzoa, A., Olthof, K.,
Ball, M., Howald, G. and Boser, C.L. (In press). ‘Connecting island
communities on a global scale: Case studies in island biosecurity’.
Western North American Naturalist.
McChesney G.J. and Tershy, B.R. (1998). ‘History and status of
introduced mammals and impacts to breeding seabirds on the California
Channel and northwestern Baja California Islands’. Colonial Waterbirds
21: 335–347.
McEachern, K., Atwater, T., Collins, P., Faulkner, K. and Richards, D.V.
(2016). ‘Managed Island Ecosystems’. In: H. Mooney and E. Zavaleta
(eds.) Ecosystems of California, pp. 755– 778. University of California
Press, Oakland, California.
Morrison, S.A. (2007). ‘Reducing Risk and Enhancing Efficiency in
Non-native Vertebrate Removal Efforts on Islands: A 25-year Multitaxa Retrospective from Santa Cruz Island, CA’. In: G.W. Witmer, W.C.
Pitt and K.A. Fagerstone (eds.) Managing Vertebrate Invasive Species:
Proceedings of an International Symposium, pp. 398–409. Fort Collins,
Colorado: National Wildlife Research Center.
Morrison, S.A., Macdonald, N., Walker, K., Lozier, L. and Shaw, M.R.
(2007). ‘Facing the dilemma at eradication's end: uncertainty of absence
and the Lazarus effect’. Frontiers in Ecology and the Environment 5:
271–276.
Morrison, S.A. (2011). ‘Trophic considerations in eradicating multiple
pests’. In: C.R. Veitch, M.N. Clout and D.R. Towns (eds.) Island
invasives: eradication and management, pp. 208–212. Occassional
Paper SSC no. 42. Gland, Switzerland: IUCN and Auckland, New
Zealand: CBB.
Pimm, S.L., Alibhai, S., Bergl, R., Dehgan, A., Giri, C., Jewell, Z., Joppa,
L., Kays, R. and Loarie, S. (2015). ‘Emerging technologies to conserve
biodiversity’. Trends in Ecology & Evolution 30: 685–696.
Regalado, A. (2017). ‘First gene drive in mammals could aid vast New
Zealand eradication plan’. MIT Technology Review [online periodical].
<www.technologyreview.com/s/603533/first-gene-drive-in-mammalscould-aid-vast-new-zealand-eradication-plan/>. Accessed 15 June,
2017.
Rout, T.M., Moore, J.L., Possingham, H.P. and McCarthy, M.A. (2011).
‘Allocating biosecurity resources between preventing, detecting, and
eradicating island invasions’. Ecological Economics 71: 54–62.
Ruiz, G.M. and Carlton, J.T. (2003). Invasive Species: Vectors and
Management Strategies. Washington, DC: Island Press.
477
E.S. Kennedy and K.G. Broome
Kennedy, E.S. and K.G. Broome. How do we prevent the obstacles to good island biosecurity from limiting our eradication ambitions?
How do we prevent the obstacles to good island biosecurity from
limiting our eradication ambitions?
E.S. Kennedy1 and K.G. Broome2
Department of Conservation, Private Bag 4715, Christchurch 8140, New Zealand. <ekennedy@doc.govt.nz>.
1
Department of Conservation, P.O. Box 10-420 Wellington New Zealand.
1
Abstract Island pest eradication and biosecurity spring from the same imperative. Yet in practice biosecurity is often
subordinated to the demands of eradication in planning processes so that it is not well prepared at the outset and is poorly
sustained in the aftermath. Why is this when, by any logic, eradication should not proceed unless the means are in place
to safeguard treated islands from renewed pest invasions? We draw on New Zealand’s conservation experience to explore
key reasons why biosecurity tends to lack the muscle and stamina required to protect eradication investments and their
priceless biological pay-offs. We take biosecurity to be an unassailable good and discuss remedies to the challenges it must
overcome if shortfalls and failures are not to become brakes on eradication ambitions. We think that eradicating pests from
islands is essentially a technical business and is thus the easier part of the pest-free equation. In contrast, the problems
associated with keeping islands pest-free in perpetuity are inherently more human than technical, though the latter are
taxing enough. Certainly, biosecurity’s work is more open-ended, less glamorous and not so dramatically rewarded. A
primary concern is that the complex social dimensions of biosecurity are usually neglected. We attribute this to limited
investment in social psychology tools and our habit of relying on biologists to manage behaviour-change in island visitors
of all kinds. As a result, biosecurity effort can be distorted by an over-determined emphasis on surveillance and incursion
response—an unduly reactive strategy but more familiar technical ground than quarantine which is inherently humanoriented. Our discussion of what works and what does not for these issues does not claim to offer universal remedies.
In addition to protecting existing assets competently, our aim is to prepare biosecurity to serve tomorrow’s eradication
ambitions, particularly those on the testing new frontier of pest removal from inhabited islands.
Keywords: behaviour-change, complacency, improvement, priority, social psychology
INTRODUCTION
In recent decades, we have witnessed an expansion in
the scale and biogeographic reach of projects to eradicate
invasive organisms from islands (Howald, et al., 2007;
DIISE, 2015). This diversifying of effort has consolidated
eradication as a requisite measure in island restoration
strategies. Successes have given comforting cues to
funders and sponsors contemplating eradication as a
conservation investment (see <www.milliondollarmouse.
org.nz>). Eradication ambitions are clearly intensifying
worldwide as operational experience accumulates and
scalable advances in tools and methods overcome technical
constraints on feasibility (<www.iucn.org/theme/species/
our-work/invasive-species/honolulu-challenge-invasivealien-species/commitments-towards-achieving-honoluluchallenge>; <www.predatorfreenz.org/>).
The rise of eradication as a restoration tool argues for
a closely correlated strengthening of biosecurity functions
(Russell, et al., 2007). These two lines of work define the
mutually dependent parts of the pest-free equation. Without
effective quarantine, surveillance and invasion response
in place to minimise the likelihood of pests recolonising,
it is reckless at best, and futile at worst, to proceed with
eradication. If island stewardship lacks good biosecurity,
defence of very substantial resource investments and
biological pay-offs becomes a lottery in which pest
organisms dictate the odds.
Biosecurity is arguably more potent than eradication
as a restoration tool since most subsequent conservation
investments make sense only if the island remains pestfree. Thus, biosecurity is the cornerstone, not an adjunct,
of eradication work (see the panel below). Yet, however
its priority is framed, we see uncomfortable signals in
rates of pest invasion and reinvasion on protected islands
(Clout & Russell, 2006; DIISE, 2015; Vincent, 2017) that
biosecurity’s practices and mind-sets have not advanced
adequately to meet eradication’s expectations of them.
Recurrent or not, some invasions will have been
inevitable. Near-shore sanctuaries will always be
vulnerable. Nor can we always predict the behaviour of
animal pests or the distances over which they can travel
by their own means (Veale, 2012). But other lapses will
have had preventable causes: inadequate resourcing,
limited preparation, hesitant follow-through, blind spots in
coverage, poorly developed tools and practices, cultures of
complacency, or simply outright neglect. These problems
signify that operational planners have not given biosecurity
the priority it requires.
These shortfalls are amenable to remedy because
they arise from judgments humans make. Uncorrected,
they result in biosecurity which is more often than not
poorly sustained once eradication work concludes. Until
the limitations in current practices are addressed, poor
biosecurity will frustrate eradication ambitions.
Island restoration experience in New Zealand confirms
these conclusions. In this paper, we discuss key reasons
in the New Zealand context why biosecurity tends to lack
the muscle and stamina of its indivisible eradication twin.
These issues relate more to the social and psychological
dimensions of biosecurity than to the technical challenges
we face in upgrading our tools and methods, though they,
too, are taxing enough.
In eradication literature, appeals for effective island
biosecurity regimes focus overwhelmingly on the
mechanics of pest interception. If mentioned, awarenessraising through educational outreach is usually promoted as
the means to invoke helpful biosecurity behaviour in island
communities and other public audiences (see for example,
Boser, et al., 2014). But a growing body of empirical studies
shows that heightened awareness does not necessarily alter
behaviour (McKenzie-Mohr, 2013). The cryptic attitudinal
barriers to biosecurity uptake are rarely assessed in public
audiences (Bassett, et al., 2016) or indeed within the ranks
of conservation practitioners. These social matters have
immediate bearing on biosecurity’s effectiveness.
We conclude our appraisal of obstacles to good
biosecurity with a brief review of the measures we are
In:
478C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 478–483. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Kennedy & Broome: Obstacles to good island biosecurity
taking to address them. Though the problems are likely to
be common to most biosecurity contexts, we do not claim
to offer universal remedies. Our overall aim is to protect
today’s pest-free islands more effectively and to prepare
biosecurity to serve tomorrow’s eradication ambitions,
particularly those on the new frontier of permanently
inhabited islands. Here the social challenges are amplified
and more diverse.
THE NEW ZEALAND CONTEXT
New Zealand’s Government has overall responsibility
for funding pre- and post-border biosecurity. The Ministry
for Primary Industries (MPI) takes the lead under the
Biosecurity Act 1993 to manage pest threats to human
health, the economy and environment (<http://www.mpi.
govt.nz/law-and-policy/legal-overviews/biosecurity/>).
Under this influential Act MPI can delegate biosecurity
duties to other central and local government agencies.
In the day-to-day division of biosecurity labour, the
Department of Conservation (DOC) has customarily acted
as the sentinel and conscience for protection of biodiversity
values. Obligations to protect valued islands from invasive
pests are explicit in its own mandating legislation, the
Conservation Act 1987 (<http://legislation.govt.nz/>).
Today, DOC has biosecurity obligations to more than
400 pest-free islands. Up to 240 of these have been cleared
historically or in more recent times of pest organisms
(DIISE, 2015). Others have never been invaded. Tenure
varies from public land administered by local or central
government to fully freehold.
This trusteeship of recognised sanctuaries and others inthe-making extends from the Kermadec Islands in the subtropics to Campbell Island in the high southern latitudes.
Islands at the extremities are generally well protected by
isolation and strict controls on access. Conversely, others
closer to New Zealand’s main landmasses are within easier
reach of humans and commensal pests.
Nationwide, about 85 of DOC’s 2000 staff contribute
in some way to island biosecurity. Typically, the majority
operate part-time as gate-keepers screening traffic to and
between pest-free islands. A small but growing number are
also handlers of pest-detection dogs—graduates of DOC’s
rigorous certification programme (DOC, 2015). Rangers
residing full-time on New Zealand’s signature pest-free
islands are obligate biosecurity gate-keepers.
Within DOC, island biosecurity operations are
supported in three ways. Two national advisors lead a welldefined improvement programme discussed shortly in this
paper. They negotiate for social and technical research too.
Specialist community rangers organise public outreach
throughout the country. And eradication veterans on DOC’s
Island Eradication Advisory Group (IEAG; Broome, et
al., 2011) act as conscience, critics and confidants for
biosecurity affairs.
Codifying of New Zealand’s island biosecurity standards
and practices first commenced in earnest in the late 1990s,
when the rising number of pest-free islands under watch
(Russell & Broome, 2016) required a more systematic
approach to the work. In addition to validating biosecurity
as a specialist function in its own right, the Standard
Operating Procedures and Manual of Best Practice (DOC,
2008) of this time strove for national consistency and a
persuasive culture of vigilance. Beforehand, biosecurity
had been left in the hands of collegially isolated, largely
untutored conservation practitioners.
over invasion risks in many parts of DOC’s jurisdiction
(Broome, 2013). Biosecurity arrangements lacked
coherence and firm, visible leadership.
Recommendations for remedy drew on examples
of good practice still in place and were formalised into
a determined programme of improvement (Broome &
Kennedy, 2014). This acquired national priority through
high-level sign-off in DOC.
The programme is still in train today. It aims to normalise
a vital culture of vigilance in every part of DOC’s structure.
It seeks, for island biosecurity, the unqualified functional
priority given to fire-fighting and Health & Safety, along
with comparable prerogatives and resourcing.
In addition to confronting unfamiliar new pest threats
such as the arrival of myrtle rust on public conservation lands
(<www.doc.govt.nz/our-work/biosecurity/myrtle-rust/>),
the upgrade programme is adjusting to new complexities
in an evolving social environment. First, contemporary
trends in the socialising of island management compound
invasion risks by partitioning control, subordinating the
biological significance of sanctuaries to other values or
by condoning independent rights of access. Second, comanagement agreements between DOC and owners,
other regulatory agencies or communities are increasingly
common. Third, many pest-free islands are passing into iwi
(Maori) ownership through Treaty of Waitangi redress for
colonial seizures of land. And it is business as usual for
DOC itself to promote the rare biota on near-shore islands
as a reason for the public to visit or camp as they wish (see
for example, <www.doc.govt.nz/parks-and-recreation/
places-to-go/auckland/hauraki-gulf-marine-park/visitingislands-and-marine-reserves/>).
LINES OF IMPROVEMENT
The action plan addresses cultural, capacity and
technical issues (Broome & Kennedy, 2014). Its original
lines of improvement have been directed internally to
ensuring that pest animals, weeds and —in special cases—
pathogens did not reach islands through DOC’s very
frequent traffic to them.
National advisors
The two national advisors appointed to guide all aspects
of the upgrade programme are extraordinary roles in the
DOC structure. These sole-purpose appointments signal
serious intent to make progress. The advisors are authorised
to think beyond the action plan to explore emerging needs
and new lines of improvement. Their operating mandate
extends across all functional divisions in DOC.
Practitioner networks
The advisors have created three regional networks
through which biosecurity practitioners can interact more
readily with their own kind. The networks aim for peermediated migration of knowledge and standards across
internal boundaries. Staff exchanges strengthen trust and
linkages between all three networks. External associates
in local government and NGOs frequently attend annual
workshops.
A declared imperative is to build a strong biosecurity
collegiate nationwide. Invoking the powerful unifying
benefits of collegial interaction promotes horizontal
accountability to peers (rather than vertically to managers)
and is intended to lift productivity under conditions of
capacity shortage at the workface.
ISSUES AT HOME
Pest-detection dogs
In 2012, DOC reacted to a disquieting rise in the
number and costs of pest invasions on protected islands
by launching a penetrating review of DOC’s biosecurity
fitness (DOC, 2012). The report into practices, attitudes and
capacity testified to a contagious culture of complacency
DOC is augmenting its pest-detection dog programme
through a formative partnership with New Zealand’s
Kiwibank (<www.doc.govt.nz/about-us/our-partners/ournational-partners/kiwibank/>). Gaps in capacity revealed
through internal review (Vincent, 2015) are to be filled
479
Island invasives: scaling up to meet the challenge. Ch 3A Strategy: Biosecurity
Truisms for the eradication–biosecurity partnership
Biosecurity is eradication’s
cornerstone
Eradication investments will
come to nothing without confident
biosecurity already in place
Ask two feasibility questions
before proceeding
Can the pest be eradicated?
Can the island be defended from
reinvasions?
Poor preparation has a long
legacy
If biosecurity is not prepared
well at the outset, it will likely be
sustained poorly in the aftermath
Anticipate long lead times
Biosecurity’s social and technical
complexities require longer
preparatory timeframes
Biosecurity requires people of Eradication experts may not have
the Right Stuff
the skills to manage biosecurity
on its very different horizons and
time-scales
Quarantine is the best
investment. Period.
Quarantine puts biosecurity on the
front foot, where it is strategically
and tactically most potent
with new handlers and dogs trained to detect more types
of organisms. Dog work will now take priority for handlers
with mixed conservation duties. New full-time handlers
will rove nationally to points of need. The programme has
recently acquired its own national advisor and manager,
appointments intended to empower the work through a
more autonomous occupational structure.
Best-practice prescriptions
One national advisor acts as the go-to keeper of bestpractice knowledge. S/he will codify all aspects of New
Zealand’s island biosecurity work through updated
operational prescriptions and standards. We intend to share
these prescriptions with the global community. Codes of
practice will be living guides amended as new knowledge
is acquired from experts, research or field experience.
Learning from informal experimenting in the field is
a further source of new knowledge, one traditionally
overlooked.
Pragmatic biosecurity plans
Outdated island biosecurity plans in all DOC regions
are to be revised. We are piloting a simplifying shortplan format (Kennedy, 2016a; 2017) designed to hasten
updates and approvals. The new format restricts plans
to defining briefly what has to be done (activities, rules,
standards and roles), without long explanatory narratives.
Thus, plans will act strictly as operational blueprints, not
as textbooks or technical manuals. ‘How-to' prescriptions
prepared independently of plans will instruct rangers in the
technical details of biosecurity tasks. The revised plan for
Maud Island (Caldwell & Higgott, 2017) exemplifies the
new approach.
CIMS-based incursion responses
We have recently adopted the Co-ordinated Incident
Management System (CIMS) as the standard mode of
response to pest invasions (Corson, 2018a, b). CIMS
clarifies response roles and organises support for ground
operations by co-ordinating inputs from relevant experts.
We assemble technical advisory groups (TAGs) as needed
to guide CIMS decision-makers on appropriate measures.
Biosecurity novices are apprenticed to CIMS teams for
training on the job.
480
Upgrade of quarantine facilities
We are ranking DOC’s 38 quarantine facilities for
complete renewal or internal refits in coming years. These
secure stores at mainland points of departure and island
landings are the primary pivot-points for movements of
DOC people and freight. Facilities on the mainland also
function as biosecurity’s public face. Design principles
(Kennedy, 2016b) are awaiting translation into new
architectural and construction standards as part of the bestpractice review.
Peer-review of biosecurity practices
Systematic audits of local biosecurity practices and
culture have resumed (see for example Kennedy &
Chappell, 2013; Kennedy & Trainor, 2016; Broome &
Corson, 2017). Experienced practitioners lead the reviews,
usually assisted by a novice. Audits are our most decisive
means of detecting lapses in standards and propagating
successful practices. As peer-reviews, they strengthen
mutual trust in DOC’s biosecurity ranks. Reviewers,
themselves, are instructed by the process of critique and
counselling. Currently, audit recommendations are not yet
binding.
As expected, these and other lines of operational upgrade
have had to compete for resources and priority within a
complex organisation attending to demands on multiple
conservation fronts. Obstacles to progress associated with
biosecurity’s social dimensions were under-estimated.
UNEXPECTED OBSTACLES TO GOOD
BIOSECURITY
Managing human behaviour is unavoidable in
the business of keeping islands pest-free. In contrast,
eradication is more a technical discipline and therefore the
easier part of the pest-free equation.
DOC’s internal biosecurity arrangements cannot be
intensified successfully, nor can they be sustained in
perpetuity, if biosecurity behaviour is not normalised in
organisational thought and practice. The same applies
to public traffic to islands. By necessity then, we have
extended our upgrade programme to invoke beneficial
behaviours in public audiences whose multiple forms
of contact with pest-free islands are less within DOC’s
control.
Kennedy & Broome: Obstacles to good island biosecurity
Of the two broad classes of audience (professional
and public), our colleagues have proven to be the more
resistant to the right behaviours. This is perplexing, since
they are closely invested in conservation and should be
acutely aware of invasive pest threats to insular biota. Yet
compliance with expected behaviours still appears to be
more obligatory than voluntary.
networks are a resolute first step towards this goal. Creative
celebrating of successes and champions will help. Loss of
occupational autonomy under DOC’s line-management
arrangements is a barrier to progress (Kennedy, 2003).
Expert mind-set
Island biosecurity in New Zealand is faltering most
conspicuously in its managing of public access to pestfree islands. DOC is still seriously under-invested here,
most obviously because the tools of social psychology
are not yet used adequately by DOC itself or trusted by
biologically-minded practitioners.
We observe an ‘expert’ mind-set at play in our
colleagues. It reasons that as conservation specialists they
are best placed to judge risk and thus how much they need
to comply with rules. Treating compliance as discretionary
translates biosecurity from an essential good into a nuisance
(rationalised perhaps as an intrusion on professional
judgement). Ough Dealy (2016) reports a similar 'expertsknow-best' phenomenon in dog owners landing their pets
illegally on pest-free islands in New Zealand’s far north.
Biosecurity’s logic is not sufficient on its own to alter this
mind-set.
Colleagues resist further through arguments that there
is little point in DOC staff submitting to quarantine checks
when the public can visit the same islands without doing so.
This surprisingly prevalent thinking underwrites pressure
to reduce quarantine standards (authors, pers. obs.). Its
disabling logic supposes that biosecurity can extinguish
all risk. We argue instead that it can only minimise the
probability of pest incursions. So, if some of an island’s
visitors are guaranteed pest-free, the risk is reduced
accordingly.
Image problems
Biosecurity is inherently prone to neglect. It has a poor
image. In contrast with the heroic character of eradication
operations, biosecurity work is less glamorous, inherently
open-ended and not so dramatically rewarded. Reputations
are simply not made in the business. Arguably then,
biosecurity does not appeal to practitioners motivated
vocationally to make a demonstrable difference for
conservation (Kennedy, 2003).
Perversely, biosecurity is a casualty of its own success.
When everything works as it should, nothing happens.
This is particularly penalising in today’s over-determined
goal-directed working environments. For resource-stressed
colleagues in DOC, intercepting very few invasive pests,
year after year, argues for shifting effort from biosecurity
to more obviously productive work.
We find as a result that our field and management staff
are inclined to treat biosecurity as an insurance policy on
which it is safe to avoid paying the premiums. Too often,
they subordinate biosecurity to lower but more immediate
priorities. Even financially punishing biosecurity failures
(for example, the > $NZ200,000 mouse invasion of Maud
Island; Broome, et al., this issue) have had only a limited
chastening effect on this habit.
OVERCOMING ATTITUDINAL OBSTACLES IN
DOC
We are in a stronger position to legislate compliance
with this captive audience. Traffic to islands is governed by
behavioural rules applying to all DOC staff, our associates
and our freight. Quarantine inspections are obligatory (see
for example, Hiscock, 2016) and must be allowed for in
operational timetables. At points of departure, authorised
biosecurity gate-keepers are mandated (regardless of
their occupational rank) to prohibit travel until quarantine
standards are met.
Ultimately, normalising biosecurity will be achieved in
this audience by playing on the powerful human instinct
to conform to peer-pressure. This influence on attitudes
and behaviour is a defining quality of collegiality, itself
articulated through peer-mediated understanding of
common purpose, values and identity. DOC’s biosecurity
OBSTACLES TO BEHAVIOUR-CHANGE IN
PUBLIC AUDIENCES
Problematic messaging
In piecemeal approaches or hesitant use of creative
messaging, outreach strategies reflect their uncertainty on
how to bring behaviour-change about. Neglecting outreach
as a quarantine measure because of its social uncertainties
denies biosecurity a powerful range of interventions.
A distorting emphasis on surveillance and response is
likely to result, particularly as these activities are more
comfortable technical ground for practitioners unskilled in
modifying public behaviour.
For instance, three years after cats and rodents were
eradicated from Great Mercury Island (1,872 ha) in New
Zealand’s southern Hauraki Gulf, on-island surveillance
and response regimes are in place (Collins & Corson, 2016)
but quarantine lacks any coherent messaging strategy to
manage visitor risk. As a result, biosecurity to manage
the differing forms of public contact with the island is
dangerously off-balance and inherently reactive.
In the absence of insights from social research, the
customary response is to fall back on orthodox messaging
through signs and pamphlets. Typically, this relies on a
mix of appeals to protect natural values and prohibitions
on unwanted behaviour, all conveyed in alienating official
language. We are not confident that this messaging or its
media are effective. Much of it amounts to shouting at
audiences.
Branding is similarly problematic. In 2009, Auckland
Council and DOC launched the Treasure Islands brand
(<www.treasureislands.co.nz>) to engender biosecurity
behaviour in the hundreds of thousands of public and
commercial travellers to 44 pest-free islands in the Hauraki
Gulf Marine Park. Uncertain of its effectiveness, brand
design was under constant review (Jack, 2011). Variations
were visible everywhere, from bill-boards to bait stations,
pamphlets to piers. Later surveys of biosecurity awareness
in island residents and visitors showed limited knowledge
of the brand as it had been communicated (Auckland
Council, 2010a, 2010b; Lysnar, 2016). Fraser, et al. (2016)
concluded that face-to-face conversations with ferry
passengers were a more effective means of outreach.
We consider that the finer-scale insights of qualitative
social research studies are more likely than quantitative
to determine which messages and media will prompt
target audiences to become willing biosecurity actors—
to quarantine their personal gear before departure. Social
research shows already that those messages will likely
resonate with audiences’ own reasons for visiting islands,
not with our notions of nobility in conservation values or
biosecurity need (see McKenzie-Mohr, 2013).
Biosecurity’s unwitting social offences
Invoking the quarantine habit in public audiences faces
cryptic psychological barriers. We regard credibility as
a crucial issue. As with our sceptical colleagues, visitors
arriving on controlled pest-free islands are disinclined to
believe that they have a mouse or a rat in their bag when
asked to check (Tyrrell, 2012). This is a pivotal moment
at which biosecurity’s legitimacy is questioned along with
the gate-keeper’s sanity.
481
Island invasives: scaling up to meet the challenge. Ch 3A Strategy: Biosecurity
Winning the contest for credibility is made more
difficult in that emptying bags and pockets in the company
of strangers can be socially awkward, particularly in the
congested conditions of an enclosed quarantine room.
At worst it may be regarded as an intrusion on personal
privacy.
Remedies for this and many other unwitting social
offences seem intractable to practitioners who are
suspicious of the arcane social sciences. This is yet another
barrier to overcome. In fact, remedies will be more likely
in the hands of social psychologists; they can no longer
be left to chance or to force of personality in biosecurity
gate-keepers.
ADDRESSING THE CHALLENGES
DOC’s investments to strengthen island biosecurity
have had to extend beyond the internal focus of its original
action plan. Thinking and initiatives are advancing along
the following lines.
Biosecurity’s functional identity and appeal
We see potential to lift biosecurity’s functional
appeal and effectiveness in-house by giving the work its
own occupational structure, dedicated line-manager and
credentials. This thinking recognises that biosecurity
is a specialist field in its own right, entitled to greater
operational autonomy, its own leadership prerogatives and
requiring operators uniquely suited to its distinctive modes
of work. Not all field ecologists have the necessary manner
or motivations. Discrete occupational identity breaks with
today’s unhelpful assumption that biosecurity is merely an
adjunct function of eradication and other specialist work.
Audits of readiness
Biosecurity readiness for eradication operations and
ongoing quarantine would benefit from more determined
auditing. An expert group resembling the IEAG appeals as
a means for authoritative scrutiny. The two should operate
in parallel, even if they share experts from time to time.
Testing our own attitudinal barriers to good practice
We see immediate value in social research to
investigate the attitudinal barriers to biosecurity uptake
in our own colleagues. Likewise, we are learning to test
the assumptions we make routinely about who our public
audiences are, what they understand and what motivates
them. Inexpert assumptions here make unreliable steppingoff points for changing public behaviour.
Seeing the improbable
We are adopting the principle that our colleagues and
public alike must see for themselves before they will accept
the improbable. This means replacing talking and preaching
with evidential photos, videos, eye-witnesses and stories.
In the quarantine inspection room on Matiu/Somes Island
in Wellington Harbour, for instance, incoming visitors can
see and handle a sobering collection of pest organisms
intercepted in bags and pockets.
Measuring the costs of failure
The time and dollar costs of incursion responses are now
recorded more carefully using new reporting templates to
prompt for essential information (Kennedy, 2015a). These
data are collated to argue for better resourcing and priority.
Each incursion is treated as an experiment from which
lessons are now extracted more systematically (Kennedy,
2015b; Kennedy, 2016c). All reports are added to DOC’s
comprehensive database of island incursions (Vincent,
2017).
482
Making biosecurity sexy
DOC’s pest-detection dogs are now employed more
purposefully as our ambassadors and champions. Their
unique ability to convey biosecurity messages by their very
presence surpasses the best of human efforts. Contact time
with public audiences of all types is now built into work
schedules.
Social research
We have commenced two qualitative social research
projects designed to find ways of stimulating beneficial
behaviour in island visitors. Both use community-based
social marketing methods (CBSM; McKenzie-Mohr,
2013) and are intended to generate solutions applicable
throughout New Zealand.
The first project is searching for effective ways to
convince recreational boat users to leave their pet dogs at
home. Dogs landed with families picnicking on pest-free
open-access islands in New Zealand’s far north are killing
endangered kiwi chicks crèched there for conservation
purposes (Ough Dealy, 2016; Ough Dealy & Greig,
2017). Findings from this modest inaugural study have
confounded most of our assumptions about the target
audience and its motivations to take dogs ashore illegally
(further information is available from the authors).
The second study aims to persuade recreational boat
users to check their vessels and gear for pests before
they leave home. This study will test our hypothesis that
willingness to check at home will be greater than at the
launch ramp where time and other pressures intrude.
Consistent with CBSM methods, this study has attempted
to isolate specific behaviours for change while exploring
the character of the target audience (Harbrow, 2017).
Conversing with audiences
Though still only a concept, we aim to converse with
rather than talk at public audiences. Experiments with
more relaxed, colloquial language on conventional signage
have been contemplated but not yet launched. Likewise,
we are considering the head-turning potential in nonconventional signage (say, 3-D models of rodents) as a
means of distinguishing our biosecurity messages from the
walls of rectangular 2-D bill-boards island visitors expect
to see at departure points.
We have also adopted the principle of conversing
continuously with audiences well in advance of their
arrival at departure points. Visitors of all kinds travelling
to controlled-access islands receive information ahead of
time on how to prepare for their quarantine inspection.
YouTube videos (<https://youtu.be/yXNOpfW7PPQ>)
are used too. Bounce-back messages on booking sites
are under consideration. Open-access islands present
greater difficulties. Here, a progression of messages along
approach routes, conversing by various media, would
avoid the unsatisfactory situation of today where visitors
are confronted with conditions on travel at the last moment,
when they are least able or prepared to comply.
Reaching back into supply chains
Following the lead of Morgan, et al. (2014), we regard
it as imperative to check freight for pests early in the supply
chain. We award pest-free warrants (<www.doc.govt.
nz/parks-and-recreation/places-to-go/auckland/haurakigulf-marine-park/know-before-you-go/treasure-islands/
pest-free-warrant/>) to businesses guaranteeing to supply
pest-free goods or services. These businesses are used and
publicised as preferred suppliers. Not only is quarantine
at source more efficient but it safeguards biosecurity gatekeepers from pressures to pass suspect goods or freight for
shipping under operational urgency at departure points.
Kennedy & Broome: Obstacles to good island biosecurity
CONCLUSION
Advances of these kinds are intended to make good on
biosecurity’s duties to its eradication partner. They spring
from lessons learnt during determined upgrading of DOC’s
island biosecurity performance. We recognise now that
biosecurity must be the cornerstone of island restoration
work, not a secondary function of pest eradication or
other conservation investments. Its effectiveness is
not assured, however, unless it is respected as a highly
specialised discipline, resourced accordingly and trusted
to credentialed practitioners attuned to the particular
demands of quarantine, surveillance and response. We
have become acutely conscious of the need to address the
discipline’s complex social dimensions. These have been
neglected at home historically because biosecurity work
was usually left to field ecologists unfamiliar with social
research tools. Biosecurity cannot avoid managing human
behaviour if it is to acquire its necessary reach and balance.
More systematic programmes of social investment are
needed to address the obstacles to good biosecurity today.
They will be outstanding preparation for the future in
which the formidable social challenges of eradicating pests
from populated islands, or indeed from entire countries
(e.g. Predator Free New Zealand 2050) will demand more
sophisticated tools and thinking than we possess now.
REFERENCES
Auckland Council. (2010a). Report on a Survey of Passengers on Sealink’s
Great Barrier Island Service, Summer 2009–2010. Unpublished internal
report. Auckland, NZ.
Auckland Council. (2010b). Report on a Survey of Boat-users in the
Hauraki Gulf, 2009–2010. Unpublished internal report. Auckland, NZ.
Bassett, I.E., Cook, J., Buchanan, F. and Russell, J.C. (2016). ‘Treasure
Islands: biosecurity in the Hauraki Gulf Marine Park’. New Zealand
Journal of Ecology 40 (2): 250–266.
Boser, C.L., Corey, C., Faulkner, K.R., Randall, J.M., Knapp, J.J. and
Morrison, S.A. (2014). ‘Strategies for biosecurity on a nearshore island
in California’. Monographs of the Western North American Naturalist
7 (1): 412–420.
Broome, K.G., Cromarty, P., Cox, A., Griffiths, R., McClelland, P.
and Golding, C. (2011). ‘The Island Eradication Advisory Group
(IEAG)—A model for effective technical support for eradication
project planning and management’. In: C. R. Veitch, M. N. Clout and
D. R. Towns (eds.) Island invasives: eradication and management, p.
524. Occassional Paper SSC no. 42. Gland, Switzerland: IUCN and
Auckland, New Zealand: CBB.
Broome, K.G. (2013). Island biosecurity: Improvements to DOC’s
Practice and Culture. Unpublished project report, DOCDM-1118748.
Hamilton, NZ: Department of Conservation.
Broome, K.G. and Kennedy, E.S. (2014). Action Plan for Island
Biosecurity. Unpublished internal plan, DOCDM-1372876.
Christchurch, NZ: Department of Conservation.
Broome, K.G. and Corson, P. (2017). Assessment of Biosecurity for
Whenua Hou Nature Reserve. Unpublished internal audit report,
DOCCM-2968470. Hamilton, NZ: Department of Conservation.
Broome, K.G., Brown, D., Brown, K., Murphy, E., Birmingham, C.,
Golding, C., Corson, P., Cox, A. and Griffiths, R. (This issue). ‘House
Mice on Islands: Management and Lessons from New Zealand’.
Caldwell, S. and Higgott, F. (2017). Te Pākeka Maud Island Biosecurity
Plan: Quarantine, Surveillance and Incursion. Unpublished operational
plan, DOCCM-2834023. Picton, NZ: Department of Conservation.
Clout, M.N. and Russell, J.C. (2006). ‘The Eradication of Mammals from
New Zealand Islands’. In: F. Koike, M.N. Clout, M. Kawamichi, M. de
Poorter and K. Iwatsuki (eds.) Assessment and Control of Biological
Invasion Risks, pp.127–141. Gland, Switzerland: IUCN.
Collins, L. and Corson, P. (2016). Ahuahu Great Mercury Island
Biosecurity Plan. Unpublished operational plan, DOCDM-1374415.
Whitianga, NZ: Department of Conservation.
Corson, P. (2018a). Island Biosecurity Incursion Responses: Flow
Diagrams by Incursion Phase. Unpublished internal guideline,
DOCCM-5387656. Rotorua, NZ: Department of Conservation.
Corson, P. (2018b). Roles and Responsibilities for Island Biosecurity
Pest Incursions. Unpublished internal guideline, DOCCM-5388157.
Rotorua, NZ: Department of Conservation.
DOC. (2008). Island Biosecurity Best Practice Manual (v.2.2). Internal
publication, DOCDM-20171. Wellington, NZ.
DOC. (2012). Island Biosecurity: Improvements to DOC’s Culture and
Practice. Unpublished project brief, DOCDM-1003645. Hamilton. NZ:
Department of Conservation.
DOC. (2015). Standard Operating Procedure for the Conservation
Dog / Handler Team. Unpublished internal SOP, DOCCM-749423.
Wellington, NZ.
DIISE. (2015). The Database of Island Invasive Species Eradications,
developed by Island Conservation, Coastal Conservation Action
Laboratory UCSC, IUCN SSC Invasive Species Specialist Group,
University of Auckland and Landcare Research New Zealand. <http://
diise.islandconservation.org>.
Fraser, D., Dabb, H. and Graham, C. (2016). Biosecurity Awareness
of Ferry Passengers Travelling to Islands in the Hauraki Gulf, New
Zealand. Auckland, NZ: Environmental & Animal Sciences, Unitec.
<http://hdl.handle.net/10652/3866>.
Harbrow, M. (2017). Auckland Boat-show Survey, 23 April–22 May 2017.
Unpublished summary of data (PDF). Wellington, NZ: Department of
Conservation.
Hiscock, J. (2016). Subantarctic Expeditions: Operational Procedures
for Management and Research Expeditions in the New Zealand
Subantarctic. Unpublished rules and guidelines, DOCDM-1351346.
Invercargill, NZ: Department of Conservation.
Howald, G., Donlan, C.J., Galván, J.P., Russell, J.C., Parkes, J.,
Samaniego, A., Wang, Y., Veitch, C.R., Genovesi, P., Pascal, P.,
Saunders, A. and Tershy, B. (2007). ‘Invasive rodent eradication on
islands’. Conservation Biology 21: 1258–1268.
Jack, S. (2011). Treasure Islands – an Overview of the Biosecurity
Campaign for the Hauraki Gulf. Unpublished internal briefing paper,
DOCDM-725070. Auckland, NZ: Department of Conservation.
Kennedy, E.S. (2003). ‘Managerialism as a Professionalising Catalyst for
the Front-line Practitioner Community of New Zealand’s Department
of Conservation’. M.Appl.Sc Thesis. Lincoln, NZ: Lincoln University.
<http://researcharchive.lincoln.ac.nz/handle/10182/1031>.
Kennedy, E.S. and Chappell, R. (2013). Preliminary Report: Audit of
DOC’s Quarantine and Surveillance Arrangements for the Islands of
the Marlborough Sounds. Unpublished audit report, DOCCM-1318761.
Christchurch, NZ: Department of Conservation.
Kennedy, E.S. (2015a). Summary of Pest Incursion Costs on Islands.
Unpublished report template, DOCDM-1560956. Christchurch, NZ:
Department of Conservation.
Kennedy, E.S. (2015b). Debrief Report for Island Incursion. Unpublished
report template, DOCDM-2619461. Christchurch, NZ: Department of
Conservation.
Kennedy, E.S. (2016a). How to Write a Genuinely Useful Biosecurity
Plan in a Jiffy. Unpublished presentation (PPT), DOCCM-2844257.
Christchurch, NZ: Department of Conservation.
Kennedy, E.S. (2016b). Quarantine Store Design. Unpublished summary
of design and structural elements, DOCCM-3186359. Christchurch,
NZ: Department of Conservation.
Kennedy, E.S. (2016c). Norway Rat Incursion, Ulva Island, September–
December 2015. Unpublished internal report using the debrief template,
DOCDM-2808849. Christchurch, NZ: Department of Conservation.
Kennedy, E. S. and Trainor, S. (2016). Audit of DOC’s Biosecurity Store
at Halfmoon Bay, Rakiura Stewart Island. Unpublished audit report,
DOCCM-2874997. Christchurch, NZ: Department of Conservation.
Kennedy, E.S. (2017). Proposed Three-part Structure for Biosecurity
Plans. Unpublished internal pamphlet, DOCCM-3186052. Christchurch,
NZ: Department of Conservation.
Lysnar, P. (2016). Summary of Findings from Hauraki Gulf Treasure
Islands Campaign Survey. Unpublished contract report. Auckland, NZ:
Environmental Services, Auckland Council.
McKenzie-Mohr, D. (2013). Fostering Sustainable Behaviour—An
Introduction to Community-Based Social Marketing. Canada: New
Society Publishers.
Morgan, S., Marshall, A., Cook, J., Waipara, N., Ward, D., McNatty, A.,
Adams, N., and Fraser, D. (2014). ‘An Assessment of Risk Pathways
from Commercial Businesses to Pest-free Islands of the Hauraki
Gulf’. Presentation to the New Zealand Biosecurity Institute (NETS)
conference, 29 July–1 August 2014, New Plymouth, New Zealand.
<http://hdl.handle.net/10652/3097>.
Ough Dealy, H. (2016). Dogs on Pest-free Island Beaches in the Bay of
Islands, Northland. Unpublished summary of interim research findings,
DOCCM-2845381. Bay of Islands, NZ: Department of Conservation.
Ough Dealy, H. and Greig, E. (2017). Visitor and Dog Monitoring on
Urupukapuka Island, 2014–2017. Unpublished internal research report,
DOCCM-3010021. Bay of Islands, NZ: Department of Conservation.
Russell, J.C. and Broome, K.G. (2016). ‘Fifty years of rodent eradications
in New Zealand: another decade of advances.’ New Zealand Journal of
Ecology 40(2): 197–204.
Russell, J.C., Clout, M.N. and Towns, D.R. (2007). ‘Early Detection
and Eradication of Invading Pests’. In: G.W. Witmer, W.C. Pitt and
K.A. Fagerstone (Eds.) Managing Vertebrate Invasive Species 43:
Proceedings of an International Symposium, pp. 268–272. Fort Collins,
Colorado, USA: USDA/APHIS Wildlife Services, National Wildlife
Research Center. <http://digitalcommons.unl.edu/nwrcinvasive/43>.
Tyrrell, L. (2012). 2012 Summer Survey Anecdotal Report (March 2012).
Unpublished survey report to Auckland Council. Unitec, Auckland.
Veale, A.J. (2012). ‘Observations of stoats (Mustela erminea) swimming’.
New Zealand Journal of Zoology 40: 166–169.
Vincent, K. (2015). Pest Detection Dogs Programme Review. Unpublished
internal report, DOCCM-1534015. Wellington, NZ: Department of
Conservation.
Vincent, K. (2017). Database of Island Invasion Incidents. Unpublished
internal database, DOCDM-53231. Wellington, NZ: Department of
Conservation.
483
M. Latofski-Robles, F. Méndez-Sánchez, A. Aguirre-Muñoz, C. Jáuregui-García, P. Koleff-Osorio, A.I.
González-Martínez, G. Born-Schmidt, J. Bernal-Stoopen and E. Rendón-Hernández
Latofski-Robles, M.; F. Méndez-Sánchez, A. Aguirre-Muñoz, C. Jáuregui-García, P. Koleff-Osorio, A.I. González-Martínez, G. Born-Schmidt,
J. Bernal-Stoopen and E. Rendón-Hernández. Mexico’s island biosecurity programme: collaborative formulation and implementation
Mexico’s island biosecurity programme: collaborative formulation
and implementation
M. Latofski-Robles1, F. Méndez-Sánchez1, A. Aguirre-Muñoz1, C. Jáuregui-García1, P. Koleff-Osorio2,
A.I. González-Martínez2, G. Born-Schmidt2, J. Bernal-Stoopen3 and E. Rendón-Hernández3
Grupo de Ecología y Conservación de Islas, A.C. Moctezuma 836 Zona Centro. Ensenada, Baja California, México.
<mariam.latofski@islas.org.mx> 2Comisión Nacional para el Conocimiento y Uso de la Biodiversidad, Liga PeriféricoInsurgentes Sur 4903, Parques del Pedregal, Ciudad de México. 3Comisión Nacional de Áreas Naturales Protegidas,
Ejército Nacional 223, Anahuac I sección, Ciudad de México.
1
Abstract Mexico’s National Strategy on Invasive Species (2010) and the National Strategy for the Conservation and
Sustainable Development of Islands (2012) embrace its steady and positive 20-year trajectory on island restoration.
Mexico has come halfway to having all islands free of invasive mammals. To sustain these results in the long-term,
biosecurity became a priority. To implement these national strategies, the National Commission for Knowledge and Use of
Biodiversity (CONABIO) and the National Commission for Natural Protected Areas (CONANP) integrated a participatory
programme to develop the country’s capacities for managing invasive alien species (IAS). With funding from the Global
Environment Facility (GEF) through the United Nations Development Programme (UNDP), and private donors, the IAS
programme is now under operation. Grupo de Ecología y Conservación de Islas, A.C. (GECI), a professional Mexican
NGO, is implementing the programme on islands, along four lines of action: (1) island biosecurity; (2) environmental
learning and capacity building; (3) control and eradication of IAS; and (4) monitoring to assess ecological recovery. While
promoting the importance of biosecurity amongst all social actors, the focus is on long-term formal implementation,
culture and everyday life. The methods therefore vary: workshops with authorities, integration of specific biosecurity
protocols, and art and conservation activities with the local communities, particularly with children and young adults. The
experience to date shows that enforcement by authorities and integration of the subject by local fishermen communities
and island users are two key factors in sustaining the valuable and tangible results achieved to date over the long term.
Keywords: biosecurity, conservation, early detection, eradication, prevention, rapid response
INTRODUCTION
As one of the world’s megadiverse countries, Mexico
acknowledges the importance of safeguarding its
biodiversity and over 10,000 endemic species (LlorenteBousquets & Ocegueda-Cruz, 2008). Invasive alien
species (IAS) pose the most important threat to biodiversity
worldwide (Reaser, et al., 2007; Towns, 2011), and have
caused 67% of the extinctions of Mexican vertebrates
(Aguirre-Muñoz, et al., 2011a). Consequently, a National
Advisory Committee for the Strategy on Invasive Species
(CANEI, for its Spanish acronym) was created in 2008. It
is comprised of governmental and academic institutions, as
well as non-profit civil society organisations. Coordinated
by the National Commission for the Knowledge and Use
of Biodiversity (CONABIO), the CANEI developed
the “National strategy on invasive species: prevention,
control and eradication” in 2010. Its vision is to address
the problems of IAS, by creating efficient prevention, early
detection and rapid response systems, as well as a legal
framework to mitigate, control and eradicate these species
(CANEI, 2010).
The nearly 4,000 Mexican islands, as do most of the
islands around the world, host a disproportionate amount
of the country’s biodiversity (Whittaker & FernándezPalacios, 2007). They are hotspots of endemism richness,
with 14 times more endemic species than the mainland
(Aguirre-Muñoz, et al., 2016a). In recognition of the need
to protect this biodiversity as well as the livelihoods of
island communities, the Mexican government has included
all islands in the National System of Natural Protected
Areas (Aguirre-Muñoz, et al., 2017a) with the recent
decree of the Islas del Pacífico de la Peninsula de Baja
California Biosphere Reserve (DOF, 2016). Therefore, the
formulation of the National Strategy for the Conservation
and Sustainable Development of the Mexican Island
Territory (2012) was an important step forward. This
national strategy sets priorities to work on three tactical
lines – sovereignty, conservation and sustainable
development – through four transverse lines of action –
knowledge, public policies, inter-institutional coordination
and financing (CANTIM, 2012).
ISLAND CONSERVATION IN MEXICO
The history of island conservation in Mexico delivers a
restoration success story. Through to 2017, 60 populations of
11 invasive mammal species have been eradicated from 39
islands, which represents over 59,000 ha restored (AguirreMuñoz, et al., 2018). Thanks to these efforts, at least 147
endemic taxa of mammals, reptiles, birds and plants are
protected. Furthermore, 227 highly vulnerable seabird
colonies are recovering from the impacts of IAS (AguirreMuñoz, et al., 2016b). A growing network of collaborating
federal government agencies, e.g. the National Commission
for Protected Areas (CONANP), CONABIO, the National
Institute of Ecology and Climate Change (INECC), and
the Department of the Environment and Natural Resources
(SEMARNAT), academic institutions, local communities,
fishing cooperatives, civil society organisations and
donors (national and international) has been fundamental
to achieving success. Working in close collaboration with
the multiple partners, Grupo de Ecología y Conservación
de Islas, A.C. (GECI) has implemented all but two of the
island eradications in Mexico and is currently executing
other eradication projects on several islands. GECI is a
Mexican civil society organisation, which works with an
interdisciplinary and comprehensive approach toward the
restoration, conservation and sustainable development of
islands (Aguirre-Muñoz, et al., 2011b).
GECI’s goal, as outlined in the IUCN’s Honolulu
Challenge, is to remove invasive mammals from all islands
of Mexico by 2030 (IUCN, 2017). To achieve it, we need
to eradicate a further 70 populations of invasive mammals
from 34 islands. To do so, we aim to eradicate invasive
mammals following restoration priorities, including where
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
484
up to meet the challenge, pp. 484–488. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Latofski-Robles, et al.: Mexico’s island biosecurity programme
endemic species are vulnerable, eradications are feasible and
risk of reinvasion is lower (Latofski-Robles, et al., 2014).
Therefore, the implementation of a National Programme
for Island Biosecurity – the policies, measures and actions
to protect island biodiversity from IAS by preventing their
arrival and establishment (Roberts, 2003; Russell, et al.,
2008) – is vital to ensure that successes achieved remain
in the long term, and that the investment in conservation
measures, such as eradications, has the highest return rates
(Broome, 2009). Implementing biosecurity will also further
Mexico’s achieved international commitments, in line
with Aichi Biodiversity Target #9 which states: “By 2020,
IAS and pathways are identified and prioritised, priority
species are controlled or eradicated, and measures are in
place to manage pathways to prevent their introduction and
establishment” (CBD, 2010). Additionally, new restoration
projects will benefit from building biosecurity capacities
beforehand. Thus, biosecurity becomes a transverse line of
action amongst all of GECIs restoration projects (AguirreMuñoz, et al., 2016b).
FORMULATING THE NATIONAL ISLAND
BIOSECURITY PROGRAMME
Islands significantly contribute to the country’s
megadiversity. They harbour 8.3% of all vascular plant
and terrestrial vertebrates (CANTIM, 2012). They also
support the livelihood of more than 200,000 people,
most of which rely on the valuable marine resources
that thrive in adjacent waters. However, some islands
have faced the negative impacts of IAS, particularly
mammalian predators, for centuries. The introduction of
such problematic species to islands in Mexico has been
mainly due to anthropogenic reasons, either intentionally
or accidentally. Before the 20th century, introduction of IAS
was mainly related to the harvesting of marine mammals
and guano mining. Nowadays, the sources of introductions
have diversified and include commercial and sport fishing,
as well as tourism related activities (Aguirre-Muñoz, et
al., 2011b). At first, restoration projects were all about
solving the problem already at hand, eliminating the IAS;
however, as we free islands of their IAS, we must change
our way of thinking and become proactive in preventing
reintroductions or new introductions. In order to halt the
introduction of IAS, intentional or accidental, we need
a society that is aware of the root causes and problems
associated with the loss of biodiversity and the ecosystem
services it provides. We need the social construction of a
new paradigm, of everyone feeling a sense of privilege
every time we visit an island and acknowledging that the
conservation of such a special place is in our own hands.
Therefore, GECI’s restoration projects are accompanied
by an environmental learning and outreach campaign that is
designed for that specific island and its local community’s
characteristics. We seek to boost the local community
identity, by publicising the island’s biodiversity, as well
as its endemic or more charismatic species. We produce
and distribute different outreach materials (e.g. posters,
photographic catalogues, wristbands, colouring books,
puzzles, etc.) that showcase the island’s uniqueness and
what you can do to protect it. We also give varied talks
to different sectors, such as schools, universities, fishing
cooperatives and tourist operators, about the restoration
project and the outcomes expected. Moreover, we learn
about the way local communities understand, interact with
and feel about their environment through their artistic
expressions. We provide the opportunity for youngsters
to express their connection to nature through music,
painting, drawing and story-telling workshops, and have
documented beautiful results.
GECI’s efforts to make island biosecurity a subject
matter and common topic amongst island users and
managers became systematic with the nationwide project
to implement the Strategy on Invasive Species in Mexico.
With funding from the Global Environment Facility (GEF)
in coordination with the United Nations Development
Programme (UNDP), the CONABIO and the CONANP
lead the inter-sectorial project to implement this Strategy.
Implementing biosecurity protocols and building capacities
on managing IAS are two priority actions established in the
Invasive Species Strategy (CANEI, 2010). The project is
implemented in priority areas of conservation and focuses
on preventing the arrival and establishment of IAS through
prevention measures, early detection systems and rapid
response (Born-Schmidt, et al., 2017).
The project began the planning stage in 2012, and
GECI, who is coordinating the island programme, started
by identifying priority protected areas for implementation
and setting action guidelines. The lines of action, with a
2015–2018 implementation horizon, are: 1) Biosecurity:
development, implementation and evaluation of
biosecurity protocols, creation of biosecurity committees;
2) Environmental learning and outreach: producing
outreach materials, developing awareness campaigns
about IAS, building capacities for local groups on early
detection and rapid response; 3) Restoration: management
of the IAS, as well as native species present; 4) Monitoring:
documenting ecosystem responses to eradication of IAS
(Aguirre-Muñoz, et al., 2013). Six priority protected areas
are our pilot project areas where the biosecurity project is
currently being implemented (Table 1, Fig. 1). The project
is being replicated in the Gulf of California, in a group of
islands known as the Midriff Islands.
DESIGNING AND IMPLEMENTING ISLAND
BIOSECURITY PROTOCOLS
In order for biosecurity to fulfil its purpose, we
need to analyse and take into account all the particular
activities that different sectors carry out on the island.
Consequently, we decided on a “bottom-up” strategy to
create site-specific biosecurity protocols in an adaptive
and participatory manner (Aguirre-Muñoz, et al., 2013).
With every sector involved in the protocol design from the
beginning, they provide the information needed to make
an informed risk analysis and detect critical control points
(González-Martínez, et al., 2017). Furthermore, by being
involved, the communities are more likely to approve and
adopt prevention measures that need to be carried out in
everyday life and with a long-term vision.
Biosecurity protocols are documents where all the
components of biosecurity are detailed; so that each
stakeholder understands what will be implemented,
and how he/she is involved. The main components of
biosecurity are prevention, early detection and incursion
response (Russell, et al., 2008). The key behind prevention
Fig. 1 Map of the islands and their coastal areas of influence
for the Biosecurity Programme.
485
Island invasives: scaling up to meet the challenge. Ch 3A Strategy: Biosecurity
Table 1 Biosecurity pilot project areas.
Island
Location
Isla Guadalupe
Biosphere Reserve
Pacific Ocean (260
km off the coast of
the Baja California
Peninsula)
Isla Cedros –
Pacific Peninsula
of Baja California
Biosphere Reserve
Pacific Ocean (25
km off the coast of
Baja California Sur
Peninsula)
Islands: Cedros &
San Benito Oeste
Archipiélago de
Revillagigedo
National Park
Islands: Socorro &
Clarión
Previous
eradications
rabbit & donkey
(2002)
horse (2004)
goat (2006)
dog (2007)
cat (in progress)
Cedros:
dog (in progress)
San Benito Oeste:
rabbit & goat
(1998)
donkey (2005)
cactus mouse
(2013)
Pacific Ocean (480 Socorro:
km off the coast of sheep (2010)
Baja California Sur) cat (in progress)
Clarión:
sheep & pig (2002)
IAS present
Local community
Plants 47
Reptiles 0
Birds 5
Mammals 2
100 people, comprising a
fishermen’s camp, a Navy
Station and GECI´s station.
Cedros:
Plants unknown
Reptiles 0
Birds 4
Mammals 6
San Benito Oeste:
Plants 9
Reptiles 0
Birds 4
Mammals 0
Socorro:
Plants 47
Reptiles 1
Birds 5
Mammals 2
Clarión:
Plants unknown
Reptiles 1
Birds 5
Mammals 1
Plants 5
Reptiles 0
Birds 0
Mammals 1
10,000 people comprising a
fishermen’s cooperative, the
Navy Station, and the salt
exporter.
Isla Espíritu Santo
– Gulf of California
Islands Protected
Area
Gulf of California
cat (2017/absence
(25 km off the coast confirmation stage)
of Baja California
goat (in progress)
Sur)
Banco Chinchorro
Biosphere Reserve
Caribbean Sea (30
km off the coast of
Quintana Roo)
Cayo Centro:
black rat &
cat (2015)
Cayo Norte Mayor
& Menor:
black rat (2012)
Plants 6
Reptiles 1
Birds 2
Mammals 0
Gulf of Mexico
(140 km off the
coast of Yucatan)
Pérez:
black rat (2011)
Plants 5
Reptiles 0
Birds 1
Mammals 0
Islands: Cayo
Centro, Cayo Norte
Mayor & Cayo
Norte Menor.
Arrecife Alacranes
National Park
Islands: Pérez,
Pájaros, Muertos,
Desterrada & Chica.
Muertos & Pájaros:
house mouse (2011)
is to set as many obstacles as possible throughout the
pathways of introduction, to reduce the probability for IAS
to get to the islands. Early detection means a surveillance
method through detection devices, such as traps, to
determine if there is an incursion. Surveillance is a longterm strategy that requires funding and local capacity
building. Finally, an incursion response plan, in case an
IAS is detected or suspected, aims not only to confirm
the incursion but also to eliminate the IAS (Moore, et al.,
2010). Biosecurity protocols contemplate, at least, the
following aspects: 1) Identifying the main potential IAS; 2)
Identifying possible pathways and vectors of introduction;
3) Establishing prevention measures on the mainland; 4)
Establishing early detection systems at disembarking sites;
5) Establishing an incursion response plan; 6) Establishing
stakeholders responsibilities (PII, 2013).
Since 2014, we have held workshops for the participative
formulation of biosecurity protocols for our pilot areas
(and others). We invite local authorities (CONANP,
486
Socorro: 40 people at the
Navy Station
Clarion: 15 people at the
Navy Station
No permanent settlement,
however during fishing
season around 90 people
camp there. Highly visited
tourist spot.
Cayo Norte Mayor: 12
people Navy Station
Cayo Centro: 3 people
CONANP station, 100
people fishermen’s camps.
Tourist visitors.
Pérez: 15 people from the
Navy Station and CONANP
station. During fishing
tournaments around 40
camp.
SEMAR, port authorities), fishermen and tourist operators,
and we go through all stages of biosecurity and discuss the
sites most visited, frequency, and type of transportation.
Afterwards, we vote on prevention measures and where to
implement them. Additionally, we do a field practice about
surveillance and early detection devices commonly used.
To date, we have six unique, specific, updated, island
biosecurity protocols, created in a participatory manner.
The protocols contain priorities for prevention measures
and the most cost-effective and site-specific tools and
methods. Protocols are currently under review by the
corresponding authorities (Latofski-Robles, et al., 2017).
Protocols were formally validated through workshops with
the Advisory Council for each island. Furthermore, we
strive to create Biosecurity Committees that are a subgroup
of said Advisory Councils. These Committees will be in
charge of implementation, evaluation and updating of the
protocols, as well as fundraising for biosecurity to continue
in the long run.
Latofski-Robles, et al.: Mexico’s island biosecurity programme
ISLAND BIOSECURITY AT WORK
The most relevant component of biosecurity is
prevention. However, all stakeholders need to communicate
and coordinate in order for it to be effective. Prevention
is closely linked to outreach and environmental education
campaigns (Parkes, 2013). An analysis of costs from the
Mexican island experience, overwhelmingly demonstrates
the importance of investing in biosecurity prevention
measures. Recent rodent eradications in Mexico, show that,
on average, it costs 20 times more to perform an eradication
project than to prevent the arrival of IAS (Aguirre-Muñoz,
et al., 2017b).
Early detection is of critical importance to discover any
elusive individual that managed to escape the prevention
measures. Thus, it also helps to evaluate the prevention
strategy. Local capacity building, strong partnerships
and straightforward communication between local
communities, island managers and other stakeholders
(e.g. tourist operators) is critical for a swift and effective
incursion response. Furthermore, the ad hoc design and
wide distribution of outreach materials for each island is
vital to raise awareness of the problem of IAS.
As our National Biosecurity Programme unfolds, we
have had two effective incursion response events that
have successfully stopped the establishment of rodents in
Arrecife Alacranes. This is a positive sign that the outreach
campaign and workshops are having an effect, and that
people are now aware that islands should be IAS-free
and their involvement is needed to achieve that (LatofskiRobles, et al., 2016; Matos, et al., 2018). Much has been
learnt from incursion events, and the lessons must be
adopted nationwide to strengthen prevention measures and
community involvement.
INSTITUTIONALISING BIOSECURITY
The successful two-decade trajectory of island
restoration in Mexico contributes to meet the country’s goals
in sustainable development and conservation (AguirreMuñoz, et al., 2016a). The National Biosecurity Programme
must become a formally recognised, institutionalised, interagency, inter-sectorial agreement for it to be effective. We
need to establish collaboration arrangements with several
agencies, such as CONANP, SEMAR, SEMARNAT, the
Federal Agency for Environmental Protection (PROFEPA),
and port authorities. Once we are all working hand in hand,
the restoration efforts for Mexico’s island biodiversity will
be reinforced and protected over the long term.
RECOMMENDATIONS AND LESSONS LEARNT
Outreach and environmental learning campaigns are of
the utmost importance, and hence need to be permanent
and not just for short periods of time. Only then, will people
become aware of the problem and actually adopt the habits
required to prevent the accidental introduction of IAS.
Working with the Protected Areas Advisory Council
is the best strategy to strengthen biosecurity protocols. It
also helps the project to become integrated with the area
manager’s work.
Communities that recently participated on an
eradication project are more likely to be interested and
active in keeping the island free of IAS.
Incursion response cases may have economic costs
that are not specifically budgeted for, so the creation of a
national biosecurity fund for emergencies is an important
step forward.
Early detection alerts are a way of evaluating if the
outreach campaign is working, so that even if it turns out
to be just a false alarm, we now know people are aware that
they should report if they see something different.
Building capacity amongst protected area managers and
users regarding island biosecurity methods and techniques
is crucial to protect the islands from the impacts of IAS.
The threat of IAS is considered as important in most of the
protected areas management plans; however, preventing
their accidental introduction is not commonly featured.
We need to sign and publish institutional collaboration
agreements between government agencies in order
to reinforce biosecurity measures and make sure all
stakeholders comply with them.
The first step toward building biosecurity capacities for
the Mexican islands was the “Island Biosecurity Workshop
for managers, park rangers and users of protected areas”
in 2014. It was held by GECI with funding from the US
Fish & Wildlife Service and the CONABIO. Twenty-six
people from all island protected areas in Mexico, gathered
in Ensenada, Baja California for three days, during
which we discussed biosecurity measures and practiced
with early detection devices in Todos Santos Sur Island.
Representatives from all agencies regarding islands came
together. There were people from CONANP, CONABIO,
the Mexican Navy, and the SEMARNAT Office for
Wildlife (DGVS). We also analysed the challenges
and opportunities to implement biosecurity protocols,
prevention measures and early detection systems (MéndezSánchez, et al., 2014).
Formulating the biosecurity protocols is the work
of many people and agencies. We would like to thank
CONABIO, SEMAR, and CONANP and each protected
area’s Advisory Council for their collaboration. The
network of funding for biosecurity continues to grow; we
thank the GEF, UNDP, Alianza WWF- Fundación Carlos
Slim, Marisla Foundation, The David & Lucile Packard
Foundation, USFWS and GIZ.
Moreover, GECI has had a solid collaboration history
with the Mexican Navy (Secretaría de Marina, Armada de
Mexico). They are invaluable partners in the conservation
of the Mexican islands, always providing their support
on projects (Aguirre-Muñoz, et al., 2017b). We have
also had talks with their central offices about the need to
adopt biosecurity measures in every port and for all ships.
During our restoration projects, we give talks to personnel
at SEMARs stations at the islands, but we also strive to
provide training, so that there is always at least one person
who knows about surveillance methods and early detection
techniques on all islands with Navy stations.
ACKNOWLEDGEMENTS
REFERENCES
Aguirre-Muñoz, A., Samaniego-Herrera, A., Luna-Mendoza, L., OrtizAlcaraz, A., Rodríguez-Malagón, M., Méndez-Sánchez, F., FélixLizárraga, M., Hernández-Montoya, J.C., González-Gómez, R., TorresGarcía, F., Barredo-Barberena, J.M. and Latofski-Robles, M. (2011a).
‘Island restoration in Mexico: ecological outcomes after systematic
eradications of invasive mammals’. In: C.R. Veitch, M.N. Clout and
D.R. Towns (eds.) Island invasives: eradication and management, pp.
250–258. Occassional Paper SSC no. 42. Gland, Switzerland: IUCN
and Auckland, New Zealand: CBB.
Aguirre-Muñoz, A., Samaniego-Herrera, A., Luna-Mendoza, L., OrtizAlcaraz, A., Rodríguez-Malagón, M., Félix-Lizárraga, M., MéndezSánchez, F., González-Gómez, R., Torres-García, F., HernándezMontoya, J.C., Barredo-Barberena, J.M. and Latofski-Robles, M.
(2011b). ‘Eradications of invasive mammals on islands in Mexico: the
roles of history and the collaboration between government agencies,
local communities and a non-government organisation’. In: C.R. Veitch,
M.N. Clout and D.R. Towns (eds.) Island invasives: eradication and
management, pp. 386–394. Occassional Paper SSC no. 42. Gland,
Switzerland: IUCN and Auckland, New Zealand: CBB.
487
Island invasives: scaling up to meet the challenge. Ch 3A Strategy: Biosecurity
Aguirre-Muñoz, A., Méndez-Sánchez, F., De la Rosa-Conroy, L., LatofskiRobles, M. and Manríquez-Ayub, A. (2013). Diagnóstico de especies
exóticas invasoras en las reservas de la biosfera y áreas naturales
protegidas insulares seleccionadas, a fin de establecer actividades de
manejo para las mismas. Tercer informe de actividades presentado a
la Comisión Nacional para el Conocimiento y Uso de la Biodiversidad
dentro de la fase preparatoria del proyecto GEF “Aumentar las
capacidades de México para manejar especies exóticas invasoras”.
Grupo de Ecología y Conservación de Islas, A.C. Ensenada, Baja
California, México. 128 pp + 2 Anexos. <http://www.biodiversidad.
gob.mx/especies/Invasoras/gef/pdf/04-diagnostico-seleccion-gruposislas.pdf>.
Aguirre-Muñoz, A., Méndez-Sánchez, F. and Latofski-Robles, M. (2016a).
‘Conservación y restauración de las islas Mexicanas’. In: M. Llano and
H. Fernández (eds.) Análisis y Propuestas para la Conservación de la
Biodiversidad en México, Pp. 66–71. Ciudad de México.
Aguirre-Muñoz, A., Samaniego-Herrera, A., Luna-Mendoza, L., OrtizAlcaraz, A., Méndez-Sánchez, F. and Hernández-Montoya, J. (2016b).
‘La restauración ambiental exitosa de las islas de México: una reflexión
sobre los avances a la fecha y los retos por venir’. In: E. Ceccon and C.
Martínez-Garza (eds.) Experiencias Mexicanas en la Restauración de
los Ecosistemas, pp. 487–512. Ciudad de México: Comisión Nacional
para el Conocimiento y Uso de la Biodiversidad. Universidad Nacional
Autónoma de México. Universidad Autónoma del Estado de Morelos.
Aguirre-Muñoz, A., Méndez-Sánchez, F., Latofski-Robles, M. and
Salizzoni-Chávez, K. (2017a) ‘Mejoras en el marco legal para la
conservación y restauración integral de las islas de México’. In: Reporte
CESOP, no. 103, Febrero 2017. pp. 19–23. Publicación mensual del
Centro de Estudios Sociales y de Opinión Pública de la Cámara de
Diputados.
Aguirre-Muñoz, A., Méndez-Sánchez, F., Latofski-Robles, M., SalizzoniChávez, K., Luna-Mendoza, L., Ortiz-Alcaraz, A., Hernández-Montoya,
J., Bedolla-Guzmán, Y. and Rojas-Mayoral, E. (2017b). ‘Avances en la
conservación y restauración integral de las islas de México y mejoras
al marco legal para consolidar la trayectoria positiva’. In: G. BornSchmidt, F. de Alba, J. Parpal and P. Koleff (eds). Principales Retos que
Enfrenta México ante las Especies Exóticas Invasoras, pp. 225–255.
Ciudad de México: Centro de Estudios Sociales y de Opinión Pública.
Cámara de Diputados / LXIII Legislatura.
Aguirre-Muñoz, A., Bedolla-Guzmán, Y., Hernández-Montoya, J.,
Latofski-Robles, M., Luna-Mendoza, L., Méndez-Sánchez, F., OrtizAlcaraz, A., Rojas-Mayoral, E., and Samaniego-Herrera, A. (2018).
‘The conservation and restoration of the Mexican Islands, a successful
comprehensive and collaborative approach relevant for global
biodiversity’. In: Ortega-Rubio, A. (ed). Mexican natural resources
management and biodiversity conservation, pp. 177-192. Springer
International Publisher AG, part of Springer Nature.
Born-Schmidt, G., Parpal-Servole, J. and Koleff, P. (2017). ‘De la
elaboración a la implementación de la estrategia nacional sobre
especies invasoras’. In: Reporte CESOP, no. 103, Febrero 2017, pp.
5–9. Publicación mensual del Centro de Estudios Sociales y de Opinión
Pública de la Cámara de Diputados.
Broome, K. (2009). ‘Beyond Kapiti – a decade of invasive rodent
eradications from New Zealand islands’. Biodiversity 10: 14–24.
CANEI (Comité Asesor Nacional sobre Especies Invasoras). (2010).
Estrategia nacional sobre especies invasoras en México, prevención,
control y erradicación. Ciudad de México: Comisión Nacional para
el Conocimiento y Uso de la Biodiversidad, Comisión Nacional de
Áreas Naturales Protegidas, Secretaría de Medio Ambiente y Recursos
Naturales.
CANTIM (Comité Asesor Nacional sobre el Territorio Insular Mexicano).
(2012). Estrategia Nacional para la conservación y el desarrollo
sustentable del territorio insular mexicano. Ciudad de México y
Ensenada, B.C: Secretaría de Medio Ambiente y Recursos Naturales,
Instituto Nacional de Ecología, Comisión Nacional de Áreas Naturales
Protegidas, Secretaría de Gobernación, Secretaría de Marina-Armada
de México y Grupo de Ecología y Conservación de Islas, A.C.
CBD (Convention on Biological Diversity). (2010). Strategic Plan for
Biodiversity 2011–2020. Aichi Targets. <https://www.cbd.int/sp/
targets/>. Accessed: 27 June 2017.
DOF (Diario Oficial de la Federación). (2016). Decreto por el que
se declara área natural protegida, con el carácter de reserva de la
biosfera, la región conocida como islas del pacífico de la península de
Baja California. Ciudad de Mexico.
González-Martínez, A.I., Barrios-Caballero, Y. and De Jesús, S. (2017).
‘Análisis de riesgo para especies invasoras en México’. In: Reporte
CESOP, no. 103, Febrero 2017, pp. 24–29. Publicación mensual del
Centro de Estudios Sociales y de Opinión Pública de la Cámara de
Diputados.
488
IUCN (International Union for Conservation of Nature). (2017).
Commitments toward achieving the Honolulu Challenge. <https://www.
iucn.org/theme/species/our-work/invasive-species/honolulu-challengeinvasive-alien-species/commitments-towards-achieving-honoluluchallenge>. Accessed: 28 June 2017.
Latofski-Robles, M., Aguirre-Muñoz, A., Méndez-Sánchez, F., ReyesHernández, H. and Schlüter, S. (2014). ‘Prioritizing restoration actions
for the islands of Mexico’. Monographs of the Western North American
Naturalist 7: 435–441.
Latofski-Robles, M., Méndez-Sánchez, F., Aguirre-Muñoz, A., JáureguiGarcía, C. and Castro-Girón, A. (2016). Diagnóstico de especies
exóticas invasoras en seis áreas naturales protegidas insulares, a fin de
establecer actividades para su manejo. Reporte de actividades del año
1 (2015) dentro del proyecto GEF 0081866 “Aumentar las capacidades
de México para manejar especies exóticas invasoras a través de la
implementación de la Estrategia Nacional de Especies Invasoras”.
Grupo de Ecología y Conservación de Islas, A.C. Ensenada, B.C.,
México.
<http://www.biodiversidad.gob.mx/especies/Invasoras/gef/
pdf/2.1-1-reporte-extenso-gef-2015.pdf>.
Latofski-Robles, M., Méndez-Sánchez, F., Aguirre-Muñoz, A., JáureguiGarcía, C., Salizzoni-Chávez, K. and Fernández-Ham, G. (2017).
Diagnóstico de especies exóticas invasoras en seis áreas naturales
protegidas insulares, a fin de establecer actividades para su manejo.
Reporte de actividades del año 2 (2016) dentro del proyecto GEF
0081866 “Aumentar las capacidades de México para manejar especies
exóticas invasoras a través de la implementación de la Estrategia
Nacional de Especies Invasoras”. Grupo de Ecología y Conservación
de Islas, A.C. Ensenada, B.C., México. <http://www.biodiversidad.gob.
mx/especies/Invasoras/gef/pdf/2.1-1-reporte-extenso-gef-2016.pdf>.
Llorente-Bousquets, J. and Ocegueda-Cruz, S. (2008). ‘Estado del
conocimiento de la biota’. In: J. Soberón, G. Halffter, and J. LlorenteBousquets, (Eds.), Capital natural de México Vol. I Conocimiento
actual de la biodiversidad, pp. 283–322. México, D.F.: Comisión
Nacional para el Conocimiento y Uso de la Biodiversidad.
Matos, J., Little, A., Broome, K., Kennedy, E., Méndez-Sánchez, F.,
Latofski-Robles, M., Irvine, R., Gill, C., Espinoza, A., Howald, G.,
Olthof, K., Ball, M. and Boser, C.L. (2018). ‘Connecting island
communities on a global scale: Case studies in island biosecurity’.
Monographs of the Western North American Naturalist 78(4): 959-972
Méndez-Sánchez, F., Aguirre-Muñoz, A., Latofski-Robles, M., JáureguiGarcía, C., Luna-Mendoza, L, Ramos-Rendón, K., De la Rosa-Conroy,
L., Félix-Lizárraga, M., Manríquez-Ayub, A., Castañeda-Rodríguez, I.
and Guzmán-Cázarez, E. (2014). Curso-taller de bioseguridad insular
para administradores guardaparques y usuarios de áreas naturales
protegidas. Final Report to U.S. Fish & Wildlife Service, Wildlife
Without Borders, Division of International Conservation. Grupo de
Ecología y Conservación de Islas, A.C. Ensenada, B.C. 28 pp + 1 Anexo.
Moore, J.L., Rout, T.M., Hauser, C.E., Moro, D., Jones, M., Wilcox, C.
and Possingham, H.P. (2010). ‘Protecting islands from pest invasion:
optimal allocation of biosecurity resources between quarantine and
surveillance’. Biological Conservation, 143(5):1068–1078.
PII (Pacific Invasives Initiative). (2013). Biosecurity Plan Template.
Resource Kit for Rodent and Cat Eradication. <http://rce.
pacificinvasivesinitiative.org/tools/Templates.html >. Accessed 3 April
2018.
Parkes, J. (2013). Strategies to detect and manage incursions of exotic
species on Mexican islands. Report presented to UNDP-Mexico and
CONABIO as part of the preparatory phase of the GEF-ENCIS Project.
Christchurch, New Zealand: Kurahaupo Consulting. <http://www.
biodiversidad.gob.mx/especies/Invasoras/gef/pdf/09-edrr-islands-en.
pdf>.
Reaser, J.K., Meyerson, L.A., Cronk, Q., De Poorter, M., Eldrege,
L.G., Green, E. and Vaiutu, L. (2007). ‘Ecological and socioeconomic
impacts of invasive alien species in island ecosystems’. Environmental
Conservation 34: 98–111.
Roberts, A. (2003). Island biosecurity SOP Best Practice Manual.
Southland, New Zealand: Department of Conservation.
Russell, J.C., Towns, D.R. and Clout, M.N. (2008). ‘Review of Rat
Invasion Biology: Implications for Island Biosecurity’. Science for
Conservation 286. Wellington, NZ: Department of Conservation.
Towns, D.R. (2011). ‘Eradications of vertebrate pests from islands around
New Zealand: what have we delivered and what have we learned?’.
In: C.R. Veitch, M.N. Clout and D. Towns (eds.) Island invasives:
eradication and management, pp. 364–371. Occassional Paper SSC no.
42. Gland, Switzerland: IUCN and Auckland, New Zealand: CBB.
Whittaker, R.J. and Fernández-Palacios, J.M. (2007). Island
Biogeography: Ecology, Evolution, and Conservation. Oxford: Oxford
University Press.
M.G. Richardson and J.P. Croxall
Richardson, M.G. and J.P. Croxall. Achieving post-eradication biosecurity on South Georgia
Achieving post-eradication biosecurity on South Georgia
M.G. Richardson and J.P. Croxall
South Georgia Heritage Trust, Verdant Works, West Henderson’s Wynd, Dundee, DD1 5BT, UK.
<mikerichardson@btinternet.com>.
Abstract The world’s largest island rodent eradication programme to date was carried out on South Georgia between
2009 and 2018 (baiting on island in 2011, 2013 and 2015) by the South Georgia Heritage Trust (SGHT). A comprehensive
survey in 2017/18 found no signs of rodents. Although SGHT planned and executed this eradication under permits issued
by (and with collaboration from) the Government of South Georgia and the South Sandwich Islands (GSGSSI), the scale
and complexity of the multi-year project was not conducive to prior agreement on post-eradication biosecurity to prevent
rodent re-invasion. Thus by 2013, two years after initial baiting, biosecurity measures for rodents remained inadequate,
relying mainly on rodent detection boxes on vessels and at the island’s main point of entry. The more substantive posteradication biosecurity measures implemented by other administrations were absent. In late 2014, after more than three
years with no rodent sign, there was unambiguous evidence of a rat within the island’s settlement. This coincided with
a vessel berthed alongside a nearby jetty. Between 2015 and 2017, SGHT formally submitted recommendations to the
GSGSSI on enhanced biosecurity provisions. Some of these recommendations have been implemented but arguably the
most important, relating to vessel berthing and secure handling of imported cargo, remain to be addressed effectively. We
summarise what remains to be done, recognising the logistic and financial challenges involved, but conscious that, until
all measures are in place, there is significant risk of re-infestation of South Georgia by invasive rodents, compromising a
decade of work (and funding) by multiple stakeholders.
Keywords: baiting, poison, rodents, sub-Antarctic
INTRODUCTION
Tentative plans for the possible eradication of rodents
from the sub-Antarctic island of South Georgia date
back to 2000 (S. Poncet, pers. comm.). The subsequent
success of the eradication programme on Campbell Island
(McClelland & Tyree, 2002; Towns & Broome, 2003)
encouraged the Government of South Georgia and the
South Sandwich Islands (GSGSSI) to undertake a feasibility
study of the practicalities of a large-scale (island-wide)
eradication of rats and mice on South Georgia (Christie &
Brown, 2007). However, due to resource limitations at the
time, the Government opted not to proceed.
Nevertheless, the small UK charity (NGO) South
Georgia Heritage Trust (SGHT) agreed to take up the
challenge and started to develop fundraising and project
management structures and initiatives to address this.
GSGSSI accepted the involvement and lead role of SGHT
in principle and practice, subject to the project conforming
to the relevant legislation and permitting processes.
In 2009, SGHT established a Steering Committee (SC)
to oversee the management of the whole operation. The SC
comprised Trustees from SGHT and the Friends of South
Georgia Island (FOSGI), key GSGSSI officials (Chief
Executive and Environmental Officer), representatives of
British Antarctic Survey (BAS), and the SGHT Project
Director (Prof. Tony Martin of the University of Dundee).
The SC met quarterly from 2010 to 2015 and its main
roles were to ensure the effective execution of the plans
for the acquisition and shipment of equipment, vessels,
helicopters and staff, and that all documentation required
by the regulatory authorities (mainly GSGSSI but also
the UK's Civil Aviation Authority (CAA)) was submitted
on time. A list of all such documentation can be found at
<http://www.sght.org/newsletters-and-publications/>.
From 2010, there followed three phases of baiting using
brodifacoum poison bait distributed by helicopters. Each
baiting phase was spaced two years apart (2011, 2013 and
2015) to allow both for further fundraising between baiting
seasons and for evaluation of methods and results before
proceeding further. This work is reported on elsewhere
(Martin, 2015; SGHT, 2016; Martin & Richardson, 2017).
In the 2017/18 austral summer, a comprehensive
monitoring survey, organised and led by SGHT in
collaboration with GSGSSI, was undertaken to determine
the results of the eradication project. The five-month survey
deployed over 1,500 inert devices (chew-sticks, tunnel and
camera traps and analogous devices) and, augmented by
trained rodent-detection dogs (which travelled 2,420 km),
covered a minimum of 8,600 ha across 120 sites. No signs
of rodents were detected, allowing the conclusion that the
eradication phases had been successful.
This paper aims to review the rodent-related biosecurity
status of South Georgia before and during the eradication
project and to summarise proposals to enhance this in the
light of events during the project, and after its successful
conclusion. It highlights the remaining measures to be
implemented to minimise the risk of inadvertent reintroduction of rodents.
RODENT BIOSECURITY AT SOUTH GEORGIA
PRIOR TO 2014
The need for biosecurity measures to be integral to any
eradication efforts on South Georgia was recognised back
in 2007, with a governmental report stating then that: 'First
and foremost, an effective and robust biosecurity regime
needs to be in place on South Georgia before eradication
is attempted' (Christie & Brown, 2007).
Although SGHT submitted Biosecurity Plans to
GSGSSI for each of the three phases of baiting, those plans
dealt with biosecurity solely in relation to the operational
requirements of the project itself – for example, the
importation into South Georgia of materials needed for the
baiting operation, or the movement of equipment, including
helicopters, within the island. The wider issue of South
Georgia's biosecurity (the responsibility of Government),
was not addressed in discussions between SGHT and
GSGSSI either before or during the earlier years of the
eradication project. In hindsight, this lapse was the result
of both organisations trying at that time to cope with the
considerable challenges of the baiting operations. Faced
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 489–493. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
489
Island invasives: scaling up to meet the challenge. Ch 3A Strategy: Biosecurity
with what was clearly going to be a multi-year, complex
operation it would have been difficult in the initial 'proof-of
concept' stages of the project to have developed a realistic
and pragmatic prescription for post-eradication biosecurity.
In consequence, biosecurity arrangements were held, at
least in respect of rodents, under relatively rudimentary
provisions. For example, the governmental Biosecurity
Protocols of 2013 and 2014 (GSGSSI, pers. comm. 2013
and 2014) did little more than stipulate the need for rat
guards on vessels, the deployment of rodent bait boxes (a
requirement for yachts only), and the requirement that all
vessels be inspected for the presence of rodents.
Despite what, in retrospect, was a deficiency in project
planning, the HR Project progressed well. By mid-2014
(more than three years after the initial baiting) the Phase
1 area (c. 14,000 ha) had been tentatively declared free of
rodents, and a relatively extensive survey in March of that
year by SGHT detected no signs of rodents in the 60,000
ha. of the more extensive Phase 2 area.
Deficiencies in the biosecurity provisions became
evident on 23 October 2014 when the unambiguous signs of
a rat were seen in newly fallen snow at King Edward Point
(KEP) – the administrative centre of the island in the heart
of the Phase 1 area. The Government rapidly set in train its
contingency plan for just such an incident. Brodifacoum
bait was spread by hand out to an arc perimeter of 1.5 km
from the sighting and many more rat traps were placed
around the KEP base area. In the event, no more sign of
this animal was seen; nor was a corpse found. This was
unfortunate since, through DNA analysis, the origin of this
lone animal could have been determined (see Piertney, et
al., 2016). The presumption was that the rat succumbed to
the poison bait.
The origins of this one known rat could only be
speculated on. It could have been a survivor (or offspring of
a survivor) from the 2011 baiting phase. However, this is
unlikely in the most inhabited part of South Georgia where
there had been no rodent signs in the preceding 3.5 years
since baiting. Alternatively, it could have been imported
in one of the small vessels based at KEP from another part
of the island yet to be baited or swam ashore from a vessel
anchored offshore. The latter scenarios are not impossible
but seem implausible. Given the coincidental timing, the
most probable source for this rat was from one or other of
two vessels that had recently tied-up alongside the nearby
KEP jetty. Records showed that one vessel had visited a
number of times between 5–22 October whilst another
vessel had arrived on 22 October and was still moored
alongside the jetty the following day at the time of the
incident (GSGSSI, in litt. to SGHT).
The general conclusion was that this latter vessel was
the most likely source of this incursion. The Government
concurred through a statement that "the rat was most
likely to have originated from a ship tied up at KEP in
the previous days, though it was impossible to prove this"
(GSGSSI, 2015 in litt. to SGHT).
Whatever the means of introduction, this rat had
managed to evade all prevention and detection measures
in place at the time – bait boxes and traps deployed both
on the vessels and extensively around the base area. Its
presence was only detected due to recent snow cover.
RODENT BIOSECURITY AT SOUTH GEORGIA
SINCE 2014
Although this incident apparently involved only a single
animal, SGHT assumed that it would rapidly trigger a major
Government-led review of biosecurity arrangements, in
order to implement more robust measures. However, it
490
was mid-2015, following completion of the last phase of
baiting, that the Government requested SGHT input to
an apparent major review of South Georgia's biosecurity
arrangements. The SGHT response, submitted in late July,
was a series of 10 recommendations to enhance islandwide biosecurity (Table 1).
These recommendations were based on the fact that,
with aircraft unable to operate into South Georgia, the
re-introduction of rodents to South Georgia could only
come about via shipping. That is: by shipwrecks on the
coast, or by animals swimming ashore from a vessel;
gaining access along mooring warps or down gangplanks;
or coming ashore in cargo or luggage. Although none of
these potential introduction pathways can be ruled out, the
greatest risk of a rodent re-introduction to South Georgia
is most likely to be via one or other of the last two routes.
SGHT's recommendations included the requirement
to maintain an adequate supply (at least three tonnes) of
in-date brodifacoum bait at KEP, the need for a series of
pre-baited box traps (which would be inspected frequently)
around the base area, and the deployment of effective ratguards on vessels moored alongside.
The Trust's four main recommendations are shown in
bold in Table 1. These were: the use of rodent-detection
dogs at ports in the Falkland Islands and on vessels destined
for South Georgia; prohibiting the mooring alongside of
vessels except for tightly prescribed activities; the erection
of rodent-proof fences around offloading jetties in South
Georgia; and the construction of rodent-proof containment
areas at KEP within which shipping containers and other
large-scale cargo could be held, and unpacked, in a
biosecure manner.
Totally eliminating the risk of a rodent reintroduction
to South Georgia cannot be guaranteed. However, SGHT
was of the view that comprehensive implementation of
its recommendations would very substantially reduce the
risk of rodents either getting to South Georgia in the first
place or, if that failed, at least preventing their escape
from the immediate surroundings of the cargo unloading/
handling areas at KEP/Grytviken. The recommendations
were considered by SGHT to be necessary, realistic,
practical, cost-effective (especially in terms of the cost of
mounting a subsequent eradication operation) and based
on international best practice.
The presumption was that these proposed provisions
would be included within a new, strengthened
governmental Biosecurity Plan. Instead, the Biosecurity
Handbook, published in December 2015 (GSGSSI, 2015)
simply re-stated the existing provisions. It took no account
of the SGHT recommendations. This caused SGHT to
re-state its case to GSGSSI in January 2016 and again in
February 2017 (SGHT, 2016; 2017 in litt. to GSGSSI).
Unfortunately, we are unaware of any substantive change in
biosecurity practices at South Georgia, with one important
exception, relating to the trial use of rodent-detection dogs
(see below).
IMPLEMENTATION OF BEST PRACTICE
BIOSECURITY AT SOUTH GEORGIA
GSGSSI's policy, in principle, over biosecurity is
rightly predicated (as is best practice) on the concept that:
"The most effective way of dealing with biosecurity is to
have pre-border measures in place...the aim is to prevent
an alien reaching the island, not try and deal with it on
arrival" (Christie & Brown, 2007). Furthermore, in its
five-year Strategy Paper (GSGSSI, 2016) the Government
advocated that "Biosecurity protocols should be reviewed
on a regular basis and best practice adopted"
Richardson & Croxall: Biosecurity on South Georgia
Table 1 Biosecurity recommendations* submitted to South Georgia Government (GSGSSI) – 2015/16 and 2017.
Recommendation
Vessels/cargo checked by rodent-detection dogs (in Falklands) then during transit
to, and on arrival at, South Georgia.
Vessels (other than yachts) must be prohibited from mooring alongside except
when unloading/loading cargo or other strictly prescribed activities; then for
minimum time only. In all other circumstances vessels must either anchor off or
moor to buoy.
When moored alongside, or to the shore, all mooring warps must have effective rat
guards fitted.
When moored alongside, gangway ashore must only be in place when necessary, and
for minimum time.
Rodent-proof fence must be constructed around every dock area (KEP/
Grytviken)
No loose cargo (other than personal effects) must be offloaded. All cargo must be
carried in sealed shipping containers which must be (a) fumigated, and (b) contain
rodent bait stations.
A rodent-proof containment area suitable for shipping containers must be
constructed at KEP. Containers must only be opened and unpacked within the
containment area.
Implementation
Trial underway in 2018
Only partial
Yes, though design of guards
needs further attention
?
None
Only partial
Under consideration.
Construction potentially
starting 2019; completion
2021?
In the event of known or suspected rodent incursion, pre-planned response action must Yes
be activated immediately. Must include setting of traps and spreading poison out to
stipulated radius from incursion.
A network of pre-baited trap boxes must be installed permanently around any dock
Yes
area and checked frequently (daily when vessel moored alongside).
Suitable quantity (minimum 3 tonnes) in-date brodifacoum bait must be held at KEP
Yes, but whether in date is
as contingency. Such bait must be replenished as appropriate.
unknown
* Recommendations in bold are the most substantive ones.
In relation to the recommendations of SGHT (Table 1)
and practices prevailing at the closest analogue operation,
that following the comprehensive and successful
eradication programme on Macquarie Island (Springer,
2016), we review the current situation at South Georgia
below:
a) Pre-border measures: cargo checking on
embarkation and in transit
In its most recent policy statement (GSGSSI, 2017),
pre-border biosecurity measures in relation to rodents rely
principally on the use of rat-guards on vessels, requiring
rodent detection boxes to be carried onboard vessels and
the use of bait stations within cargo shipping containers.
However, even taken together, we contend that these
measures are unlikely to be effective. For example, despite
GSGSSI having trialled a number of rat-guard designs,
none to date has proved capable of coping with the harsh
weather conditions prevailing in South Georgia.
At Macquarie Island, the deployment of rodentdetection dogs is now routine. Dogs check all cargo twice
before it departs Australia and then again during passage
to, and on arrival, at Macquarie where the environs of the
research station are then also subject to survey by dogs.
With financial assistance from SGHT, GSGSSI embarked
in early 2018 on a trial deploying rodent-detection dogs
at embarkation points in the Falkland Islands and on
ships destined for South Georgia. This is a very welcome
initiative which, it is hoped, will be converted into a
permanent procedure.
Nevertheless, the most likely pathway for a rodent
to gain access to South Georgia is from vessels moored
alongside a jetty at KEP/Grytviken either via offloaded
cargo or by simply "jumping ship".
b) Vessel mooring
At Macquarie Island, the risk of further rodent
invasion is reduced still further by there being no harbour
or jetty facilities on the island. This means that, unlike
South Georgia, all cargo transfers from ship to shore are
performed either by helicopters or amphibious lighters,
enabling more stringent biosecurity checks to be made.
Vessels anchoring well offshore, beyond the swimming
distance of rats, reduce the risk that any shipborne rodents
may gain direct access to the island through their ability
(documented in both the UK and Falklands) to swim up
to 1–2 km between, or out to, islands. The Macquarie
situation has the added advantage that ship movements
are confined largely to transits between Tasmania and the
island. This again enables far greater biosecurity control.
At South Georgia, in contrast, whilst ships depart
to South Georgia from a variety of locations (e.g. South
American ports), those that are currently allowed to tieup alongside at KEP are invariably governmental vessels
arriving from the Falklands, where the embarkation
ports are known to be infested with rats and lack fully
appropriate facilities for biosecure handling of cargo.
SGHT has recommended that the practice of mooring
alongside, the most likely route for a rodent incursion,
should be prohibited, except for cargo handling and
other closely prescribed activities (such as undertaking
necessary mechanical repairs to a vessel or for safety).
The current criteria allowing alongside mooring include
activities such as the "transfer of personnel, or "allowing
for crew rest periods". Given the biosecurity risks that
alongside mooring poses, convenience per se should not be
a valid justification for continuation of this practice. This
is particularly so given that the many thousands of tourists
who visit South Georgia (and KEP/Grytviken) annually do
so from vessels anchored offshore.
491
Island invasives: scaling up to meet the challenge. Ch 3A Strategy: Biosecurity
c) Cargo-handling ashore
Biosecurity handling facilities at KEP are currently
restricted to a single shed where small-scale cargo can be
unpacked and checked within a confined space. There are
no facilities, in the form of a rodent-proof fenced area,
within which large-scale cargo (i.e. shipping containers)
can be stored and then opened securely. Such a facility,
coupled with rodent-proof fencing around the jetty areas at
KEP/Grytviken, would provide some measure of constraint
for any rodent that either managed to escape from a vessel
or survived inside a shipping container.
It is evident that the rodent incursion of October 2014
was of an animal that had apparently circumvented both
shipboard measures and the numerous traps and bait boxes
around KEP. Those measures, at least then, had proved
wanting.
BIOSECURITY COSTS AND RISKS
It is important to contrast the respective costs of
eradication and biosecurity. Investment in the eradication
project by SGHT has been considerable, with direct costs
of around £7.5 million, rising to c. £10 million when
indirect costs are included. Over 80% of the project
costs have been raised through charitable donations and
sponsorship; although GSGSSI provided extensive staff
and logistic assistance throughout the project it made no
other contribution to direct costs. In contrast, we estimate
that the capital costs of the additional recommended
biosecurity measures are unlikely to exceed £0.5 million.
In December 2016, SGHT offered to fundraise to help pay
for those capital costs.
SGHT recognises that implementation of the more
substantive measures would come with additional costs
(including ongoing maintenance costs), alternative risks,
and the need for specific design considerations to meet
South Georgia's harsh conditions. For example, any rodentproof structure on the island must be able to withstand
extremely high wind loadings, ice and snow accumulation
as well as the attention of other wildlife such as southern
elephant seals (Mirounga leonina).
Even significantly reducing the practice of alongside
mooring would not entirely eliminate risk. The alternatives
are either anchoring offshore or mooring to a suitable
buoy. Large vessel buoyage is no longer available at South
Georgia and its provision and maintenance would be both
expensive and not without liability for the regulatory
authority (GSGSSI). Vessels at anchor can also be at risk.
Weather conditions at South Georgia can change at short
notice and be severe. In March 2000, three long-lining
fishing vessels were driven ashore one night in Cumberland
Bay in extreme weather conditions. One managed to
re-float; the other two were completely wrecked on an
inshore reef. Whether there were rodents on either of those
vessels, and whether they escaped ashore is not known
but the incident emphasises that some risks of rodent reintroduction will always remain. This makes it even more
imperative to address those risks which can be mitigated
or eliminated.
CONCLUSIONS
The South Georgia Habitat Restoration Project has
been the largest island rodent eradication yet undertaken.
The overall effectiveness of the three seasons of baiting
over five years with brodifacoum has recently been
confirmed following a comprehensive monitoring survey
in the 2017/18 season. This found no signs of rodents in
any of the baited areas, leading to the conclusion that the
eradication programme has been successful. Concurrently,
492
the increase in the numbers and distribution of some
breeding birds (including endemic species) since baiting,
has already been dramatic.
Notwithstanding that result, two major lessons can
be taken from this important project; one highly positive,
the other less so. On the former, large scale eradications
(e.g. Campbell Island (McClelland & Tyree, 2002) and
Macquarie Island (Springer, 2016)) have usually relied
on the extensive resources of governments. In contrast,
the South Georgia project has demonstrated that, through
extensive fundraising and competent project planning and
implementation, even relatively modest or small-sized
NGOs can take on the challenge of large-scale eradications.
The downside of the South Georgia operation has
been the lack of a close synergy between eradication
and biosecurity. Again, previous large eradications have
had the benefit of intra-governmental co-ordination with
often the same governmental agency (e.g. New Zealand's
Department of Conservation or the Tasmanian Parks
and Wildlife Service) having responsibility for both
elements. The problems in the case of South Georgia were
complicated by the fact that two organisations, of highly
contrasting status, undertook or were responsible for the
eradication and biosecurity elements.
To ensure that these two equally important aspects are
taken forward in close harmony, we make the following
recommendations for future rodent eradication projects.
That:
●
adequate biosecurity measures must be in place
before, during and after eradication;
●
in those instances where responsibilities for
eradication and biosecurity may reside with
different organisations, agreement must be reached
in advance between those entities; and that:
●
such agreements should set out the respective
responsibilities, objectives and timetables for both
parties before eradication is allowed to commence.
Experience from South Georgia has shown that in the
absence of any such prior agreements, eradication and
biosecurity may get out of step either in their timing or
effectiveness – or both. Such a situation creates a potential
risk that the considerable investment in eradication and its
corresponding environmental benefits may be jeopardised
subsequently by inadequate biosecurity provisions.
ACKNOWLEDGEMENTS
As Chair of the SC, MGR would like to thank all
members of that Committee, not least the Project Director,
Prof. Tony Martin of Dundee University. We would also pay
tribute to the Island Foundation, Prof. Frederick Paulsen
and the Paulsen Foundation, without whose massive
generosity, particularly in start-up costs, the whole project
could never have commenced. Fundraising also relied in
a large part on FOSGI (the US counterpart of SGHT).
Our thanks go also to all staff involved throughout the
near-decade of the project and to the hundreds of donors,
large and small, individuals, trusts and foundations, who
have helped bring this eradication project to a successful
conclusion.
REFERENCES
Christie, D. and Brown, D. (2007). Recommendations and Concerns
Regarding Proposed South Georgia Rat Eradication. Stanley, Falkland
Islands: Report prepared for the Government of South Georgia and
the South Sandwich Islands. <http://www.sght.org/newletters-and
publications/>.
Richardson & Croxall: Biosecurity on South Georgia
GSGSSI. (2015). Biosecurity Handbook. Stanley, Falkland Islands
Government of South Georgia and the South Sandwich Islands. <http://
www.gov.gs/docsarchive/Environment/Biosecurity/Biosecurity%20
Protocols%202015.pdf>.
GSGSSI. (2016). South Georgia and the South Sandwich Islands Strategy
2016–2020. Stanley, Falkland Islands: Government of South Georgia
and the South Sandwich Islands. <http://www.gov.gs/docsarchive/
GSGSSI/Strategy/Final%20Published%20Strategy%20-%20PDF%20
Version.pdf>.
GSGSSI. (2017). Biosecurity Handbook 2017–2018. Stanley, Falkland
Islands: Government of South Georgia and the South Sandwich
Islands.
<http:www.gov.gs/docsarchive/Environment/Biosecurity/
Biosecurity%20Handbook2017-2018FINAL.pdf>.
Martin, A.R. (2015). Reclaiming South Georgia. South Georgia Heritage
Trust, Dundee.
Martin, A.R. and Richardson, M.G. (2017). ‘Rodent eradication scaled
up: clearing rats and mice from South Georgia’. Oryx: doi.org/10.1017/
S003060531700028X.
McClelland, P. and Tyree, P. (2002). ‘Eradication: the clearance of
Campbell Island’. New Zealand Geographic 58: 86–94.
Piertney, S.B., Black, A., Watt, L., Christie, D., Poncet, S. and
Collins, M.A. (2016). ‘Resolving patterns of population genetic and
phylogeographic structure to inform control and eradication initiatives
for brown rats Rattus norvegicus on South Georgia’. Journal of Applied
Ecology 53: 332–339.
SGHT. (2016). The South Georgia Habitat Restoration Project; Final
Report. Dundee, Scotland, UK: South Georgia Heritage Trust.
<http://www.sght.org/wp-content/uploads/2018/07/SGHT-HabitatRestoration-Project-final-report.pdf>.
Springer, K. (2016). ‘Methodology and challenges of a complex multispecies eradication in the sub-Antarctic and immediate effects of
invasive species removal’. New Zealand Journal of Ecology 40: 273–
278.
Towns, D.R. and Broome, K.G. (2003). ‘From small Maria to massive
Campbell: forty years of rat eradication from New Zealand islands’.
New Zealand Journal of Zoology 30: 377–398.
493
K.X.T. Bellis, R.T. Peet, R.L. Irvine, G. Howald and G.J. Alsop
Bellis, K.X.T.; R.T. Peet, R.L. Irvine, G. Howald and G.J. Alsop. Beyond biodiversity: the
cultural context of invasive species initiatives in Gwaii Haanas
Beyond biodiversity: the cultural context of invasive species initiatives
in Gwaii Haanas
K.X.T. Bellis1, R.T. Peet2,3, R.L. Irvine2, G. Howald4,5 and G.J. Alsop1,3
Council of the Haida Nation, P.O. Box 98, Queen Charlotte, BC, Canada, V0T 1S0. 2Gwaii Haanas National Park
Reserve, National Marine Conservation Area Reserve, and Haida Heritage Site, P.O. Box 37, Queen Charlotte, BC,
Canada, V0T 1S0. <robyn.irvine@canada.ca>. 3Gwaii Haanas Archipelago Management Board. 4Island Conservation,
2100 Delaware Ave, Suite 1, Santa Cruz, CA, USA, 95060. 5Island Conservation, 1531 Appleridge Rd, Kelowna, British
Columbia, Canada, V1W 3A5.
1
Abstract Haida Gwaii is a remote island archipelago located off the north Pacific coast, approximately 100 km from
mainland British Columbia, Canada. The southern part of the archipelago, Gwaii Haanas, is designated a National Park
Reserve, National Marine Conservation Area Reserve, and Haida Heritage Site, and is cooperatively managed by the
Government of Canada and the Council of the Haida Nation. Haida Gwaii has been the site of many invasive mammal
introductions including rats, racoons and deer. It was also where the first successful North American rat eradication
took place, on Langara Island in the 1990s. Since then, a number of invasive mammal eradication and control projects
have taken place in Gwaii Haanas. Most recently, a project to eradicate invasive deer on six islands began in 2017 and
is ongoing. The conservation gains from eradications are well documented in various systems around the globe, but the
cultural context in Gwaii Haanas is unique. The Gwaii Haanas experience presents an example of how active restoration
can move beyond biodiversity to develop IAS projects that are also culturally meaningful.
Keywords: cooperative management, cultural resource, Gwaii Haanas, Haida Gwaii, traditional knowledge, traditional
use
INTRODUCTION
Invasive alien species (IAS) are a significant driver
of declines in species and ecosystem services worldwide
(Pejchar & Mooney, 2009). In Canada, IAS negatively
affect 22% of endangered species (Wilcove, et al., 1998;
Venter, et al., 2006). Although invasion science spans
the fields of social science, ecology, economics and
conservation biology (Simberloff, et al., 2013), it rarely
considers the cultural context in which invasive species
eradications take place. For example, the impacts of IAS
on cultural practices including the harvest of traditional
foods and medicines are not usually considered.
In this paper, we describe IAS management in Gwaii
Haanas, a land-and-sea protected area located in northern
North America that is cooperatively managed by the Council
of the Haida Nation and the Government of Canada. The
Gwaii Haanas experience shows how cooperation between
indigenous and federal governments in IAS initiatives
allows projects to expand beyond biodiversity to become
culturally meaningful as well.
STUDY AREA
Haida Gwaii is an archipelago of over 350 islands
located approximately 100 km off the coast of British
Columbia, Canada. Gwaii Haanas is a 5,000 km2 protected
area in the southern third of Haida Gwaii. It is known for
its diverse ecosystems, distinctive flora and fauna, rich
marine life, and living Haida culture.
Gwaii Haanas first garnered international attention in
1985, when the Council of the Haida Nation (CHN) led
a non-violent blockade on Athlii Gwaay (Lyell Island)
to protest against logging in the area. The CHN declared
Gwaii Haanas (terrestrial and marine) a Haida Heritage
Site in the same year. Soon after, the Government of
Canada designated the Gwaii Haanas terrestrial area a
National Park Reserve, and cooperative management of
the land began in 1993 when the two governments signed
the Gwaii Haanas Agreement. In 2010, the CHN and the
Government of Canada signed the Gwaii Haanas Marine
Agreement (2010), which committed the two governments
to cooperative management of the Gwaii Haanas marine
area. In the same year, Gwaii Haanas was designated
a National Marine Conservation Area Reserve by the
Government of Canada.
Gwaii Haanas is cooperatively managed by the Council
of the Haida Nation and the Government of Canada
through the Archipelago Management Board (AMB).
The AMB is made up of equal representation from the
Haida and Canadian governments and is responsible for
all aspects of planning, operation, management and use of
Gwaii Haanas. Decisions are made by consensus and based
on the constitutions of both nations (Canadian Constitution
Act, 1982, Haida Nation Constitution, 2014).
ECOLOGICAL CONTEXT
Haida Gwaii is ecologically significant in a global
context. Supporting approximately 1.5 million breeding
seabirds of 12 different species, it is the nesting location for
one half of the entire global breeding population of ancient
murrelets (Synthliboramphus antiquus) and for one fifth
of the world’s breeding Cassin’s auklets (Ptychoramphus
aleuticus) (Harfenist & Cober, 2003). Many of these
important seabird breeding areas are situated within Gwaii
Haanas.
There are 10 extant native mammals in Haida Gwaii,
including the genetic sub-variant black bear (Ursus
americanus carlottae). Currently, there are 12 introduced
terrestrial vertebrates on Haida Gwaii, including two rats
(Rattus rattus and R. norvegicus) and the Sitka black-tailed
deer (Odocoileus hemionus), which have been the focus of
recent eradication efforts in Gwaii Haanas.
Although Gwaii Haanas is protected in legislation,
seabird habitat quality has continued to decline as a result
of the impacts of IAS such as rats, raccoons (Procyon
lotor vancouverensis) and Sitka black-tailed deer. Indeed,
introduced species have been identified as the main threat
to ecological integrity in Gwaii Haanas (AMB, 2018).
Impacts from deer browsing include simplification of the
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
494
up to meet the challenge, pp. 494–496. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Bellis, et al.: Invasive species initiatives in Gwaii Haanas
forest structure and lack of regeneration of culturally and
ecologically important species such as western redcedar
(Thuja plicata) and yellow cedar (Chamaecyparis
nootkatensis) (Pojar, 2008). Rats have had similarly
devastating impacts on seabirds. For example, some
islands that previously had murrelet colonies numbering
up to 8,000 breeding pairs, are now effectively at zero
(Rodway, et al., 1994).
HAIDA CULTURAL CONTEXT
Archaeological evidence of human habitation in Gwaii
Haanas goes back more than 12,000 years (Fedje, et al.,
2011). It is estimated that perhaps more than 30,000 Haida
lived in villages and seasonal camps across the archipelago.
Following European contact, diseases such as smallpox
and influenza reduced the Haida population to as low as
550 and the remaining people gathered in two villages.
These villages have since developed into the present-day
towns of Old Massett and Skidegate. The current Haida
population estimate is 5,000, and approximately half live
on Haida Gwaii.
The Council of the Haida Nation is mandated to
steward the lands and waters of Haida territory on behalf
of the Haida Nation, including the perpetuation of Haida
language, culture, art and traditional ways, for future
generations (Haida Nation, 2017). No treaty was ever
signed for Haida Gwaii and, in 2002, the Haida Nation
filed a legal case with the Supreme Court of Canada for
the title to Haida Gwaii. This case challenges the idea that
Haida Gwaii is owned by the Canadian government.
While this ownership dispute is resolved in the courts,
Gwaii Haanas continues to be managed by the AMB. This
is possible because the dispute over title to Gwaii Haanas is
explicitly laid out in the Gwaii Haanas Agreement (1993),
which also describes how the two parties will set aside
that disagreement in order to focus on shared objectives
concerning the care, protection and enjoyment of the
archipelago.
IAS PROJECTS IN GWAII HAANAS
Haida Gwaii was the first place in North America to
carry out a successful rat eradication, on Langara Island
in 1995, and remains the world’s largest bait station-based
eradication. The Council of the Haida Nation provided
direction, and several Haida community members worked
on this eradication, which used bait station transects
throughout the 32.7 km2 Island.
In Gwaii Haanas, rats were removed from four islands
through the SGin Xaana Sdiihltl'lxa (Night Birds Returning)
project (2009–2016). All were successful, and today three
islands remain rat free and one has been reinvaded. The
origin of the reinvasion is unclear but genetic testing results
show that it is not related to rats from the adjacent island
or those that were present prior to the eradication in 2011.
In 2017, a deer eradication project, Llgaay gwii sdiihlda
(Restoring Balance), began on six islands in Gwaii Haanas.
This project is currently in progress and aims to restore the
forest understorey community that has been decimated by
deer browsing.
With the deer population reduction, there is evidence
of recovery including culturally important plants such as:
● Ts’iihlinjaaw (devil’s club – Oplopanax horridus) –
medicine;
● SGiidllGuu (huckleberry – Vaccinium parvifolium) –
medicine, berries;
● Hldaan (blueberry – Vaccinium alaskaense and V.
ovalifolium) – wooden pegs, berries;
● Sk’idGan (salal – Gaultheria shallon) – berries.
A HAIDA PERSPECTIVE ON IAS
Invasive species on Haida Gwaii have a direct impact
on Haida culture, including:
● Loss of culturally significant plants and animals
including western redcedar, a species important for
carving, weaving and house building;
● Lack of opportunity to access medicinal and edible
plants;
● Loss of opportunities to pass on knowledge of
traditional harvesting teachings between generations;
and
● Loss of traditional food sources such as fruiting
plants, seabirds and seabird eggs.
These losses affect Haida citizens’ ability to exercise
their rights and practice their culture.
The primary objective of the Council of the Haida
Nation (CHN) is to achieve legal title to Haida Gwaii and
the surrounding waters. However, the CHN also works to
achieve conservation gains in Haida Gwaii. For example,
Land Use Orders and forest management based on
ecosystem management principles have been implemented
to protect culturally significant areas and to ensure Haida
values are maintained in areas where logging occurs. In
addition, a 1000-year cedar strategy is in development.
This strategy will ensure that there are large, monumental
cedars available in perpetuity for Haida Nation citizens to
utilise for pole or canoe carving, weaving or house building
projects. The CHN has also established new protected
areas and designed effectiveness monitoring programmes,
collecting data to support decision-making and research.
While the CHN has participated in many IAS-related
projects, collaborating with provincial and federal
governments, it is now working to develop a broader vision
and targets concerning IAS on Haida Gwaii. This involves
setting priorities and assessing current-day challenges. For
example, the invasive Sitka black-tailed deer has become
a significant food source for the Haida community and
its hides, hooves and antlers are now incorporated into
ceremonies. Therefore, initiatives need to balance the
impacts of IAS on culturally important species while also
considering the value these species have in the presentday culture. Generally, CHN-led IAS initiatives will focus
on species that impact cultural activities, with a goal of
managing IAS on Haida Gwaii in order to create healthy
ecosystems for future generations.
HAIDA PLANNING PRINCIPLES
The CHN applies several basic principles to all planning
initiatives, including invasive species management. These
principles are:
● Ts’uu (western redcedar – Thuja plicata) –
construction, carving;
Yahguudang – Respect
● Kayd (Sitka spruce – Chamaecyparis nootkatensis) –
roots, pitch, construction;
Respect for each other and all living things is rooted in
our culture. We take only what we need, we give thanks,
and we acknowledge those who behave accordingly.
● K’aang (hemlock – Tsuga heterophylla) – fish hooks,
food;
495
Island invasives: scaling up to meet the challenge. Ch 3B Strategy: Collaboration
‘Laa guu ga kanhllns – Responsibility
We accept the responsibility passed on by our ancestors
to manage and care for our sea and land. We will ensure
that our heritage is passed onto future generations.
Gina 'waadluxan gud ad kwaagid – Interconnectedness
Everything depends on everything else. The principle of
interconnectedness is fundamental to integrated planning
and management. This comprehensive approach considers
the relationships between species and habitats and
accounts for short-term, long-term and cumulative effects
of human activities on the environment. Interrelationships
are accounted for across spatial and temporal scales and
across agencies and jurisdictions.
Fedje, D., Mackie, Q., Lacourse, T. and McLaren, D. (2011). ‘Younger
Dryas environments and archaeology on the Northwest Coast of North
America’. Quaternary International 242: 452–462.
Gwaii Haanas Agreement. (1993). ‘Gwaii Haanas Agreement’. <www.
haidanation.ca/wp-content/uploads/2017/03/GwaiiHaanasAgreement.
pdf>. Accessed July 2017.
Gwaii Haanas Marine Agreement. (2010). ‘Gwaii Haanas Marine
Agreement’. <www.pc.gc.ca/en/pn-np/bc/gwaiihaanas/info/index/~/me
dia/662F1B36258E462DA46C2E8372B62E3A.ashx>. Accessed July
2017.
Haida Nation Constitution. (2014). ‘Constitution of the Haida Nation’.
<www.haidanation.ca/wp-content/uploads/2017/03/HN-ConstitutionRevised-Oct-2014_official-unsigned-copy.pdf>. Accessed July 2017.
Giid tlljuus – Balance
Harfenist, A. and Cober, A. (2003). Seabird Colonies Background
Report for Haida Gwaii/Queen Charlotte Islands Land Use Plan.
Report prepared for Ministry of Forests, Lands and Natural Resource
Operations, Queen Charlotte, BC. <www.for.gov.bc.ca/tasb/slrp/lrmp/
nanaimo/haidagwaii/docs/Seabird-Rpt-Low-Resolution.pdf>. Accessed
July 2017.
The world is as sharp as the edge of a knife. Balance
is needed in our interactions with the natural world. If we
aren't careful in everything we do, we can easily reach a
point of no return. Our practices and those of others must
be sustainable.
Pojar, J. (2008). ‘Changes in Vegetation of Haida Gwaii in Historical
Time’. In: A.J. Gaston, T. E. Golumbia, J-L. Martina, and S.T. Sharpe
(eds.) Lessons from the Islands: Introduced Species and What They Tell
Us about how Ecosystems Work. Proceedings from the Research Group
on Introduced Species 2002 Symposium, pp. 32–38. Queen Charlotte,
BC: Government of Canada.
Gina k’aadang.nga gii uu til k’anguudang – Seeking
wise counsel
Pejchar, L. and Mooney, H.A., (2009). ‘Invasive species, ecosystem
services and human well-being’. Trends in Ecology and Evolution
24(9): 497–504.
Our elders teach us about traditional ways and how to
work in harmony. Like the forests, the roots of our people
are intertwined. Together we consider new ideas and
information in keeping with our culture, values and laws.
Isda ad dii gii isda – Giving and receiving
Reciprocity is a respected practice in our culture,
essential in our interactions with each other and the natural
world. We continually give thanks to the natural world for
the gifts that we receive.
CONCLUSION
The ecological impacts of IAS eradications are well
studied and documented by the ecological research
community. Less-often considered is the importance of
eradications to cultural integrity and the continuity of
indigenous cultures such as the Haida Nation. The Gwaii
Haanas experience demonstrates how partnerships with
indigenous governments can broaden the scope of, and
support for, IAS projects and make them culturally as well
as ecologically meaningful.
ACKNOWLEDGEMENTS
The authors would like to say haaw’a (thank you) to the
islanders and international research community members
that have initiated and maintained the focus on biosecurity,
invasive species research and restoration on Haida Gwaii.
This includes but is not limited to Chris Gill of Coastal
Conservation, Jean-Louis Martin and others from the
RGIS group, the concerned and involved citizens of Haida
Gwaii and staff of Gwaii Haanas and the Council of the
Haida Nation.
REFERENCES
AMB. (2018). ‘Draft Gwaii Haanas Land-Sea-People Management Plan
for the National Park Reserve, National Marine Conservation Area
Reserve, and Haida Heritage Site’. <https://www.pc.gc.ca/en/pn-np/bc/
gwaiihaanas/info/public>.
Canadian Constitution Act (1982). ‘The Constitution Act, 1982, being
Schedule B to the Canada Act 1982 (UK)’. <http://canlii.ca/t/ldsx>.
496
Rodway, M.S., Lemon, M.J.F. and Kaiser, G.W. (1994). British Columbia
Seabird Colony Inventory: Report 6 – Major Colonies on the West Coast
of Graham Island. Canadian Wildlife Service, Pacific and Yukon Region
Technical Report Series 95. Ottawa, ON, Canada: The Queen’s Printer.
Simberloff, D, Martin, J.-L., Genovesi, P., Maris, V., Wardle, D.A.,
Aronson, J., Courchamp, F., Galil, B., Garcia-Berthou, E., Pascal, M.
and Pysek, P. (2013). ‘Impact of biological invasions: What’s what and
the way forward’. Trends in Ecology and Evolution 28(1): 58–66.
Venter, O., Brodeur, N.N., Nemiroff, L., Belland, B., Dolinsdk, I.J. and
Grant, J. (2006). ‘Threats to endangered species in Canada’. BioScience
56(11): 903–910.
Wilcove, D.S., Rothstein, D., DuBow, J., Phillips, A. and Losos, E.
(1998). ‘Quantifying threats to imperilled species in the United States’.
BioScience 48(8): 607–615.
D.C. Duffy and C. Martin
Duffy, D.C. and C. Martin. Cooperative Natural Resource and Invasive Species Management in Hawaiʽi
Cooperative natural resource and invasive species management
in Hawai'i
D.C. Duffy1 and C. Martin2
Pacific Cooperative Studies Unit, Department of Botany, University of Hawaiʻi at Mānoa, 3190 Maile Way, Honolulu
HI 96822, USA. <dduffy@hawaii.edu>. 2Pacific Cooperative Studies Unit/Coordinating Group on Alien Pest Species,
University of Hawaiʻi at Mānoa, 3190 Maile Way, Honolulu HI 96822, USA.
1
Abstract From the arrival of Polynesians before 1200 AD until Western contact in 1778, Hawaiian land use and resource
distribution were centred on the “ahupuaʻa” system of watershed-based management areas that made their inhabitants
nearly self-sufficient. Those units were grouped into larger regions, each called a “moku,” led by lower chiefs who
were in turn governed by the high chief or king of each island. While societal and environmental taboo or “kapu” were
enforced from above, day to day neighbourhood cooperation served to protect resources, produce food, and sustain up
to one million people before Western contact. Following the arrival of Europeans, land and resource management, and
governance based on native Hawaiian core beliefs, were replaced by a centralised Western market economy. Modern
land ownership, agency mandates and legal jurisdictions provide artificial walls that keep people from moving, but do
not constrain invasive species, nor are they effective for managing public trust resources such as water or native species.
Over time government and conservation organisations have come to view decentralised cooperation as a key to protecting
the 50% of Hawaiian terrestrial, native habitat that persists. Current cooperative efforts serve to protect watersheds via
watershed partnerships, and to detect and address invasive species through invasive species committees on each island.
These and other projects are facilitated by the University of Hawaii’s Pacific Cooperative Studies Unit, which plays a
unique role in coordinating the use of private and government grant dollars to hire and manage 460 conservation staff in
cooperative conservation projects state-wide.
Keywords: ahupuaʻa, invasive species, cooperative, watershed partnerships, invasive species committees
INTRODUCTION
The state of Hawaiʻi comprises eight large islands, seven
of them inhabited. The island of Oʻahu, home to nearly
one million of the state’s 1.4 million residents, is the seat
of the central state government (<http://files.hawaii.gov/
dbedt/economic/databook/db2016/db2016.pdf>). Each of
the islands has different land use and economic histories.
While most islands have a few major land owners, almost
all the islands are currently mosaics of federal, state,
county, and private properties, making it difficult to mount
effective responses to invasive alien species (IAS), as these
recognise neither property boundaries nor jurisdictions, or
to manage public trust resources such as native species,
habitats, and water resources (Ikuma, et al., 2002; Rago
& Sugano, 2015). This paper presents case studies that
explore how Hawaiʻi has responded to conservation
challenges through cooperative efforts by both institutions
and individuals. The list is illustrative, not exhaustive.
Natural history and the Hawaiian period
Hawaiʻi is one of the most isolated archipelagos on the
planet. From still-forming Hawaiʻi Island, the archipelago
progresses in age through the main islands, to the islets
and atolls of Papahānaumokuākea National Monument,
and the submerged Emperor Seamounts, representing 70
million years of passage over a tectonic hot spot (Heliker,
1989). Before the arrival of humans, new species became
established every 175,000–15,000,000 years (Ziegler,
2002). Isolation and subsequent adaptation to a wide
variety of ecological zones and habitats over millions
of years produced a stunning biodiversity vulnerable to
outside perturbations (Carlquist, 1974; Duffy & Vargas,
2017).
Polynesians settled in Hawaiʻi by 1200 AD (Kirch,
2011; Wilmshurst, et al., 2011). Initial populations were
small and the first Hawaiians subsisted as hunter-gathers
with limited agriculture. Even these initial actions had a
massive effect on biodiversity in a terrestrial ecosystem
that had not known mammalian predators. With increasing
human populations and the extinction of terrestrial protein
sources such as flightless birds and land crabs, agriculture
became more important, requiring communal investment
in infrastructure such as fish ponds and irrigation systems
(Kirch, 1985; Paulay & Starmer, 2011). Land was
often divided into mountain-to-sea pie-shaped wedges
(ahupua‘a) with larger units called moku on each island.
Although trading occurred, ahupuaʻa tended to be internally
balanced systems (Andrade, 2008). Ahupua‘a were
administered by konohiki, resource managers appointed
by the aliʻi (rulers) of large districts or entire islands
(Gonschor & Beamer, 2014). There was also a division
between the realm of man (wao kanaka), the agricultural
and community areas, and the realm of gods (wao akua),
the upper forests where entry was granted only to specially
trained individuals following strict protocols. While tenure
of the aliʻi was subject to the political winds of fortune, the
residents (maka‘ainana) of the ahupuaʻa were permanent.
Together with the konohiki, they made decisions about
use of local resources ranging from montane forests and
irrigated uplands down to coastal ponds and inshore waters
and these decisions were regulated by social/religious
strictures (kapu) (Mueller-Dombois, 2007).
The arrival of humans greatly increased the rate of
species’ arrivals, either deliberately for food or other
economic or cultural advantages, or as accidentals,
incidental to travel and commerce. The first settlers
traveling east from Polynesia brought about 30 plant
species and several animals, including Polynesian pigs (Sus
scrofa) and Pacific rats (Rattus exulans), some perhaps as
stowaways.
Despite the capacity of Hawaiian society to mobilise
large numbers of people at the island or moku level to
engage in major community efforts such as building heiau
(temples) and fishponds (Kirch, 1985), we have no direct
information on how pre-contact Hawaiians reacted to the
impacts of invasive species. For example, archaeological
evidence suggests that Pacific rats caused major changes
in lowland ecosystems by eating tremendous amounts of
seeds, damaging or killing plants, and preying on ground-
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 497–502. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
497
Island invasives: scaling up to meet the challenge. Ch 3B Strategy: Collaboration
nesting birds and other species (Athens, 2009). Kepelino
(1932: 86) reported oral traditions that rats were a major
problem for sweet potato (Ipomoea batatas) crops in
the lowlands. The aliʻi organised rat hunting contests,
suggesting opportunity, if not necessity (Athens, 2009;
Handy & Handy, 1972).
Post-European arrival
The arrival of Captain Cook in 1778 led to rapid
changes, including the introduction of diseases such as
smallpox to which Hawaiians had little resistance, and
the introduction of Western ideals and religion that muted
Hawaiian language, culture, and beliefs (Busnell, 1993).
Just seventy years later, the Great Mahele (land division) of
1848 placed two-thirds of the crown lands in private hands,
the majority non-Hawaiian, as most Hawaiians could not
conceive of a world where they needed to claim rights to
the land they had always lived on. This alienation of land
further weakened the traditional societal structure and the
kapu restrictions that controlled use of natural resources
(Chinen, 1958; LaCroix & Roumasset, 1990).
Land ownership and political power became
concentrated in the hands of the “Big Five” corporations
which were primarily involved in an export economy
centred on sugar production (Dorrance & Morgan,
2000). These shifts eventually led to the overthrow of
the Monarchy (Kame‘eleihiwa, 1992) and the resulting
“plantation system” came to dominate Hawaii’s social,
political and economic systems with a top-down political
structure centred on the island of Oʻahu.
Extensive deforestation occurred as lands were
converted to cane fields and streams were diverted for
irrigation. The drive to export goods also led to further
impacts on forests, with harvesting of sandalwood
(Santalum paniculatum) and pulu, the fibre from native
tree ferns (Cibotium menziesii), in the wao akua areas
previously regarded as sacred and off-limits (Cuddihy &
Stone, 1990). Land devoted to sugar production peaked
in the 1940s and economically viable production ceased
by 2015 (Dorrance & Morgan, 2000). The end of sugar
as a crop left large portions of lower-elevation landscapes
fallow or being converted into housing tracts and tourist
developments on the coasts. Sugar has been replaced by
tourism and the military as drivers of the economy, but the
state retains its Oʻahu-centric political orientation left over
from plantation days (Kalapa, 1992).
Polynesian pigs were initially barnyard animals, but
after 1778 they mixed with introduced European strains
and soon found their way into upper-elevation forests.
Captains Cook and Vancouver left cows (Bos taurus),
sheep (Ovis aries) and goats (Capra hircus) as gifts for
Hawaiians (Tomich, 1986). Cows were placed under royal
protection by King Kamehameha after their introduction
in 1794. Protection lasted until 1830 by which point the
population had greatly expanded and caused significant
deforestation (Tomich, 1986). Cats (Felis catus) were
brought to the islands both as novelties and to curb rodent
populations (Duffy & Capece, 2012). With the rise of the
sugar plantations, species such as cane toads (Rhinella
marina), mongoose (Urva auropunctata) and parasitic
wasps were introduced to reduce rats and insects that
feed on the cane. While their effectiveness is debatable,
their negative consequences are not (Doty, 1945; Peck,
et al., 2008). Accidental introductions included black and
Norway rats (R. rattus and R. norvegicus) (Tomich, 1986),
and mosquitoes (Culicidae) that were stowaways on ships.
Earthworms (Lumbricidae) and ants (Formicidae) were
absent from pre-contact Hawaiʻi, but probably arrived
in soil and plants, as did numerous other invertebrates
with largely undocumented but likely enormous impacts
(Gillespie & Reimer, 1993).
498
More than 100 plant species arrived in the 60 years
following Cook’s arrival (Nagata, 1985). More recently
Loope & Kraus (2009) reported that, during 1995–2003,
89 species per year became established. To date, over
10,000 plant species have been introduced for cultivation
in the islands (Imada, et al., 2005). Birds and mammals
were introduced for hunting and for human entertainment
(Walker, 1967; Long, 1971). Accidental introduction and
deliberate smuggling of herptiles have been a problem,
with some becoming invasive (McKeown, 1996; Kraus
& Cravalho, 2001; Kraus, 2009). Aquatic species have
arrived as deliberate introductions for fisheries, through
the aquaculture/pet/aquarium trade, in ballast water and as
biofouling (Eldredge & Smith, 2001; Brasher, et al., 2006;
Carlton & Eldredge, 2009). Finally, pathogens have been
a continuing problem for both humans and the rest of the
biota since European contact (e.g. Wilbar, 1947; Warner,
1968; Bushnell, 1993).
Responding to alien invasive species
The Kingdom of Hawaiʻi enacted the first biosecurity
measure for the islands, banning the import of coffee
beans to prevent alien disease from affecting the islands’
own crops (Holt, 1996). Later, import of sugar cane and
other grasses was restricted because they might bring
in new diseases and pests of the dominant agricultural
crop (Territory of Hawaiʻi, 1941). By 1975, deliberate
introductions of organisms had to be approved by the
Department of Agriculture. This rule remains in place;
however, the vast majority of plants and plant parts are
still not effectively restricted from entry (Loope & Kraus,
2009).
King Kalākaua (Kalākaua 1876) began a programme
of fencing to exclude feral ungulates from watersheds to
protect the water supply. By the turn of the century and the
fall of the monarchy, the territorial legislature recognised
the continued impact on watershed forests by feral animals
and the unregulated harvest of forest products. This led
to a massive re-planting of fast-growing non-native trees
with the hope that this would sustain watershed function.
However, several of the trees became invasive (<https://
www.nature.org/media/hawaii/the-last-stand-hawaiianforest.pdf>; Cox, 1992; Woodcock, 2003; Kaiser, 2014).
During the territorial period and following statehood,
legislation created several state governmental agencies
to address alien invasive species (AIS) and to protect or
manage natural resources. With various name changes
over time, the Department of Health dealt with disease
vectors such as rats and mosquitoes, while the Department
of Land and Natural Resources dealt with establishing
forest reserves, managing aquatic and hunting resources,
and reducing the impact of invasive species on stateowned watersheds and native ecosystems. The Department
of Agriculture dealt with invasive species of importance
to agriculture, and the importation of agricultural goods
and species. No agency was or is responsible for a
holistic assessment or response to the continued arrival of
additional invasive species (Rago & Sugano, 2015).
In the decades since these laws were created, there
have been major changes in Hawaii’s economic drivers,
agricultural crops, frequency and quantities of imports,
and the rise of air cargo (2.72%/year from 1990 to 2016:
DBEDT, 2017), with a resulting increase in magnitude of
risk from invasive species. Unfortunately changes to the
laws and policies that reduce or address invasive species
risks or impacts have been piecemeal and insufficient, with
gaps within or between agency mandates (Miller & Holt,
1992; Ikuma, et al., 2002; Loope & Kraus, 2009; Rago &
Sugano, 2015).
Duffy & Martin: Cooperative species management in Hawai’i
At the federal level, the National Park Service and
the National Wildlife Refuge System, both under the
Department of Interior, have broad mandates to manage
invasive species and protect natural resources and habitats
within their holdings, but their authority and actions
historically have been confined within their property
boundaries. The Department of Defense (DOD), another
large landowner, did not focus effort or attention on
mitigating impacts on the natural environment or protecting
natural resources unless they interfered with military
activities, as did a dengue outbreak during World War II
when martial law allowed the agency to ignore property
rights to deal with the outbreak (Wilbar, 1947). More
recently, DOD has become more active and pre-emptive,
in part because of federal laws such as the U.S. Endangered
Species Act of 1973 (ESA) and the National Environmental
Policy Act of 1969 (NEPA), where actions that may have
an impact on Federally listed endangered species and noncompliance might restrict the military mission.
For prevention of new alien invasive species, among
other mandates, the Department of Homeland Security
Customs and Border Protection is responsible for
regulating the importation of goods and conveyances from
foreign sources into the U.S., while the U.S. Department
of Agriculture focuses on foreign and domestic
agricultural imports that may carry pests and diseases.
The U.S. Department of Interior (U.S. Fish and Wildlife
Service) uses the Lacey Act and ESA to reduce the risk
of invasive species being imported into Hawaii in specific
circumstances (e.g. the Injurious Wildlife Provisions of the
Lacey Act).
With the rise of rapid world trade, Hawaii’s borders
have become increasingly permeable to invasive species
(Loope & Kraus, 2009). Responses to such species
are difficult, as potentially invasive species are rarely
discovered on a single property where the landowner has
the knowledge, skills, interest, and funding to address the
species before it spreads. Landowners can also be hesitant
to allow government officials onto their properties to search
for or control invasive alien species, so government often
fails to detect invasive alien species before they spread
(Kraus & Duffy, 2010). Action against a new invasive
alien species largely depends on its location and whether
the species is perceived as falling within the mandate of a
particular agency. Further, the bureaucratic process for the
addition of new species to official lists mandating control
does not keep up with the pace of arrivals (Penniman,
et al., 2011). Finally, cooperation between state, federal
and county authorities has at times been limited and
intermittent (Warren, 2006). In consequence in the last
two decades, cooperative, often informal approaches have
increasingly supplemented top-down formal efforts. We
present four such cooperative models that range from the
intergovernmental, to agreements between landowners, to
groups open to anyone sharing a common objective.
threats actually reached them, making parks legal but not
ecologically sustainable islands. Following initial surveys
documenting the native flora and fauna and threats to these
species, including from non-native species, UH scientists
built a small test ungulate exclosure, which produced rapid
recovery of native plants. NPS engaged the CPSU to build
more fences, removing ungulates, and monitoring the
subsequent recovery. Based in part on this work, fencing as
a management tool was adopted by NPS, other federal and
state agencies, and non-profit organisations. In response to
increasing recognition of threats to Hawaii’s biodiversity
and perceived gaps caused by narrow agency mandates
and jurisdictions, CPSU morphed into what is now the
Pacific Cooperative Studies Unit (PCSU) working with a
range of state and federal agencies, as well as non-profits
and private companies and individuals. PCSU provides
research, resource management and outreach expertise via
collaborative projects, while also increasing employment
opportunities in conservation. Over the last 20 years it
has grown from 150 employees to more than 450 (Fig. 1),
mentoring and staffing a range of organisations dealing
with invasive and endangered species (see below).
Coordinating Group on Alien Pest Species (CGAPS)
A second key development was the formation of the
Honolulu-based interagency Coordinating Group on Alien
Pest Species (CGAPS) in 1995. In 1992, The Nature
Conservancy of Hawaii (TNCH) and the Natural Resources
Defence Council published a report on Hawaii’s biosecurity
measures and the gaps that would likely lead to the arrival
and establishment of major new pests such as brown tree
snakes (Boiga irregularis) and red fire ants (Solenopsis
invicta) (Miller & Holt, 1992). The report concluded that,
although there were funding and policy gaps, the most
serious problem was a lack of interagency, and sometimes
intra-agency, communication and cooperation, and that
many such gaps could be addressed through a coordinated
effort (Miller & Holt, 1992; Holt, 1996). These reports,
and other events, led to the crafting of the Hawaii Alien
Species Action Plan in 1993–94 with the help of more than
80 agency and NGO leaders under a steering committee
that morphed into CGAPS (Nakatani & Wilson, 1995).
Today, CGAPS continues to facilitate interagency and
NGO communication and cooperation through quarterly
meetings, and its steering committee plans and conducts
collaborative projects to catalyse action on invasive
species. CGAPS was originally administered by the Hawaii
Department of Agriculture with staff time contributed
by TNCH and agencies, but it now has a rotating chair
structure, and staff and logistics supported by PCSU.
One of its recent projects was the crafting of a Plant
Health Emergency Response Plan, which laid out how the
US Department of Agriculture and Hawaiʻi Department of
Agriculture could engage other federal, state and county
Pacific Cooperative Studies Unit (PCSU)
One of the earliest natural resource management
organisations in Hawaii to extend beyond top-down
management was the Cooperative National Parks
Studies Unit (CPSU), which formed in 1973 through an
agreement between the National Park Service (NPS)
and the University of Hawaii. CPSU initially provided
collaborative research and technical support for Hawaiʻi
Volcanoes and Haleakala National Parks. Following
the passage of three key federal laws: the National
Environmental Policy Act, the Endangered Species Act,
and the General Authorities Act of 1970 (<https://www.nps.
gov/parkhistory/hisnps/NPSHistory/timeline_annotated.
htm>), NPS lacked the internal capacity to conduct
the research necessary to respond to these mandates. It
also lacked a mandate to protect national parks before
Fig. 1 Growth of the Pacific Cooperative Studies Unit.
499
Island invasives: scaling up to meet the challenge. Ch 3B Strategy: Collaboration
agencies, and non-governmental organisations if a serious
plant pest were to arrive, requiring a response beyond
what the two federal agencies could provide (Loope &
Shluker Ryon, 2013). To test the plan, CGAPS conducted
a discussion-based “tabletop” exercise in November
2013, using the then-fictitious discovery of the coconut
rhinoceros beetle (Oryctes rhinoceros, CRB). The tabletop
discussion included more than 40 participants and resulted
in a report that outlined legal and procedural questions
that arose (Coordinating Group on Alien Species, 2013).
Coincidentally, a month later, in December 2013, CRB
were detected in a trap at Joint Base Pearl Harbor-Hickam,
triggering the first use of the Plant Health Emergency
Response Plan. Although eradication has yet to be
achieved, the emergency response has been successful in
many ways, from the initial staffing of the response coled by the two responsible agencies, and supplemented by
multiple federal, state, and non-government partners, to the
containment of the beetles to West Oʻahu.
Invasive Species Committees
Invasive Species Committees (ISCs) are voluntary
partnerships on each island that address incipient (new)
invasive plants and animals on an island-wide basis. The
five invasive species committees (Kauaʻi Invasive Species
Committee, Oʻahu Invasive Species Committee, Maui
Invasive Species Committee, Molokaʻi Invasive Species
Committee, and Big Island Invasive Species Committee),
represent perhaps the best collaborative efforts in the
islands. Their steering committees are essentially selfrecruited, made up of interested private individuals and
groups as well as representatives of county, state and federal
agencies. Together, each island’s steering committee
provides strategic direction to a paid staff and field crew
for island-wide work on early detection and control or
eradication of high-risk invasive species. A critical function
of ISCs is to obtain right of entry to private lands through
education and negotiation. The logistic, fiscal and staffing
aspects of each ISC are handled by the Pacific Cooperative
Studies Unit.
The first committee sprang from a pioneer effort on
Maui Island. A melastome tree (Miconia calvescens) had
been identified on Tahiti as a major threat to intact native
forests (Meyer, 1996). Biologists returning from a visit to
Tahiti recognised that the species occurred on Maui and
might represent a similar local threat to Hawaiʻi (Gagné, et
al., 1992). This led to the formation of an ad hoc Melastome
Action Committee in 1991 to address the problem on Maui
(Conant, et al., 1997; Medeiros, et al., 1997). The effort
subsequently expanded to Big Island (Tavares, 1998).
In recognition that there might be additional IAS threats
(Miller & Holt, 1992), the MAC expanded to other species,
and in 1997 the Maui Invasive Species Committee was
formed, soon followed by Oʻahu, Kauaʻi, Big Island and
Molokaʻi committees (Martin, 2003).
The ISCs focus on early detection and rapid response
leading to eradication of incipient invasive species, and
they also conduct outreach and education to help the public
reduce the impacts of established species. All species are
chosen for their local, not state, importance, based on
evaluation criteria that include risk to an island’s economy
or ecology, and feasibility of control or eradication
(Penniman, et al., 2011). The number of staff in each ISC
varies, but generally each has a field team, an outreach
specialist, a GIS/data specialist, and an overall manager
who is responsible for government relations, obtaining
funding, and working with its steering committee and
PCSU (Krauss & Duffy, 2010). Since formation, the
ISC managers have worked together to develop standard
methods, coordinate funding and reports, and even share
field crews when advantageous. During the 2008 recession,
500
they redistributed funding to keep all the ISCs staffed and
active. Being local to each island, these committees enjoy
strong county and legislative support, but funding remains
a persistent problem as new invasive species continue
to arrive while many of the old ones persist. As of 2010,
27 populations of emerging invasives had been removed
by the ISCs, but efforts for others are likely to be drawn
out because of reinvasions, persistent seed-banks, or the
continued discovery of isolated individuals (Kraus & Duffy,
2010; Penniman, et al., 2011). In addition, the ISCs have
worked with the Hawaii Ant Lab (another project of PCSU)
and HDOA to survey for and control incipient populations
of species such as little fire ants (Wasmannia auropunctata)
and coqui frogs (Eleutherodactylus coqui) on islands
where they are not yet established. These early detection
and rapid response functions have resulted in dozens of
local eradications of these pests before they could establish
populations.
Watershed Partnerships
Isolated oceanic islands like the Hawaiian archipelago
have limited freshwater supplies. Native forests in Hawaiʻi
retain water better than do island forests dominated by
introduced species (Giambelluca, et al., 2009; Kagawa, et
al., 2009; Cavaleri, et al., 2014), so protection of watersheds
is a prudent investment toward the persistence of human
populations in the islands, as well as for the maintenance
of the archipelago’s unique biota.
Maui has been the incubator of a number of innovations
in Hawaii and so it is not surprising that the first joint effort
to manage and improve watersheds across ownerships, the
East Maui Watershed Partnership (EMWP), was established
in 1991 through the efforts of The Nature Conservancy of
Hawaiʻi and the state Department of Land and Natural
Resources (Loope & Reeser, 2001). There are now ten
watershed partnerships or associations on five islands,
covering 2.2 million acres and involving 75 land-owning
partners, ranging from state and federal agencies to NGOs
to private companies and individuals (<http://hawp.org/
partnerships/>) (Fig. 2). These partners may have differing
objectives in land management so watershed partnerships
focus on those they hold in common, rather than imposing
the agendas of a minority of partners (cf. Ostrom, 1990).
Like CGAPS and the ISCs, most of the watershed
partnerships are informal, with the landowners and
agencies functioning as steering committees to determine
objectives, and with a manager and staff to address
the objectives. PCSU also provides the structure and
administrative capacity for most of these partnerships.
Fig. 2 Extent of watershed partnerships in the Hawaiian
archipelago in 2015 (<http://hawp.org/wp-content/
uploads/2011/12/WP_2015.png>).
Duffy & Martin: Cooperative species management in Hawai’i
The main task of the partnerships is the long-term
protection of forests in watershed recharge areas, by
removing alien plants and animals directly or through the
installation and upkeep of hundreds of miles of protective
fencing to exclude ungulates from sensitive areas.
Watershed partnerships can do this at the landscape level
without the constraints of political or property boundaries.
Indirectly, by protecting native habitat across wide areas,
the watershed partnerships have also become critical for
the conservation of endangered species. Working with
the ISCs, the partnerships are also key to locating and
dispatching newly arrived potentially invasive species.
DISCUSSION
Before contact with Western culture and during the
monarchy, environmental rules in Hawaiʻi were mandated
at the island or moku level. The introduction of Western
concepts such as private property and land ownership,
coupled with the abolition of the kapu system and transition
of the monarchs to Christianity, resulted in major changes
in the management of natural resources. Major increases
in resource extraction and land clearance were permitted
and abetted by the republic and territorial governments
controlled by the sugar companies. After statehood,
strong federal laws dealing with pollution and wetlands
functioned through command and control enforcement.
Most recently there has been a recognition that such topdown approaches are less effective for non-point problems
such as pollution, habitat destruction and managing
endangered and invasive species outside government
lands (Lubell, et al., 2002). Wider involvement is needed
to ensure buy-in by “stake holders” who must be part of
solutions (John, 1994). Some of these more participatory
approaches are mandated by U.S. law, such as interagency
consultation over endangered species, habitat conservation
plans for endangered species, and public comments
on federal government actions under the National
Environmental Policy Act. Within government, ad hoc
cooperation between agency partners through entities such
as CGAPS and PCSU has often proven more nimble than
statutory constructs. Outside formal government, limitedaccess partnerships of landowners such as watersheds
partnerships and invasive species committees open to all
have proven highly responsive to local conditions.
Overall, partnerships can yield multiple advantages.
They can help bring together resources such as outside
expertise (e.g. CPSU/PCSU) or undertake landscape-scale
management by pooling resources and reducing artificial
boundaries (WPs and ISCs), or they can see “the big
picture” with collaborators working together to identify
and address issues (CGAPS).
Partnerships are not automatic panaceas. Partnerships
require flexibility and trust, and a recognition that they
may not be appropriate for every problem. Partnerships
also require a roughly equal distribution of power and
resources. If one partner is dominant, then the partnership
becomes merely an advisory group or rubber stamp.
Partnerships require a working consensus on approaches.
Islands have a limited spectrum of economic activity
compared to the mainland so people are more likely to
share a common perspective and recognition of the value
of the local indigenous environment than may occur at the
continental scale. However, even in Hawaiʻi, issues such
as air-dropped rodent control agents, genetically-modified
organisms, and biocontrol may remain too controversial
for partnership approaches. Feral cat management on the
islands of Oahu and Kauai is a particularly contentious
issue, but there is hope for an emerging consensus. Twenty
years ago, fencing was similarly controversial but it has
now become an accepted approach to land protection.
In terms of logistics, partnerships can falter without
a lead person and staff whose jobs are to move the
partnership’s goals forward. Partnerships appear to be
less effective when they are burdened with managing such
staff, as administrative concerns divert time and energy
away from the “big picture” and away from consensus
building around common objectives (Lubell et al., 2002).
In Hawaiʻi, the Pacific Cooperative Studies Unit has
frequently supplied the stable logistics and organisational
underpinning for such partnerships, providing professional
staffing, and the ability to handle financial, legal and
regulatory requirements. This appears to provide a
flexibility not always present in government agencies
where funding can vary from year to year and priorities
change from one political administration to the next.
Although partnerships and cooperative efforts can
be powerful tools for conservation and we can generate
general rules about what works and what doesn’t, all such
efforts are local, dependent on the local economy, local
politics and the local environment. It is important that
we better document what has worked and what hasn’t for
Hawaii, both for other areas that might wish to explore the
use of cooperation in conservation, and as anthropogenic
climate change brings new challenges to islands.
ACKNOWLEDGEMENTS
Among many others, we thank J. Canfield, B. Harry, A.
Holt, L. Loope, T. Penniman, and C. Smith for discussions
over the years on the challenges and opportunities of
collaborative efforts in Hawaii. The views expressed here
are our own and do not necessarily represent those of any
institution or agency. We dedicate this paper to Lloyd
Loope who was seer, savant and sensei to our community.
REFERENCES
Andrade, C. (2008). Haena: Through the Eyes of the Ancestors. Honolulu:
University of Hawaiʻi Press.
Athens, J.S. (2009). ‘Rattus exulans and the catastrophic disappearance
of Hawai’i’s native lowland forest’. Biological Invasions 11: 1489–
1581.
Brasher, A.M., Luton, C.D., Goodbred, S.L. and Wolff, R.H. (2006).
‘Invasion patterns along elevation and urbanization gradients in
Hawaiian streams’. Transactions of the American Fisheries Society 135:
1109–1129.
Bushnell, O.A. (1993). The Gifts of Civilization: Germs and Genocide in
Hawaii. Honolulu: University of Hawaii Press.
Carlton, J.T. and Eldredge, L.G. (2009). Marine Bioinvasions of Hawaii.
Bishop Museum Bulletins in Cultural and Environmental Studies.
Honolulu: Bishop Museum Press.
Cavaleri, M.A., Ostertag, R., Cordell, S. and Sack, L. (2014). ‘Native
trees show conservative water use relative to invasive trees: results
from a removal experiment in a Hawaiian wet forest’. Conservation
Physiology 2: doi:10.1093/conphys/cou016.
Chinen, J.J. (1958). The Great Mahele: Hawaii's Land Division of 1848.
Honolulu: University of Hawaii Press.
Conant, P., Medeiros, A.C. and Loope, L.L. (1997). ‘A Multi-agency
Containment Program for Miconia (Miconia calvescens), an Invasive
Tree in Hawaiian Rain Forests’. In J.O. Luken and J.W. Thieret (eds.)
Assessment and Management of Invasive Plants, pp. 249–254. Berlin:
Springer-Verlag.
Cox, T. (1992). ‘The Birth of Hawaiian Forestry: The Web of Influences’.
In: J. Dargavel and R. Tucker (eds.) Changing Pacific Forests:
Historical Perspectives on the Forest Economy of the Pacific Basin. pp.
116–25. Durham, NC: Forest History Society.
Coordinating Group on Alien Species (2013). After Action Report/
Improvement Plan. <http://www.cgaps.org/wp-content/uploads/AARIP-CGAPS-FINAL-110813.pdf>.
Cuddihy, L.W. and Stone, C.P. (1990). Alteration of Native Hawaiian
Vegetation – Effects of Humans, their Activities and Introductions.
Honolulu: Cooperative National Park Resources Studies Unit,
University of Hawaii at Manoa.
DBEDT (Department of Business, Economic Development & Tourism).
(2017). Air Cargo in Hawaii’s Economy: Developments in Hawaii’s
Air Cargo Industry 1990–2016. Honolulu: Department of Business,
Economic Development & Tourism.
Dorrance, W.H. and Morgan, F. (2000). Sugar Islands: The 165-Year
Story of Sugar in Hawaiʻi, Honolulu: Mutual Publishing.
501
Island invasives: scaling up to meet the challenge. Ch 3B Strategy: Collaboration
Doty, R.E. (1945). ‘Rat control on Hawaiian sugar cane plantations’.
Hawaiian Planter’s Record 49: 71–239.
Duffy, D.C and Capece, P. (2012). ‘Biology and impacts of Pacific island
invasive species. 7. The domestic cat (Felis catus)’. Pacific Science 66:
173–212.
Duffy, D.C. and Vargas, H. (2017). ‘From the vagile to the sedentary:
disease implications and new host relationships on islands’. Chapter
5. In: P.G. Parker (ed.) Disease Ecology. Galapagos Birds and their
Parasites, pp. 113–138. New York: Springer.
Eldredge, L.G and Smith, C.M. (2001). Guidebook of Introduced Marine
Species in Hawaii. Bishop Museum Technical Report 21. Honolulu:
Bishop Museum.
Gagné, B.H., Loope, L.L. Medeiros, A.C. and Anderson, S.J. (1992).
‘Miconia calvescens: A threat to native forests in the Hawaiian Islands’
(Abstract). Pacific Science 46: 390–391.
Giambelluca, T.W., Martin, R.E., Asner, G.P., Huang, M., Mudd, R.G.,
Nullet, M.A., DeLay, J.K and Foote, D. (2009). ‘Evapotranspiration
and energy balance of native wet montane cloud forest in
Hawai‘i’. Agricultural and Forest Meteorology 149: 230–243.
Gillespie, R.G. and Reimer, N.J. (1993). ‘The effect of alien predatory ants
(Hymenoptera: Formicidae) on Hawaiian endemic spiders (Araneae:
Tetragnathidae)’. Pacific Science 47: 21–33.
Gonschor, L. and Beamer, K. (2014). ‘Toward an inventory of ahupua‘a
in the Hawaiian kingdom: A survey of nineteenth- and early twentiethcentury cartographic and archival records of the island of Hawai‘i’.
Hawaiian Journal of History 48: 53–87.
Handy, E.S.C. and Handy, E.G. (1972). ‘Native planters in old Hawaii:
Their life, lore and environment’. Bernice P. Bishop Museum Bulletin
233: 1–641.
Heliker, C. (1989). ‘Volcanic origins of the Hawaiian Islands.’ In:
C.P. Stone and D.B. Stone (eds.) Conservation Biology in Hawaii. pp.
11–15. Honolulu: University of Hawaii Press.
Holt, A. (1996). ‘An Alliance of Biodiversity, Agriculture, Health, and
Business Interests for Improved Alien Species Management in Hawaii.’
In: O.T. Sandlund, P.J. Schei, and A. Viken (eds.) Proceedings of the
Norway/UN Conference on Alien Species, pp. 155–160. Trondheim:
Directorate for Nature Management and Norwegian Institute for Nature
Research.
Ikuma, E.K., Sugano, D. and Mardfin, J.K. (2002) Filling the Gaps in
the Fight Against Invasive Species. Honolulu: Legislative Reference
Bureau. <http://lrbhawaii.org/reports/legrpts/lrb/rpts02/gaps.pdf>.
Imada, C.T., Staples, G.W. and Herbst, D.R. (2005). Annotated Checklist
of Cultivated Plants of Hawai‘i. <http://www2.bishopmuseum.org/
HBS/botany/cultivatedplants/>.
John, D. (1994). Civic Environmentalism. Washington, D.C.:
Congressional Quarterly Press.
Kagawa, A, Sack, L., Duarte, K. and James, S. (2009). ‘Hawaiian native
forest conserves water relative to timber plantation: Species and stand
traits influence water use’. Ecological Applications 19: 1429–1443.
Kaiser, B.A. (2014). ‘Watershed conservation in the long run’. Ecosystems
17: 698–719.
Kalākaua, D. (1876). An Act for the Protection and Preservation of Woods
and Forests. Honolulu: The Hawaiian Kingdom.
Kalapa, L.L. (1992). ‘Why is it that the State Seems to “Bully” the
Counties?’ In: R.W. Ross (ed.) The Price of Paradise. Honolulu: Mutual
Publishing.
Kame‘eleihiwa, L. (1992). Native Land and Foreign Desires. Honolulu:
Bishop Museum Press.
Kepelino, K. (1932). M.W. Beckwith (ed.) Kepelino’s Traditions of
Hawaii. Bishop Museum Bulletin 95. Honolulu: Bishop Museum Press.
Kirch, P.V. (1985). Feathered Gods and Fishhooks: An Introduction to
Hawaiian Archaeology and Prehistory. Honolulu: University of Hawaii
Press.
Kirch, P.V. (2011). ‘When did the Polynesians settle Hawai‘i? A review
of 150 years of scholarly inquiry and a tentative answer’. Hawaiian
Archaeology 12: 3–26.
Kraus, F. and Cravalho, D. (2001). ‘The risk to Hawaiʻi from snakes’.
Pacific Science 55: 409–417.
Kraus, F. (2009). Alien Reptiles and Amphibians: A Scientific Compendium
and Analysis. Invading Nature – Springer Series in Invasion Ecology,
Vol 4. Dordrecht, Netherlands: Springer.
Kraus, F. and Duffy, D.C. (2010). ‘A successful model from Hawaii for
rapid response to invasive species’. Journal for Nature Conservation
18: 135–141
LaCroix, S.J. and Roumasset, J. (1990). ‘The evolution of private
property in nineteenth-century Hawaii’. Journal of Economic History
50: 829–852.
Long, J.L. (1971). Introduced Birds of the World: The Worldwide History,
Distribution, and Influence of Birds Introduced to New Environments.
New York: Universe Books.
Loope, L. and Reeser, D. (2001). ‘Crossing Boundaries at Haleakala:
Addressing Invasive Species through Partnerships’. In: D. Harmon
(ed.) Crossing Boundaries in Park Management: Proceedings of the
11th Conference on Research and Resource Management in Parks and
Public Lands, pp. 29-34. Hancock, Michigan: George Wright Society.
502
Loope, L. and Kraus, F. (2009). ‘Preventing Establishment and Spread of
Invasive Species in Hawaii: Current Status and Needs for the Future’. In:
T.K. Pratt, C.T. Atkinson, P.C. Banko, J.D. Jacobi and B.L. Woodworth
(eds.) Conservation Biology of Hawaiian Forest Birds: Implications for
Island Birds, pp. 359–380. New Haven: Yale University Press.
Loope, L. and Shluker Ryon, A.D. (2013). State of Hawai’i Collaborative
Plant Health Emergency Management Plan (PHERP) and Reference
Manual.
<http://www.cgaps.org/wp-content/uploads/Hawaii-PlantHealth-Emergency-Response-Plan-Final-12-30-2013.pdf>.
Lubell, M., Schneider, M, Scholz, J.T. and Mete, M. (2002). ‘Watershed
partnerships and the emergence of collective action institutions’.
American Journal of Political Science 46: 148–163.
Martin, C. (2003). Island-based Partnerships and Statewide Coordination
to Protect Hawaii. <http://www.cgaps.org/wp-content/uploads/2002Island-Based-and-Statewide-Report.pdf>.
McKeown, S. (1996). A Field Guide to Reptiles and Amphibians in the
Hawaiian Islands, Los Osos, California: Diamond Head Publishing.
Medeiros, A.C., Loope, L.L., Conant, P. and McElvaney, S. (1997).
‘Status, Ecology, and Management of the Invasive Tree Miconia
calvescens DC (Melastomataceae) in the Hawaiian Islands’. In: N.L.
Evenhuis and S.E. Miller (eds.) Records of the Hawaii Biological
Survey for 1996. Bishop Museum Occasional Papers No. 48, pp. 23–36.
Honolulu: Bernice Pahuahi Bishop Museum.
Meyer, J.-Y. (1996). ‘Status of Miconia calvescens (Melastomataceae),
a dominant invasive tree in the Society Islands (French Polynesia)’.
Pacific Science 50: 66–76.
Miller, S. and Holt, A. (1992), The Alien Pest Species Invasion in Hawaii:
Background Study and Recommendations for Interagency Planning.
Honolulu: The Nature Conservancy of Hawaii and the Natural
Resources Defense Council.
Mueller-Dombois, D. (2007). ‘The Hawaiian Ahupua‘a Land Use System:
Its Biological Resource Zones and the Challenge for Silvicultural
Restoration’. In: N.L. Evenhuis and J.M. Fitzsimons (eds.) Biology of
Hawaiian Streams and Estuaries. Bishop Museum Bulletin in Cultural
and Environmental Studies 3, pp. 23–33. Honolulu: Bernice Pahuahi
Bishop Museum.
Nagata, K.M. (1985). ‘Early plant introductions in Hawaii’.
Hawaiian Journal of History 19: 35–61.
Nakatani, J. and Wilson, M. (1995). Hawaii Alien Species Action Plan.
<http://www.cgaps.org/wp-content/uploads/CGAPS-formation-toimplement-Hawaii-ASAP-1995.pdf>.
Ostrom, E. (1990). The Logic of Collective Action. New York: Cambridge
University Press.
Paulay, G. and Starmer, J. (2011) ‘Evolution, insular restriction, and
extinction of oceanic land crabs, exemplified by the loss of an endemic
Geograpsus in the Hawaiian islands’. PLOS ONE 6(5): e19916.
Peck, R.W., Banko, P.C., Schwarzfeld, M., Euaparadorn, M. and Brinck,
K.W. (2008). ‘Alien dominance of the parasitoid wasp community
along an elevation gradient on Hawaiʻi Island’. Biological Invasions
10: 1441–1455.
Penniman, T.M., Buchanan, L. and Loope, L.L. (2011). ‘Recent plant
eradications on the islands of Maui County, Hawaiʽi’. In: C.R. Veitch,
M.N. Clout and D.R. Towns (eds.) Island invasives: eradication and
management, pp. 325–331. Occassional Paper SSC no. 42. Gland,
Switzerland: IUCN and Auckland, New Zealand: CBB.
Rago, F. and Sugano, D. (2015). Can’t see the forest for the (albizia) trees:
an invasive species update. Honolulu: Legislative Reference Bureau.
Stone, C.P. and Stone, D.B. (eds.) (1989). Conservation Biology in
Hawaii. Honolulu: University of Hawaii Press.
Tavares, K. (1998). ‘Big Island Melastome Action Committee: Miconia
calvescens control program overview’. Proceedings of the First
Regional Conference on Miconia Control (August 26–29, 1997): 52–64.
Territory of Hawai’i (1941). Concerning the Introduction of Sugar Cane,
Bamboo and Other Grasses or Parts Thereof into Hawaii from Any
Other Part of the United States. Effective 12/12/41. Honolulu: Division
of Entomology, Board of Commissioners of Agriculture and Forestry.
Tomich, P.Q. (1986). Mammals in Hawaiʻi. Honolulu: Bishop Museum
Press.
Walker, R.L. (1967). ‘A brief history of exotic game bird and mammal
introductions into Hawaii with a look to the future’. ʻElepaio 28: 39–43.
Warner, R.E. (1968). ‘The role of introduced diseases in the extinction of
the endemic Hawaiian avifauna’. Condor 70: 101–120.
Warren, P. (2006). Biosecurity Systems of Hawaiʻi.: An Evaluation
and Recommendations for Reform. Honolulu: Hawaiʻi Conservation
Alliance.
Wilbar, C.L. (1947). ‘Control of dengue in Hawaii’. American Journal of
Public Health 37: 663–674.
Woodcock, D. (2003). ‘To restore the watersheds: Early twentieth
century tree planting in Hawaii’. Annals of the Association of American
Geographers 93: 624–635.
Ziegler, A.C. (2002). Hawaiian Natural History, Ecology, and Evolution.
Honolulu: University of Hawaii Press.
J.R. Mauremootoo, S. Pandoo, V. Bachraz, I. Buldawoo and N.C. Cole
Mauremootoo, J.R.; S. Pandoo, V. Bachraz, I. Buldawoo and N.C. Cole. Invasive species management in Mauritius: from
the reactive to the proactive – the National Invasive Species Management Strategy and its implementation
Invasive species management in Mauritius: from the reactive to the
proactive – the National Invasive Species Management Strategy
and its implementation
J.R. Mauremootoo1, S. Pandoo2, V. Bachraz3, I. Buldawoo4 and N.C. Cole5,6
InSpiral Pathways, Bristol, UK. <john@inspiralpathways.com>. 2UNDP, Port Louis Mauritius. 3National Parks
and Conservation Service, Ministry of Agro Industry and Food Security (MAIFS) Réduit, Mauritius. 4Entomology
Division, MAIFS, Réduit, Mauritius. 5Durrell Wildlife Conservation Trust, Jersey. 6Mauritian Wildlife Foundation,
Vacoas, Mauritius.
1
Abstract This account provides the context behind the need to implement an integrated, cross-sectoral approach to
Invasive Alien Species (IAS) management in the Republic of Mauritius (RoM). The challenge of increased travel, tourism
and transport is enumerated and the history of IAS entry, establishment and spread in the RoM before and since the
formulation of the National Invasive Alien Species Strategy and Action Plan for the Republic of Mauritius (NIASSAP)
(2010-2019) is reviewed to judge the effectiveness of biosecurity measures at the national and sub-national level. New
incursions appear to have increased since 2010. Examples include, the papaya mealybug (Paracoccus marginatus; 2013),
the Oriental fruit fly (Bactrocera dorsalis), a species that had previously been eradicated from the island (2013 and
2015) and foot and mouth disease (FMD) (Aphthae epizooticae; 2016). There have been some effective responses. A
biological control agent was released against the papaya mealybug and fruit production has recovered, FMD has been
eradicated and a campaign for eradication of B. dorsalis is underway. However, management approaches remain reactive
and sectorally-driven with little cross-fertilisation of ideas and approaches. Biological control, for example, has been very
actively pursued in the agricultural sector but has not been officially undertaken for environmental weeds since 1982.
The documented incursions represent biosecurity failures that the NIASSAP was designed to address but it has yet to be
systematically implemented. The growing impact of new biological invaders on all sectors of the Mauritian economy has
stimulated a revival of interest in biosecurity at the governmental level and in 2016 the Government submitted a US$17M
UNDP/GEF project: Mainstreaming IAS Prevention, Control and Management, which will provide the incremental cost
to review, update and effectively initiate the implementation of the NIASSAP.
Keywords: early detection and rapid response, mainstreaming, management, NISSAP, pathways, prevention, Republic
of Mauritius, risk analysis
INTRODUCTION
The expanding IAS threat in the Republic of
Mauritius and the adoption of the NIASSAP
NIASSAP (‘the proactive’), its implementation to date and
future prospects.
The Republic of Mauritius (RoM) comprises the main
island of Mauritius and Rodrigues, about 560 km to the east
of Mauritius, their associated islets, and the outer islands
of Agalega, Tromelin, Cargados Carajos (St Brandon) and
the Chagos Archipelago. Mauritius and Rodrigues form
part of the Mascarene Islands chain located in the Western
Indian Ocean. The Mascarenes belong to one of the 25
internationally recognised biodiversity ‘hotspots’ (Myers,
et al., 2000). Tropical climate, diverse topography and over
a million of years of isolation have resulted in the evolution
of a diverse biota with a high degree of endemism.
Invasive alien species (IAS) constitute a major threat to
the remaining biodiversity in the RoM (Florens, 2013;
Virah-Sawmy, et al., 2009; Cheke & Hume, 2008). IAS
also have serious economic and health impacts, especially
if the definition of IAS is broadened to include agricultural
pests and zoonotic diseases. This broad conception of IAS
was used when developing the country’s National Invasive
Species Strategy and Action Plan 2010–2019 (NIASSAP)
(RoM, 2010), officially adopted by Cabinet in 2010. The
NIASSAP is based on the premise that the problems of
biological invasions are cross-sectoral in nature, so there
is a need for a harmonised approach to biosecurity that
cuts across traditional sectoral boundaries. Making use
of the ‘biosecurity umbrella’ will help to ensure that all
activities relating to species introductions and spread are
based upon a coordinated and science-based approach that
is underpinned by the assessment and management of risk.
This paper describes some of the RoM’s IAS invasion
trends, its expanding and diversifying IAS pathways, and
examples of IAS management successes and challenges
(‘the reactive’) as a backdrop to the development of the
A brief history of alien species establishment in
Mauritius
Vertebrate establishment
Human actions resulted in the introduction of vertebrates
to Mauritius even before the first documented landing on
the island, by the Dutch in 1598 (Cheke, 1987). Black
rats (Rattus rattus) probably established themselves on
Mauritius via shipwrecks and may have been responsible
for the extinction of many endemic animal species even
before colonisation. Between the first Dutch landing
and settlement in 1638 two major animal invaders, the
Javanese macaque (Macaca fascicularis) and the feral pig
(Sus scrofa) became established in Mauritius. During the
Dutch period (1638–1710), major introductions included
Javan deer (Cervus javanicus) and cats (Felis catus) which
became feral.
During French rule (1721–1810), introductions with
significant negative economic and environmental effects
included the brown rat (Rattus norvegicus), the Asian house
shrew (Suncus murinus) and the tenrec (Tenrec ecaudatus).
The steady rate of vertebrate introductions continued
under the British (1810–1968) with introductions including
the Indian wolf snake (Lycodon capucinus), the redwhiskered bulbul (Pycnonotus jocosus), African landsnails
(Achatina spp.) and the small Indian mongoose (Urva
auropunctata).
Significant vertebrate deliberate and accidental
introductions since independence include the Madagascar
giant day gecko (Phelsuma grandis), the gold-dust day
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 503–509. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
503
Island invasives: scaling up to meet the challenge. Ch 3B Strategy: Collaboration
gecko (Phelsuma laticauda) and the red-eared slider
(Trachemys scripta elegans). All of these introductions
are believed to be due to the pet trade. It would appear
that the numbers of vertebrates establishing in the wild in
Mauritius is showing no signs of a levelling off (Fig. 1).
Also of concern is the spread of vertebrates and all
other taxa between the islands and islets that make up
the Republic of Mauritius. Rodrigues Island, the outer
islands and Mauritian and Rodriguan islets harbour only
a sub-set of the invasive vertebrates found on Mauritius
Island. This has conservation implications. For example:
carnivorous mammals have never established on Round
Island thus saving several endemic reptile species from
extinction (Bullock, 1986); Flat Island was home to 80% of
the world’s population of Bojer’s skink (Gongylomorphus
bojerii) until 2010 when shrews were accidentally
introduced (possibly in building materials) from the
Mauritian mainland causing their local extinction (Cole
& Payne, 2015); and Rodrigues does not have Javanese
macaques which, if introduced, would further threaten
their already fragile native biodiversity. These examples
illustrate the importance of effective inter-island pathway
biosecurity.
Plant establishment
Since colonisation, more than 1,600 plant species
have been introduced to Mauritius. Many of these
introductions have been desirable and others have been
essential as Mauritius only has one native plant species,
the hurricane palm (Dictyosperma album) that has so far
been exploited on a commercial scale. Heeroo (2000)
assembled all introduced plants records from the Mauritius
Herbarium between 1888 and 2000 and found that 804
of the 1,619 species were classified as 'weedy species',
141 being 'agricultural weeds' and 674 being 'naturalised
weeds' (Fig. 2). It should be noted that there can be a
turnover of weedy species so, assuming that the records
are comprehensive, the cumulative number listed is likely
to be higher than the actual numbers of weedy species in
the field. Herbarium records can only approximate the rate
at which species establish themselves as they are heavily
dependent on collection effort but it would appear that new
naturalisations levelled off between the 1980s and 2000.
Data from 2000 onwards need to be consolidated to clarify
recent trends.
Of the naturalised species, about 30 currently dominate
the country’s natural vegetation in terms of numbers of
individuals and biomass. Some of the principal invasive
woody and shrubby plants in Mauritius and Rodrigues
include Psidium cattleianum (Chinese guava) which
constitutes the vast majority of the biomass in much of
Fig. 1 Cumulative records of vertebrate establishment
in Mauritius (pre-1600–2016). Source: Cheke & Hume
2008; Nik Cole (pers. obs.).
504
Mauritius’ humid forest (Florens, et al., 2016), Ravenala
madagascariensis (ravenale) which forms monotypic
stands in similar climatic zones (Baret, et al., 2013),
Hiptage benghalensis (liane cerf) a woody climber which is
increasing in abundance in less humid forests (C. Griffiths
pers. comm. 2015) and Syzigium jambos (jamrosa) which
dominates many riverine landscapes in Mauritius and is
one of the most widespread plant invaders in Rodrigues.
Entry establishment and spread of additional species
can exacerbate an already bad situation. For example,
species belonging to the genus Prosopis, a known invasive
group (Richardson, 1998) have been planted for erosion
control on dry mountain slopes and a proposal for the
plantation of up to 3,200 ha of Arundo donax (giant reed) is
being considered despite its known invasiveness (Csurhes,
2009).
Insect plant pest establishment
Williams & Ganeshan (2001) documented the
acceleration in insect pest establishment in Mauritius
from the 1970s. Data from the Entomology Division of
the Ministry of Agro Industry and Food Security (MAIFS)
indicates that this rate has continued, averaging about
one new pest record per year (Fig. 3). Recent insect pest
introductions include the papaya mealybug (Paracoccus
marginatus), detected in 2013, and the yellow sugar
cane aphid (Sipha flava), detected in 2015. These newly
established pests represent a well-documented burden
on the country's agricultural sector which has become
extremely reliant on the use of synthetic pesticides with all
their concomitant drawbacks (Abeeluck, et al., 2009). The
impacts of newly-established pests on native biodiversity
have not been studied.
Fig. 2 Cumulative records of weedy species from Mauritius
herbarium records 1888–1999. Source: Heeroo 2000.
Fig. 3 New insect pest records in Mauritius 1901–2016.
Source: Entomology Division, MAIFS.
Mauremootoo, et al.: Mauritius National Invasive Species Management Strategy
An example of a repeated insect pest incursion is that of
the oriental fruit fly (Bactrocera dorsalis): Native to Asia,
B. dorsalis is one of the world's most destructive pests of
fruit with over 300 host species. Bactrocera dorsalis is
now found in at least 65 countries and continues to spread
via infested fruit, either as cargo or carried illegally by
airline passengers (CABI, 2017). It was first detected in
Mauritius in 1996 and, following an eradication campaign
involving bait spraying, male annihilation, fruit collection
and destruction, Mauritius was declared free of B. dorsalis
in 1998 (Seewooruthun, et al., 2000). A further incursion
was detected in 2013 and eradicated using similar methods
in 2014. The pest was discovered once again in 2015 but at
many more locations that previously, so eradication using
established methods was not possible. The population
is currently being contained and suppressed while an
irradiation facility for the breeding of sterile males is being
constructed. The first release is scheduled for February
2018 to treat an area of 400 km2 with the release of 15
million males per week for at least eight months. In the
medium term, this programme will be expanded into
eradication campaigns for the eight other fruit fly species
present in Mauritius (P. Sookar pers. comm. 2017). These
planned eradications are being accompanied by increased
pest screening of all imported fruit and vegetables.
Disease establishment
A number of zoonotic diseases have been introduced
into Mauritius in recent years. The country experienced a
major outbreak of chikungunya, a debilitating mosquitoborne virus, in 2006 (Ramchurn, et al., 2008), it’s first ever
outbreaks of African swine fever in 2007 (Lubisi, et al.,
2009) and in 2016 the first foot and mouth disease outbreak
in 100 years (Hamuth-Lauloo, et al., 2016). All three
diseases are no longer present in the RoM but the outbreaks
had major social, political, economic and environmental
impacts.
IAS pathways are expanding and diversifying
The major IAS pathways for Mauritius are international
shipping and air travel but there is also a risk posed by the
Fig. 4 Passenger arrivals in (A) Mauritius Island and (B)
Rodrigues Island 1983–2016 (prior to 1994 figures
exclude cruise travellers). Source: Statistics Mauritius
(2017).
unknown numbers of pleasure crafts that land informally
on the Mauritian mainland, Rodrigues and their associated
islets and are therefore unregulated.
The volume and diversity of traffic along air and
sea pathways into and within the RoM has increased
substantially over the past decades. Passenger arrivals
into Mauritius Island have increased nearly tenfold from
177,665 in 1983 to 1,684,835 in 2016 (Fig. 4a). Of these
arrivals, 409,608 were returning Mauritian residents.
Arrivals into Rodrigues over the same period have
increased more than fourteenfold from 6,556 to 94,270
(Fig. 4b). Most flights to Rodrigues come from Mauritius
Island with an additional scheduled service from Réunion.
In June 2017, Air Mauritius, the country’s national carrier,
was running scheduled services to 24 destinations in 15
countries.
Despite the large increase in absolute numbers, the
proportion of travellers from different regions of the
world has been relatively consistent (Fig. 5). However,
it is highly likely that the diversity of passenger origins
has increased substantially although this is not possible to
conclude definitively as comprehensive data on passenger’s
original port of embarkation is only available from 2013.
According to Statistics Mauritius (2017), passengers began
their journey in at least 110 countries or territories in 2016.
Air travel to and through Mauritius is likely to further
increase through the continued growth of the tourism
industry and the efforts Mauritius is making to position
itself as an air travel hub for the fast-growing Africa-Asia
market. More and more people coming from more and
more biogeographic zones has biosecurity implications for
a variety of frontline agencies – currently the airport entry
point is staffed by representatives of the the Mauritius
Revenue Authority (Customs Department), the Ministry of
Agro Industry and Food Security (Plant Protection Office)
and Ministry of Health and Quality of Life.
Imports into Mauritius by weight have increased nearly
seven-fold from 905,398 tonnes in 1974 to 6,007,056
tonnes in 2016 (Fig. 6). At the same time the number of
countries exporting to Mauritius has increased from 33
to 61. However, in contrast to the continued growth in
tonnage, the increase in numbers of exporting countries
levelled off from 2000 (Fig. 7). The relative importance of
exporting countries by monetary value of their exports to
Mauritius has changed with two trends being particularly
evident: the growth in exports from Asia (from 37–53%)
and decline in exports from Europe (from 40–25%) (Fig. 8).
These changes have implications for biosecurity including
the increased risk that comes with greater volumes of
movement, increases in the numbers of source locations
and an increase in sources from the warmer parts of the
world. However, the precise nature of the risks involved
Fig. 5 Foreign passenger arrivals per region of residence.
Source: Statistics Mauritius (2017).
505
Island invasives: scaling up to meet the challenge. Ch 3B Strategy: Collaboration
cannot be ascertained from aggregate figures. For example,
goods such as high value electronics and processed grains,
which can be relatively ‘clean’, pose lower risks than
‘dirtier’ imports such as semi-processed and unprocessed
food and timber, and used machinery which can harbour a
wide range of invasive species.
Nearly all official shipping to Rodrigues comes from
Mauritius Island, which simplifies pathway analysis.
However, pleasure boats and artisanal fishing boats also
operate between Mauritius and Rodrigues islands and their
offshore islets. Precise numbers of local and international
visitors to islets have not been recorded but are known
to be in the hundreds of thousands per year. Biosecurity
practices are adopted for organised tours of islet nature
reserves such as Ile aux Aigrettes and for conservation
missions to Round Island and Gunner’s Quoin but similar
protocols have yet to be formally adopted by private tour
operators, pleasure craft owners or fishers.
IAS MANAGEMENT SUCCESSES AND
CHALLENGES
Prevention: keeping white grubs out of Mauritius
The white grub (Hoplochelus marginalis), a beetle
indigenous to Madagascar, was accidently introduced to
Réunion, 150 km to the west of Mauritius, in 1973 in potted
ornamental plants and became a threat to Réunion’s sugar
industry in the 1980s (Jeuffrault, et al., 2004). Sugar cane
is Mauritius’ principal crop and the generalist white grub
also affects other crops and wild grasses. Collaboration
between Mauritius and Réunion has prevented the white
grub from reaching Mauritius and includes the following
measures: reduction in the population densities of white
grubs in Réunion by the use of the fungal pathogen
Beauveria brongnartii; sustained public awareness
campaigns; changes in flight and boat departure times in
summer when the beetle actively flies around dusk and is
attracted to light; regular inspections and spraying around
the Mauritius port and airport areas. This systematic
approach reflects sugar cane’s economic importance and
the priority given to agriculture.
Early detection and rapid response: stopping redbacks
in their tracks
Fig. 6 Volume of imports in tonnes 1974–2016. Source:
Statistics Mauritius (2017).
Fig. 7 Number of countries that export to Mauritius 1974–
2016. Source: Statistics Mauritius (2017).
The redback spider (Latrodectus hasseltii) is a
venomous Australian spider, responsible for far more bites
requiring antivenom than any other creature in Australia. It
was found on Gunner's Quoin, an islet of key conservation
importance, 8 km from the north coast of Mauritius in
2008 by scientists carrying out conservation activities
(N. Cole, pers. obs.). The individual spider and three egg
cocoons were found, the spider was collected to confirm
identification and the cocoons destroyed, although two had
previously hatched. Systematic searches were conducted
and subsequently three more spiders and additional egg
cocoons were detected and destroyed. Since 2010, there
have not been any new detections despite intensive surveys
every four to six months. It is not known how the redback
was introduced but it is suspected that it could have been a
stowaway on private yachts from Australia that are known
to travel in the region. Since 2010, 15 invasion events by
10 invertebrate and vertebrate species have been detected
on six islets surrounding Mauritius. The periodic presence
of biologists on these islets has in most cases permitted
rapid response resulting in seven of these invasion events
being prevented from establishing or subsequently
eradicated with another two eradication efforts ongoing.
Increased use of the islets for tourism and leisure activities
have been identified as the most significant IAS pathway.
Effective biosecurity systems do not exist for most islets
with the exception being Round Island which is managed
for strict conservation purpose, is difficult to access, and is
permanently staffed by conservationists.
Eradication: elimination of foot and mouth disease
from Mauritius and Rodrigues
Fig. 8 Value of imports by region 1974–2016. Source:
Statistics Mauritius (2017).
506
The following is a summary of the detailed account
given by Hamuth-Lauloo, et al. (2016). From 7–27 July
2016, 62 cases of cattle illness had been reported in
Rodrigues. On 31 July, a team from Mauritius observed Foot
and Mouth (FMD) symptoms in cattle and pigs. This was
confirmed by blood tests on 1 August. In the meantime, two
consignments of livestock had been exported to Mauritius
Island. The presence of FMD was confirmed in Mauritius
on 5 August. The most probable source of FMD was frozen
buffalo meat imported from India via Mauritius. The
response comprised of stamping out, movement control,
disinfection, quarantine, surveillance, destruction of
animal products, official disposal of carcasses, by-products
and waste, zoning and vaccination and no FMD cases have
Mauremootoo, et al.: Mauritius National Invasive Species Management Strategy
been detected from both Rodrigues and Mauritius since
December 2016. Inspection of export facilities in India
could have prevented the outbreak, more rapid diagnosis
and better inter-island quarantine could have reduced its
severity and spread and a contingency plan would have
resulted in a more coordinated response than was the case.
Management: the use of biological control within an
IPM approach in the agricultural sector
eradicate those alien species which threaten ecosystems,
habitats or species” (UN, 1992). Towards this end,
MAIFS established the National Invasive Alien Species
Committee (NIASC) in August 2003. One of the priorities
for the NIASC, which comprises representatives from the
agriculture, biodiversity conservation, health, environment
and education sectors as well as the private sector, was the
production of a National Invasive Alien Species Strategy
and Action Plan (NIASSAP) for the Republic of Mauritius
(RoM, 2010). Funding was secured for its development
from 2008 and the NIASSAP was approved by cabinet
in 2010. The NIASSAP presents a vision in which the
negative impacts of IAS on the economy, environment and
society of the RoM are avoided, eliminated or minimised.
The strategy was based on the assumptions that an effective
biosecurity system is built upon a risk analysis framework
and that its success depends upon effective collaboration
between all those concerned with invasion pathways.
As outlined, there has been an increased rate of insect
pest introduction to Mauritius since the 1970s. This has
been one of the reasons for the growing use of pesticides
in Mauritian agriculture. However, at the same time, the
country, notably through the sugar sector, which barely
uses insecticides, has made grounds in integrated pest
management (IPM), advocating a package of measures
designed to reduce the prophylactic use of pesticides. One
of the main planks of this approach has been the use of
biological control. This has been reflected in the consistent
use of biological control agents (parasitoids, pathogens,
and biopesticides) in recent decades. A major recent
success was the introduction of the parasitoid Acerophagus
papayae in 2013 to control the papaya mealybug.
The Strategy comprises ten interlinked elements: five
hierarchical “Management Elements” and five “CrossCutting Elements”. The management elements are those
“on the ground actions” that directly address the Strategy’s
vision. The cross-cutting elements are enabling actions.
The sectoral nature of IAS management – the case of
biological control
The NIASSAP Management Elements, with their
accompanying goal or goals are listed in order of priority
based on the maxim that “prevention is better than cure”:
Approaches developed in one sector are not necessarily
adopted and adapted to other sectors. An example of this is
biological control which is actively pursued in agriculture
but not in the conservation sector. The priority given to
biological control in agriculture, using protocols based
on International Sanitary and Phytosanitary Measures
(ISPMs), reflects the sector’s economic importance and the
clear direction offered to the plant protection sector through
the International Plant Protection Convention (IPPC), to
which the country has been a signatory since 1971. There
has been no deliberate introduction of a biocontrol agent
against an invasive plant that threatens native biodiversity
since 1982, despite the fact that biocontrol of environmental
weeds in Mauritius has a very successful history with a full/
partial success rate of 80% (Fowler, et al., 2000). Ironically,
two recent examples of possible biosecurity failures are
likely to have had positive impacts on biodiversity in
the RoM. Firstly, there is the movement of Teleonemia
scrupulosa (lantana lace-bug), a biological control agent
for Lantana camara (vieille fille) already present in
Mauritius, to Rodrigues which has hugely reduced the
vigour of L. camara in areas of conservation importance
and on rangeland in Rodrigues. Secondly, the spread of
the biocontrol agent Cibdela janthina (mouche bleu) from
Réunion to Mauritius, which may have arrived in 2015
(Florens, et al., 2017), has the potential to substantially
reduce the vigour of Rubus alceifolius (giant bramble) a
major invasive plant in Mauritian forests. C. janthina could
have conceivably dispersed naturally from Réunion but the
chances of the L. camara agent dispersing naturally from
Mauritius to Rodrigues are very low. Whatever the case,
biosecurity systems need to be tightened but responsible
biological control for invasive plants must to be part of an
integrated approach to invasive plant management.
THE DEVELOPMENT OF A NATIONAL INVASIVE
ALIEN SPECIES STRATEGY AND ACTION PLAN
Following its accession to the Convention on Biological
Diversity in 1992, the Mauritian conservation community
was very actively engaged in the Global Invasive Species
Programme (GISP) which operated between 1997 and 2011
to encourage the adoption of measures in line with CBD
Article 8h: “Each Contracting Party shall, as far as possible
and as appropriate prevent the introduction of, control or
1. Prevention – to minimise the number of unintended
and intended IAS introductions to the RoM;
2. Early detection and rapid response – to minimise
the number of IAS that go on to have harmful
consequences once they are introduced to the RoM;
3. Eradication – an agreed framework for eradication
priorities in place, eradications undertaken as
necessary and results disseminated;
4. Control and management – to contain the distribution
and abundance of IAS in the RoM to a long-term
acceptable level; and
5. Restoration – to undertake ecosystem restoration
where necessary in the RoM to achieve long-term
ecosystem goals.
The Cross-Cutting Elements, again listed with their
goal or goals, are:
6. Legal, policy and institutional frameworks – to have
a coordinated policy and management framework
that minimise the risk of IAS;
7. Capacity building and education – to make available
appropriately skilled personnel to implement all
aspects of IAS management in the country;
8. Information management and research – (i) To have
a clear understanding of the impacts of IAS that
have become established in the RoM; (ii) to have
ready access to critical information that will support
IAS management programmes and (iii) to provide
a strong scientific basis for decision-making and
resource allocation;
9. Public awareness and engagement – all stakeholders
in the RoM should have a high level of awareness
of IAS risks and the benefits of IAS prevention and
management;
10. International cooperation – (i) the RoM should
have access to the necessary information, technical
and financial support and other resources it needs
to effectively meet its international obligations;
(ii) Mauritian IAS experiences and lessons learnt
are effectively disseminated to help IAS initiatives
regionally and internationally and (iii) the RoM is
not a source of IAS for other countries
507
Island invasives: scaling up to meet the challenge. Ch 3B Strategy: Collaboration
Partial implementation of the NIASSAP (2010–2017)
The NIASSAP has yet to be systematically
implemented. Major reasons for this were the fact that lead
agencies were not designated to carry out each action and
timelines, milestones and estimates of resources required
were not agreed upon. The National IAS Committee
only met sporadically between 2010 and 2015 and was
only made statutory in 2015 under the Native Terrestrial
Biodiversity and National Parks Act (2015).
However, actions in line with the NIASSAP have
been undertaken in Mauritius since 2010, some of which
have been outlined above, but they were not implemented
because of the NIASSAP.
The prospects for effective implementation of the
NIASSAP received a boost with a broadly costed and
timetabled provision for its implementation under the
National Biodiversity Strategy and Action Plan (NBSAP)
2017–2025 as the National contribution to Aichi Target 9:
“By 2025, the NIASSAP is revised and fully implemented
through adequate financial and human resources
commensurate to the existing challenges, and the impacts
caused by IAS are minimised” (RoM, 2017). Linked to
the above, from 2015–2018, has been the development of
a UNDP/GEF VI Project to mainstream IAS prevention,
control and management (US$20.89M project: US$3.89M
from Global Environment Facility and US$17M from
National co-financing).
Next steps: mainstreaming IAS prevention, control
and management
The objective of the ‘IAS Mainstreaming Project’
is to safeguard globally significant biodiversity in
vulnerable ecosystems through the prevention, control and
management of IAS in the RoM. This will be achieved
through four outcomes which are summarised below
together with key outputs and activities that contribute to
intended outcomes:
1. By 2024, the RoM has a gender sensitive policy,
regulatory and institutional framework and capacity to
manage IAS effectively:
● The NIASSAP is reviewed and revised, with
progress assessed, gaps identified and activities fully
costed with precise timelines for implementation for
both terrestrial and marine IAS;
● Existing legislation is strengthened for more effective
control and management of IAS;
● A cross-sectoral policy coordination framework
is established for the incorporation of IAS issues
into the legal and policy framework of all relevant
agencies;
● A technical secretariat for IAS is established;
● Capacity is strengthened in key agencies and
organisations;
● Financial sustainability of the apex agency and IAS
operations are secured through the development
and application of new market-based and fiscal
mechanisms and incentives to support IAS
management.
2. By 2024, the government effectively prevents and
manages IAS threats based on risks:
● National and inter-island biosecurity priorities and
resource needs, including baselines are established;
● A comprehensive risk assessment system is in place
and being used in the Republic of Mauritius, to
(1) assess the risks that new species proposed for
508
importation to the RoM or moved between its islands
may become invasive (border control), and (2) assess
the risks associated with species already present
in the RoM but which may not yet have become
invasive there;
● Species identified by formal risk assessment as
having high invasiveness potential in the Republic of
Mauritius are refused permission for importation or
for translocation between its islands;
● Procedures for controlling the unregulated (illegal)
importation of species to the Republic of Mauritius
or between its islands are improved (effective
quarantine system with sanctions for deliberate
infractions);
● Species present in the Republic of Mauritius, with
high invasive potential but still present only in
limited areas, are prioritised for management and,
where feasible, eradication by means of a formal risk
assessment process, including, as far as possible,
their declaration as “harmful”, “prohibited”, or
similar;
● Pilot biodiversity conservation and ecological
restoration operations developed on key islets and in
Rodrigues;
● Equipment and infrastructure updated to help ensure
that priority biosecurity measures are effectively
implemented.
3. By 2024, planning, management and decision-making
by all relevant stakeholders are informed by knowledge
management and learning:
● Review and survey of the status of IAS pathways,
IAS distributions, the cross-sectoral economic,
environmental and cultural impact of IAS and the
successes and lessons learnt from past and ongoing
IAS prevention, early detection and rapid response,
eradication, control and mitigation and restoration;
● Up-to-date lists of terrestrial and marine invasive
species of all taxa present in the Republic of
Mauritius are completed and publicly available, and
a system for their regular updating is in place and is
being used;
● Pathways of introduction of new species into
Mauritius and between the islands of the Republic of
Mauritius are identified, their relative importance is
quantified, and they are prioritised for management
action to reduce the rate of arrival of new species;
● A national IAS information system is developed and
operationalised to monitor and inform risk-based
management of species, pathways and ecosystems
based on agreed protocols;
● A national IAS gateway is developed to provide
rapid access and dissemination of information to
enhance deployment of coordinated actions between
institutional partners on IAS management;
● A national IAS communications and awareness
strategy and action plan is developed and
implemented;
● IAS tools and manuals are developed to complement
training courses and for use in day to day IAS
management operations, and guidelines are
developed to embed IAS issues into key sectors
whose activities have IAS implications.
Project risks include increased liberalisation of
movement and trade, the continuation of a fragmented
sector by sector and case by case approach, lack of
support for strengthened biosecurity measures at different
levels, and economic and political pressure being used to
Mauremootoo, et al.: Mauritius National Invasive Species Management Strategy
circumvent decision-making based on a transparent risk
analysis process. It is clear, therefore, that the NIASSAP
represents an ambitious and costly undertaking, but the
costs of not systematising IAS prevention and management
(business as usual) are likely to be considerably higher.
ACKNOWLEDGEMENTS
Thanks to staff of the Ministry of Agro-Industry and
Food Security (National Parks and Conservation Service,
Forestry Service, Plant Quarantine Service and Entomology
Division) and the Mauritian Wildlife Foundation for
helping to provide information for this paper and for their
contributions towards putting invasive species on the
national radar. Thanks to the reviewers for their valuable
comments on an earlier draft of this paper. A special thanks
to all the dedicated staff from the concerned agencies who
undertake the daily activities to minimise invasive species
impacts in the Republic of Mauritius.
REFERENCES
Abeeluck, D., Dunhawoor, C. and Unmole, L. (2009). ‘Development
and implementation of integrated pest management in Mauritius: An
overview’. University of Mauritius Research Journal 15: 85–97.
Baret, S., Baider, C., Kueffer, C., Foxcroft, L.C. and Lagabrielle, E.
(2013). ‘Threats to Paradise? Plant Invasions in Protected Areas of
the Western Indian Ocean Islands’. In: L.C. Foxcroft, P. Pyšek, D.M.
Richardson and P. Genovesi (eds.) Plant Invasions in Protected
Areas: Patterns, Problems and Challenges, pp. 423–447. Dordrecht,
Netherlands: Springer.
Bullock, D.J. (1986). ‘The ecology and conservation of reptiles on Round
Island and Gunner’s Quoin, Mauritius’. Biological Conservation 37:
135–156.
CABI. (2017). ‘Bactrocera dorsalis (oriental fruit fly), Invasive Species
Compendium’. <http://www.cabi.org/isc/datasheet/17685>. Accessed
30 April 2017.
Cheke, A.S. (1987). ‘An Ecological History of the Mascarene Islands,
with Particular Reference to Extinctions and Introductions of Land
Vertebrates’. In A.W. Diamond (ed.) Studies of Mascarene Island Birds,
pp. 5–89. Cambridge: Cambridge University Press.
Jeuffrault, E., Rolet, A., Reynaud, B., Manikom, R., Georger, S., Taye, T.,
Chiroleu, F., Fouillaud, M. and Vercambre, B. (2004). ‘Vingt ans de lutte
contre le ver blanc de la canne à sucre à la Réunion: Un succès, mais
il reste des questionnements scientifiques pour confirmer la durabilité
de la lutte biologique’. Phytoma–La Défense des Végétaux 573: 16–19.
Lubisi, B.A., Dwarka, R.M., Meenowa, D. and Jaumally, R. (2009).
‘An Investigation into the first outbreak of African swine fever in the
Republic of Mauritius’. Transboundary and Emerging Diseases 56:
178–188.
Myers N., Mittermeier, R.A., Mittermeier, C.G. et al. (2000). ‘Biodiversity
hotspots for conservation priorities’. Nature 403: 853–858.
Ramchurn, S.K., Goorah, S.S.D., Mungla, D., Ramsurrun, B., Pydiah, V.
and Summun, A. (2008). ‘A study of the 2006 Chikungunya epidemic
outbreak in Mauritius’ Internet Journal of Medical Update 3(1): 11–21.
Richardson, D.M. (1998). ‘Forestry trees as invasive aliens’. Conservation
Biology 12(1): 18–26.
RoM. (2010). ‘The National Invasive Alien Species Strategy and Action
Plan for the Republic of Mauritius’. Port Louis: Ministry of Agro
Industry and Food Security.
RoM. (2017). ‘National Biodiversity Strategy and Action Plan (NBSAP)
2017 – 2025’. Port Louis: Ministry of Agro Industry and Food Security.
Seewooruthun, S.I., Permalloo, S., Gungah, S., Soonnoo, A.R. and Alleck,
M. (2000). ‘Eradication of an Exotic Fruit Fly from Mauritius’. In: K-H.
Tan (ed.) Area-Wide Control of Fruit Flies and Other Insect Pests: Joint
Proceedings of the International Conference in Area-Wide Control
of Insects Pests, May 28−June 5, 1998, and the Fifth International
Symposium on Fruit Flies of Economic Importance, June 1–5, 1998, pp.
389–394. Pulau Pinang, Malaysia: Penebrit Universiti Sains.
Statistics Mauritius (2017). ‘2015–2016 Statistics Mauritius Annual
Report’. Port Louis, Mauritius: Ministry of Finance & Economic
Development.
UN. (1992). ‘Convention on Biological Diversity’. Nairobi. <https://
www.cbd.int/convention/text/default.shtml>. Accessed 30 April 2017.
Virah-Sawmy, M., Mauremootoo, J.R., Marie, D., Motala, S.M. and
Sevathian, J-C. (2009). ‘Rapid degradation of a Mauritian rainforest
following 60 years of plant invasion’. Oryx 43(4): 599–607.
Williams, J.R. and Ganeshan, S. (2001). ‘Insects introduced into
Mauritius during the 20th century: Natural, accidental and deliberate
introductions’. Revue Agricol et Sucrière de l’Ile Maurice 78: 33–50.
Cheke, A.S. and Hume, J.P. (2008). Lost Land of the Dodo: An Ecological
History of Mauritius, Réunion and Rodrigues. London: T & AD Poyser.
Cole, N. and Payne, C. (2015). ‘Gongylomorphus bojerii, The IUCN Red
List of Threatened Species 2015’. <http://dx.doi.org/10.2305/IUCN.
UK.2015-4.RLTS.T62251A13482733.en>. Accessed 25 June 2017.
Csurhes, S. (2009). ‘Weed Risk Assessment: Giant Reed (Arundo donax)’.
Australia: The State of Queensland: Department of Employment,
Economic Development and Innovation. <http://www.feedipedia.org/
node/18477>. Accessed 10 April 2017.
Florens, F.B.V. (2013). ‘Conservation in Mauritius and Rodrigues:
Challenges and Achievements from two Ecologically Devastated
Oceanic Islands’. In: N.S. Sodhl and P.H. Raven (eds.) Conservation
Biology: Voices from the Tropics, pp. 40–50. Winchester: WileyBlackwell.
Florens, F.B.V., Baider, C., Martin, G.M.N., Seegoolam, N.B., Zmanay,
Z. and Strasberg, D. (2016). ‘Invasive alien plants progress to dominate
protected and best-preserved wet forests of an oceanic island’. Journal
for Nature Conservation 34: 93–100.
Florens, F.B.V., Bissessur, P., Bunsy, Y. and Ramdonee, D. (2017).
‘Cibdela janthina (Klug 1834) (Hymenoptera: Tenthredinidae),
Réunion’s biocontrol agent of Rubus alceifolius Poir., recorded on
Mauritius’. African Entomology, 25(1): 271–274.
Fowler, S.V., Ganeshan, S., Mauremootoo, J.R. and Mungroo, Y. (2000).
‘Biological Control of Weeds in Mauritius: Past Successes Revisited
and Present Challenges’. In: N.R. Spencer (ed.) Proceedings of the X
International Symposium on Biological Control of Weeds, pp. 43–50.
Bozeman, Montana: Montana State University.
Hamuth-Lauloo, S.B.A., Sibartie, D. and André, A.A. (2016). ‘Report
of the Fact-Finding Committee on the Outbreak of Foot and Mouth
Disease’. Government of Mauritius: Port Louis. <http://agriculture.
govmu.org/English/Documents/Report/full%20report.pdf>. Accessed
30 April 2017.
Heeroo, R. (2000). ‘Documentation of the Weed Status of Exotic Plants
in Mauritius based on the Records of the Mauritius Herbarium’. BSc
dissertation. Réduit: University of Mauritius.
509
J.-Y. Meyer and M. Fourdrigniez
Meyer, J.-Y. and M. Fourdrigniez. Islander perceptions of invasive alien species: the role of socioeconomy and culture in small isolated islands of French Polynesia (South Pacific)
Islander perceptions of invasive alien species: the role of socioeconomy and culture in small isolated islands of French Polynesia
(South Pacific)
J.-Y. Meyer1 and M. Fourdrigniez2
Délégation à la Recherche, Government of French Polynesia, B.P. 20981, 98713 Papeete, Tahiti, French Polynesia.
<jean-yves.meyer@recherche.gov.pf>. 2Groupement Espèces Envahissantes, Bioconsulting, B.P. 50902,
98716 Pirae, Tahiti, French Polynesia.
1
Abstract Islands, often celebrated as natural laboratories for evolution and ecology, also provide unique experimental
grounds for societal studies. Although biological invasions are widely recognised as one of the main causes of biodiversity
erosion and a driver of global change, the human perception of invasive species may vary at regional and local levels,
especially in societies with different levels of socio-economic development and cultures. This study was conducted in
French Polynesia (South Pacific), a territory formed by 120 tropical and subtropical oceanic islands (76 being inhabited)
divided into five archipelagos (Austral, Marquesas, Society, Tuamotu, and Gambier Is), comprising both highly populated
and urbanised islands (such as Tahiti in the Society Is) and less populated and very small islands, sometimes very isolated
(without airstrips) and where traditional life style and strong dependence on natural resources still persist. During an
eight-month education and prevention campaign targeting alien plant and animal species legally declared invasive in
French Polynesia, public meetings were organised on 19 small islands for a total of 2,045 consulted people in 41 different
villages. Negative, positive and neutral comments made by participants on some invasive species present in their islands
were recorded. Our results show that their perceived status differs from one archipelago to another, or even among
islands in the same archipelago, with more positive comments (i.e. species benefits) on more isolated islands. Perception
of invasiveness varied according to societal and cultural values (e.g. utilitarian or aesthetic), and often depends on the
species’ date of introduction (“indigenisation” of old introduced plants and animals). These surveys can provide useful
baseline information on the degree to which local island communities are likely to support invasive species management,
to get involved in prevention, surveillance and control efforts, and to avoid potential conflicts of interest between different
stakeholders in small but sometimes complex insular societies.
Keywords: conflicts, indigenisation, invasiveness, isolation, prevention, social dimension, values
INTRODUCTION
The human or social dimension is increasingly
recognised as a crucial issue for the effective management
of invasive alien plants and animals (McNeely, 2000;
Marshall, et al., 2011; Estévez, et al., 2014). Indeed, many
control, eradication or prevention programs have been
delayed or even failed because of differing public attitudes
and feelings towards the targeted invasive species. The
various stakeholders (such as foresters, pastoralists,
horticulturists, pet shop managers, conservationists and
environmentalists) may have different or opposite views
of species status (e.g. “noxious/harmful” versus “useful/
beneficial” species) and strong opposition by some
influential groups of people or even single individuals may
occur. Control or eradication programs of animals such
as feral cats (Felis catus), feral deer (Cervus spp.), pigs
(Sus scrofa), or grey squirrels (Sciurus carolinensis) (see
references in McNeely, 2000; Estévez, et al., 2014), and
of plants such as gorse (Ulex europaeus) in New Zealand
(Hill, 1989) or strawberry guava (Psidium cattleianum)
in the Hawaiian islands (Veitch & Clout, 2000; Warner
& Kinslow, 2013) and La Réunion (Mascarene Is, Indian
Ocean) are well-documented examples of social conflicts
of interests, often associated with “controversies” reported
in public and media opinions.
Thus, studying human perceptions and attitudes
towards invasive species is often useful and sometimes
an important prerequisite before starting often costly and
long-term management programmes. Many recent studies
have been conducted in “western” and/or well-developed
regions/countries, such as Europe, Canada and USA
(Bremner & Park, 2007; Garcia-Llorente, et al., 2008;
Selge, et al., 2011; Fischer, et al., 2014), and New Zealand
(Fraser, 2001; Russell, 2014), using questionnaires or
interviews addressed to different stakeholders among
different socio-professional categories. A few other studies
have been conducted in developing countries where
invasive species may sometimes constitute a natural
resource rather than a nuisance (e.g. the potential use
of water hyacinth (Eichhornia crassipes) as biofuel in
south-east Asia, Bhattacharya & Kumar, 2010). The case
of “true” island countries and territories (excluding large
continental islands such as Australia, Madagascar, or Great
Britain) is even less studied, although they are highly
vulnerable to the impacts of invasive alien species, with
many cases of native species’ extinction and extirpation
and stronger conservation challenges. Moreover, islands,
often celebrated as natural laboratories for evolution and
ecology, may also provide unique experimental grounds
for societal and cultural studies, as they also harbour a high
cultural diversity and different levels of socio-economic
development. In this study conducted in the small tropical
oceanic islands of French Polynesia (South Pacific), we
tested the two following hypotheses:
Does human perception of invasive species vary with
island isolation, human population and socio-economic
development?
What is the influence of cultural (traditional) values on
public attitudes toward introduced species in small remote
islands?
MATERIAL AND METHODS
This study was conducted in French Polynesia, a
European Overseas Country and Territory (OCT) located
in the South Pacific, formed by about 120 small tropical
oceanic islands (76 being inhabited by a total of ca.
276,000 inhabitants in 2017) divided into five archipelagos
(Austral, Marquesas, Society, Tuamotu, and Gambier Is),
and dispersed over a marine area as wide as Europe (Fig. 1).
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
510
up to meet the challenge, pp. 510–516. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Meyer & Fourdrigniez: Islander perceptions of invasive species
An oral PowerPoint presentation listing and describing
the 46 legally declared invasive species (38 of which
were present in the surveyed small islands) was delivered,
without providing details on their ecological and socioeconomical impacts. Two main following questions were
asked to the participants:(1) do you know or have you seen
these species in your island? (2) do you consider them
invasive (i.e. abundant and/or spreading) in your island,
and where (i.e. which locations)?
Although no direct question was asked about species
perceptions and associated values, comments were given
by participants related to the negative impacts of species
on biodiversity and other sectors (e.g. agriculture, health),
and also their positive impacts (past and current benefits),
which were systematically recorded.
RESULTS
Effects of island isolation, human population and
socio-economic development
Fig. 1 French Polynesia and its 120 tropical oceanic
islands located in the South Pacific. The names of the 19
surveyed small islands are underlined.
This OCT comprises both highly populated and urbanised
islands (such as Tahiti, the largest with a land area of 1,045
km² and over 183,000 inhabitants) and very small isolated
islands (sometimes without airstrips such as Tahaa in the
Society, Fatu Iva and Tahuata in the Marquesas, Makatea
in the Tuamotu, Rapa in the Austral Is with an area of only
40 km² and 515 inhabitants), which are less populated
and developed, where traditional lifestyles and strong
dependence on natural marine and terrestrial resources still
persist. As an example, coconut plantations for copra and
coconut oil production remain the main source of income
in the Leeward Islands (Society), the Tuamotu atolls and
the Marquesas high volcanic islands (IEOM, 2017). The
island isolation or “remoteness” (distance from the most
urbanised and populated island of Tahiti in km) and the
number of regular flights per week departing from Tahiti or
“connectivity” were used as proxies for the socio-economic
development of each surveyed island.
Environmental matters and issues fall to the authorities
of the French Polynesian Government, (i.e. they are
different from French laws and regulation texts), with a
“Code de l’Environnement de la Polynésie Française” voted
by the Assembly of French Polynesia in 2003, including a
chapter specifically dedicated to invasive alien species. A
total of 46 species including 35 plants and 11 animals have
been legally declared “a threat to biodiversity” in French
Polynesia (Table 1) because of their significant negative
impacts on the endemic fauna and flora. New introduction,
culture or propagation, as well as inter- and intra-island
transportation, of these species is banned in all islands of
French Polynesia and control or eradication programmes
have been set up. Their presence on each inhabited
island was compiled based on literature, plant and animal
databases and local expertise (Fourdrigniez, et al., 2014).
During a communication, education, prevention and
capacity building campaign conducted (by the second
author M.F.) between May and December 2014 (about
eight months), public meetings were organised on 19
small islands (< 400 km² and 10,000 inhabitants) within
41 different villages. A total of about 2,045 people were
consulted (Table 3). These meetings were held at the city
halls (“mairie” in French) or community houses during
the morning or the evening, and were attended mainly by
adults (for a total of 1,781) and some schoolchildren.
The total number of legally declared invasive alien
species known to be present in each surveyed island
(according to Fourdrigniez, et al., 2014) in the four
archipelagos of the Leeward (Society), Austral, Marquesas
and Tuamotu Is does not decrease with island remoteness
(Fig. 2), comprising 44 of the 46 invasive alien species
(Table 3). Invasive species diversity also does not increase
with island size (Table 3) although the two largest remote
islands of Hiva Oa and Nuku Hiva in the Marquesas (>
300 km² of land area) have a high proportion of species
(between 50–56% of the total), probably related to their
higher habitat diversity (ranging from coastal vegetation
and littoral forest to dry-mesic forests, valleys and slopes
rainforests, and montane cloudforests and summit ridges
up to 1,200 m elevation, Lorence, et al., 2016) compared
to the other surveyed islands. There is a relatively weak
correlation between invasive species and the number
of inhabitants (R²=0.48, P-value < 0.01, Fig. 3a), which
becomes stronger with the number of regular flights
departing from Tahiti per week (R²=0.53, P-value <
0.05, Fig. 3b), i.e. with human and goods transportation
connection and frequency. This “connectivity” between
Tahiti and the other French Polynesian islands constitutes
a very good proxy for the socio-economic development of
isolated islands. If the Tuamotu atolls are removed from the
analysis, the correlation coefficient is significantly higher
(R²=0.72). Indeed, the atolls and raised atolls have fewer
invasive species mainly because of their small terrestrial
areas, their calcareous substrate and strong insolation
Fig. 2 Relationship between the number of invasive alien
species on Tahiti and the 19 surveyed small islands
according to distance from Tahiti: Leeward Is (Society
Is) >170–310 km from Tahiti; Tuamotu Is >220–350 km;
Austral Is >500–700 km; Marquesas Is >1,000–1,500
km (Spearman test, P-value = 0.000995).
511
Island invasives: scaling up to meet the challenge. Ch 3B Strategy: Collaboration
Table 1 List of the 46 invasive alien species legally declared a “threat to biodiversity in French Polynesia” (according to
the French Polynesia “Code de l’Environnement”) and their presence in Tahiti and the other 19 surveyed small islands
(Fourdrigniez, et al., 2014).
ANIMALS: INVERTEBRATES AND VERTEBRATES (N=11)
Kingdom
Scientific name
Common name
Tahiti
Surveyed islands (%)
Insects
Wasmannia auropunctata*
Little fire ant
X
0 (0%)
Molluscs
Euglandina rosea*
Rosy wolfsnail
X
7 (36.8%)
Birds
Acridotheres tristis*
Common myna
X
5 (26.3%)
Bubo virginianus
Great horned owl
-
1 (5.3%)
Circus approximans
Swamp harrier
X
4 (21.1%)
Pycnonotus cafer*
Red-vented bulbul
X
5 (26.3%)
Reptiles
Trachemys scripta*
Red-eared slider
X
4 (21.1%)
Mammals
Mus musculus*
House mouse
X
12 (63.2%)
Rattus exulans
Pacific rat
X
19 (100%)
Rattus norvegicus
Norway rat
X
13 (68.4%)
Rattus rattus*
Black rat
X
12 (63.2%)
VASCULAR PLANTS (N=35)
Family
Scientific name (synonyms)
Habit
Euphorbiaceae
Antidesma bunius
Tree
Myrsinaceae
Ardisia elliptica*
Small tree
X
2 (10.5%)
Moraceae
Castilla elastica
Tree
X
4 (21.1%)
Cecropiaceae
Cecropia peltata*
Tree
X
6 (31.6%)
Chrysobalanaceae
Chrysobalanus icaco
Small tree
X
5 (26.3%)
Rubiaceae
Cinchona pubescens*
Tree
X
0 (0%)
Hydrocharitaceae
Egeria densa
Aquatic herb
X
0 (0%)
Myrtaceae
Eugenia uniflora
Small tree
X
14 (73.7%)
Fabaceae
Falcataria (syn. Albizia) moluccana
Large tree
X
13 (68.4%)
Fabaceae
Flemingia strobilifera
Shrub
X
14 (73.7%)
Agavaceae
Furcraea foetida
Erect herb
X
7 (36.8%)
Crassulaceae
Kalanchoe pinnata
Erect herb
X
18 (94.7%)
Verbenaceae
Lantana camara*
Shrub
X
15 (78.9%)
Fabaceae
Leucaena leucocephala*
Small tree
X
19 (100%)
Convolvulaceae
Merremia peltata
Liana (woody vine)
X
8 (42.1%)
Poaceae
Melinis minutiflora
Grass
X
16 (84.2%)
Melsatomataceae
Miconia calvescens*
Small tree
X
3 (15.8%)
Asteraceae
Mikania scandens (syn. M. micrantha)*
Vine
X
0 (0%)
Mimosaceae
Mimosa diplotricha (syn. M. invisa)
Shrub
X
7 (36.8%)
Passifloraceae
Passiflora maliformis
Liana (woody vine)
X
11 (57.9%)
Passifloraceae
Passiflora rubra
Vine
-
1 (5.3%)
Passifloraceae
Passiflora suberosa
Vine
X
2 (10.5%)
Asteraceae
Pluchea symphytifolia
Shrub
X
4 (21.1%)
Myrtaceae
Psidium cattleianum*
Small tree
X
10 (52.6%)
Myrtaceae
Rhodomyrtus tomentosa
Small tree
X
0 (0%)
Rosaceae
Rubus rosifolius
Shrub
X
4 (21.1%)
Anacardiaceae
Schinus terebinthifolius*
Tree
X
0 (0%)
Araliaceae
Schefflera actinophylla
Tree
X
5 (26.3%)
Bignoniaceae
Spathodea campanulata*
Large tree
X
7 (36.8%)
Myrtaceae
Syzygium cumini
Tree
X
19 (100%)
Myrtaceae
Syzygium jambos
Tree
X
14 (73.7%)
Bignoniaceae
Tecoma stans
Small tree
X
9 (47.4%)
Polygonaceae
Triplaris weigeltiana
Large tree
X
0 (0%)
Fabaceae
Vachelia (syn. Acacia) farnesiana
Small tree
X
4 (21.1%)
Myrtaceae
Waterhousea floribunda
Tree
X
1 (5.3%)
*Listed among the “100 of the World’s Worst Invasive Alien Species” (Lowe, et al., 2000).
512
Tahiti
X
Surveyed islands (%)
0 (0%)
Meyer & Fourdrigniez: Islander perceptions of invasive species
Table 2 Number of surveyed islands, villages and people (adults) consulted during public meetings in the different
archipelagos of French Polynesia.
Archipelagos
Number of surveyed islands (names)
Leeward Is (Society Is)
Tuamotu Is
Austral Is
Marquesas Is
4 (Maupiti, Tahaa, Huahine, Bora Bora)
5 (Niau, Kaukura, Makatea, Tikehau, Rangiroa)
4 (Raivavae, Rimatara, Rurutu, Tubuai)
6 (Nuku Hiva, Ua Pou, Ua Huka, Hiva Oa, Fatu Iva,
Tahuata)
19
Total
Number of No of participants
villages
(adults)
9
494
9
479
10
414
13
394
41
1,781
Table 3 Number and density of invasive alien species (IAS) legally declared “a threat to biodiversity in French Polynesia” in
relation to geographic and demographic characteristics of islands, and plane transportation frequency or “connectivity”
with Tahiti: island with an international airport; islands with a domestic airport or airstrip; 2012 population census
(<www.ispf.pf>).
ARCHIPELAGO
(distance from
Tahiti in km)
SOCIETY
(170-310 km)
TUAMOTU
(220-350 km)
AUSTRAL
(500-700 km)
MARQUESAS
(1,000-1,500 km)
TOTAL
Island (number of flights
per week departing from
Tahiti)
Area (ha)
Population
(2012)
Tahiti
Tahaa (61 via Raiatea)
104,510
183,480
1.76
44 (96%)
0.04
Huahine (37)
9,020
7,480
5,220
6,303
0.58
0.84
28 (60.9%)
26 (56.5%)
0.31
0.35
Bora Bora (74)
2,930
9,598
3.27
26 (56.5%)
0.89
Maupiti (9)
1,140
1,223
1.07
19 (41.3%)
1.67
Rangiroa (20)
Makatea
7,900
2,567
0.32
10 (21.8%)
0.13
Niau (2)
2,950
2,100
68
226
0.02
0.11
15 (32.6%)
10 (21.8%)
0.51
0.48
Tikehau (10)
2,000
529
0.26
7 (15.2%)
0.35
Kaukura (2)
1,100
475
0.43
8 (17.4%)
0.73
Tubuai (14)
4,500
2,170
0.48
24 (52.2%)
0.54
Rurutu (12)
3,235
2,322
0.72
23 (50%)
0.71
Raivavae (7)
2,035
940
0.46
16 (34.8%)
0.79
Rimatara (5)
953
873
0.91
17 (36.9%)
1.78
Nuku Hiva (15)
33,950
2,967
0.03
23 (50%)
0.07
Hiva Oa (15)
31,550
2,184
0.07
26 (56.5%)
0.08
Ua Pou (9)
10,560
2,175
0.21
16 (34.8%)
0.74
8,340
621
0.07
14 (30.4%)
0.17
8,500
6,100
250,863
611
703
222,688
0.07
0.11
0.89
19 (41.3%)
13 (28.3%)
46 (100%)
0.22
0.21
0.02
Ua Huka (6)
Fatu Iva
Tahuata
20
which constitute demanding ecological conditions for
both introduced animals and plants. The Austral high
volcanic islands have a cooler climate due to their southern
geographical location (mean annual temperature between
18°C for Rapa Iti and 20°C for the other islands) which
may also prevent the establishment and invasion of some
“truly” tropical species. If the Austral islands are removed
from the analysis, the correlation coefficient is slightly
higher (R²=0.57).
Perceptions of invasive species in different
archipelagoos and islands
The total number of negative, positive and neutral
comments (50) recorded by participants for each species
was analysed for all the 19 surveyed islands. Comments
Population
density (/ha)
IAS number IAS density
(%)
(/ha)
were reported only for 15 of the 38 species occuring in
the islands, most of them were positive (Fig. 5). More
comments were made in the isolated islands of the Austral
Is (> 500–700 km from Tahiti) and the Marquesas Is (>
1,000–1,500 km) with lower socio-economic development
but where people seem to show a stronger interest in the
use of available natural resources (Fig. 4), compared to the
Leeward Is in the Society Is. Comments in the Tuamotu Is
were the lowest and the number of reported invasive species
is also the smallest (between 7 and 15 species, i.e. 15–33%
of the total). It is noteworthy that all comments made on
invasive species were positive in the Tuamotu atolls (Fig.
4), meaning they are more considered as “useful” for
people than “noxious/harmful”. In all surveyed islands
and archipelagos, positive comments exceeded negative
ones, but this rather surprising result might be biased as
513
Island invasives: scaling up to meet the challenge. Ch 3B Strategy: Collaboration
it is subjected to an active control programme in Tahiti to
protect the threatened endemic flycatcher Pomarea nigra
(Monarchidae) (Blanvillain, et al., 2003). For vascular
plants, the 29 invasive species were not considered as
“noxious/harmful” in all the surveyed islands where they
are present. There were many positive comments for
ornamental plants or fruiting trees, especially in the most
remote islands of the Austral and the Marquesas (Table 4).
It is interesting to note that the perceived status of
invasive alien species differs from one archipelago to
another, but also among islands in the same archipelago,
such as the climbing liana Passiflora maliformis in the
Austral Is because of its edible fruits or the large tree
Falcataria moluccana in the Marquesas as a timber tree
(Table 4). Both species are currently being controlled in
areas of high conservation values in Tahiti.
DISCUSSION
Island invasibility, species invasiveness and socioeconomic development
Perception of invasiveness is complex because of
diverse mental representations by different key interest
groups and socio-economic contexts (Garcia-Llorente,
et al., 2008). An understanding of human dimensions is
necessary to avoid potential social conflicts in invasive
species management (Estévez, et al., 2014; Russell, 2014).
Fig. 3 A. Relationship between the number of invasive alien
species and the number of inhabitants (2012 population
census) in the 19 surveyed islands (Spearman test,
P-value = 0.001407). B. Relationship between the number
of invasive alien species and the plane transportation
frequency (number of flights per week from Tahiti) in the
16 surveyed islands with a domestic airport.
most people agreeing with the invasiveness status did not
make specific negative comments (e.g. for the three species
of rats – Rattus spp.). To avoid this bias towards positive
comments, future studies should explicitly ask participants
for their inputs on the ecological and socio-economical
impacts of the targeted invasive species.
One animal species, the common myna (Acridotheres
tristis), has received only positive comments. This bird,
first introduced to Tahiti in the early 1900s (Meyer, 2003)
is indeed considered as a useful animal because it eats
introduced wasps and ticks especially in the Leeward
Islands of the Society archipelago (e.g. in Huahine), whereas
Fig. 4 Percentage of the positive, neutral and negative
comments for the invasive alien species recorded in the
19 surveyed islands.
514
Our results conducted on small islands of French
Polynesia show that the number of invasive alien species
is not decreasing with island remoteness (i.e. distance
from Tahiti) and island size, but is more correlated with
human development (e.g. the number of inhabitants and
the frequency of transportation connection with Tahiti)
and habitat diversity, as documented in other islands
elsewhere (Kueffer, et al., 2010). The island of Tahiti
can be considered as a “transportation hub” in the South
Pacific, with an international airport opened in 1960 and
direct flight connections to Rarotonga (Cook Is), Australia,
New Zealand, New Caledonia, California and Hawaii
(USA), Chile and Japan; and a major trade port in 1962
with goods imported from Europe, North and South
America and South-east Asia. The increasing development
of commercial trade during the past decades (from 330,000
tons in 1989 to 980,000 tons in 2015, ISPF, 2016) was
associated with a dramatic increase of accidental plant
and animal introductions. Invasive insects such as fruit
flies (Bactrocera spp., Tephrididae), the glassy-winged
sharpshooter (Homalodisca vitripennis, Cicadellidae) and
the little fire ant (Wasmannia auropunctata, Formicidae),
first introduced to Tahiti between the 1970s and the 1990s
(Meyer, 2003), have subsequently spread to many other
Fig. 5 Percentage of the positive, neutral and negative
comments for the invasive alien species recorded in the
19 surveyed islands.
Meyer & Fourdrigniez: Islander perceptions of invasive species
Table 4 Examples of positive and negative comments for some invasive alien plants introduced by Europeans in the
surveyed islands with their date of first introduction or record in Tahiti, French Polynesia (Baas Becking, 1950; Jacquier,
1960).
Scientific name
Positive comment(s)
Eugenia uniflora
Falcataria
(syn. Albizia)
moluccana
Flemingia
strobilifera
Furcraea foetida
Lantana camara
Negative
comment(s)
Island(s)
Edible fruits, wood used for Rimatara, Tubuai
fish tools
Honeybee-forage plant,
Fatu Iva,
wood used for boats
Raivavae
Alters feral
goat meat
Dries out
rivers
Fatu Iva
Rurutu
1936
Flower used in necklaces
Spreads in
gardens
-
Tahuata
1937
-
?
-
-
1853
-
-
1845
Suppresses
orange and
coffee trees
-
Fatu Iva,
Tubuai
Formerly used for ropes &
traditionnal dance skirts
Ornamental garden plant
Leucaena
leucocephala
Passiflora
maliformis
Forage for cattle, improves
soil erosion control
Edible fruits used for jams
Syzygium cumini
Edible fruits
Syzygium jambos
Edible fruits
Island(s)
Nuku Hiva, Ua
Huka, Rimatara
Rimatara,
Rurutu, Tubuai
Nuku Hiva, Ua
Huka
Nuku Hiva Ua
Huka
Rimatara
Tikehau,
Makatea
Tubuai
French Polynesian islands through inter-island boat and/or
plane transportation.
The perceived status of the 46 legally declared invasive
species, a small subset of the total number of invasive
species in French Polynesia (e.g. with more than 80 plants
considered as invasive, Fourdrigniez & Meyer, 2008),
differs from one archipelago to another, or even among
islands in the same archipelago. They are more positively
considered in the most isolated islands with lower socioeconomic development and/or where natural resources
are extremely limited, e.g. in atolls where invasive woody
plants are used as tools or for wood construction, such
as Leucaena leucocephala. This is very similar to the
different attitudes of urban versus rural residents to pest
species management in western developed countries or in
Australia and New Zealand in the Pacific region (Fraser,
2001; Johnston & Marks, 1997). When abundant, invasive
alien species are often seen as potential natural resources
by islanders whereas when they are less common or rare,
people agreed to eradicate introduced species. Species
prioritisation that includes socio-economic values may thus
contribute to a better efficiency in control or eradication by
gaining support of local communities in remote islands.
Importance of cultural values
Human perceptions and attitudes vary with time,
places, societies, economic conditions and culture (Dalla
Bernardina, 2010; Fitzgerald, et al., 2007). The importance
of cultural (traditional or ancestral) values of introduced
species in the Pacific islands is well illustrated by animal
species that were introduced by the first humans during
their migration and colonisation, and became invasive
with time, with sometimes dramatic impacts on the native
biodiversity. Feral pigs (Sus scrofa) are still a source of
dispute between conservationists and native Hawaiians
who hunt them as in the past (Van Driesche & Simberloff,
2016), and Pacific rats (Rattus exulans) are considered a
treasure brought to New Zealand by their Maori ancestors,
thus may be worshipped and of high significance (Haami,
1994; Veitch & Clout, 2000). Some plants introduced
-
Date of first
introduction or
record
1848
?
1880
1890
by the first Polynesians for ritual, aesthetic or utilitarian
values (Whistler, 2009) have also spread into native
lowland forests in French Polynesia and Hawaii, including
the candlenut tree (Aleurites moluccana, Euphorbiaceae)
and the bamboo (Schizostachyum glaucifolium, Poaceae)
where they are considered as either invasive (Smith, 1985)
or part of the Polynesian social heritage (Larrue, et al.,
2010).
Our survey indicates that the date of species introduction
in the islands of French Polynesia, more particularly in
Tahiti (Baas Becking, 1950; Jacquier, 1960), seems to be
an important factor explaining attitudinal differences, as
old introduced species seem to be more widely accepted or
positively considered by people, because of their long coexistence (more than one century). This is the case of the small
tree Leucaena leucocephala and the shrub Lantana camara
which were introduced by Europeans in French Polynesia
in 1845 as a fodder plant and 1853 as an ornamental garden
plant respectively, and often still considered as beneficial
species (Table 4). This phenomenon is sometimes, but
incorrectly, called “indigenisation”, as these naturalised
species (“naturalisation” is defined as an ecological proces
where the alien plant species establishes and becomes
incorporated within the natural flora, Richardson, et al.,
2000) are not becoming indigenous or native but part of the
human culture or natural heritage. It should be refered to
as “heritagisation” (“patrimonialisation” in French) which
describes a socio-cultural, legal or political process where
an area, a good or a species is transformed into an object of
the natural, cultural or religious heritage with conservation
or restoration value.
One of the crucial challenges in invasive species
management is the active involvement, engagement and
support of local communities (Hart & Larson, 2014), as
well as resolving or at least avoiding potential conflicts of
interest between different stakeholders. The small Pacific
islands, including French Polynesia, provide an excellent
ground for testing new methodologies and initiatives in
complex insular societies. Based on the results of this
survey, we propose that an “invasive species perception
515
Island invasives: scaling up to meet the challenge. Ch 3B Strategy: Collaboration
index” should be included in feasibility studies to manage
biological invasions in isolated inhabited islands.
A first step to integrate the local socio-economic and
cultural dimensions of invasive species in the islands of
French Polynesia was the creation of a network during
and following this survey (called “Te Rau Mata Arai” in
Tahitian, literally the “numerous watchful eyes”). Its aims
are the prevention, detection, surveillance and control of
invasive alien species by identifiying local, key people
in each island (a total of 36 on the 19 surveyed islands)
including local government and city council representatives,
members of nature protection groups, small entrepreneurs,
and other civil society actors.
ACKOWLEDGEMENTS
We are grateful to all the participants of the survey
conducted in the 19 small islands of French Polynesia
in 2014, the city councils (“mairies”) and the local
representatives of the Service du Développement Rural
(Department of Agriculture of the Government of French
Polynesia) for organising the public meetings. We also
thank the Délégation à la Recherche and the Direction de
l’Environnement (Government of French Polynesia) for
the moral and funding support, Charles Chimera (HawaiiPacific Weed Risk Assessment, Hawaii, USA) for revising
the English, Robin Pouteau (Institut Agronomique néoCalédonien, Nouméa, New Caledonia) for performing
the statistical tests, and two anonymous reviewers for
their constructive comments that helped to improve the
manuscript.
REFERENCES
Baas Becking, L.G.M. (1950). Liste Préliminaire de Plantes Introduites
à Tahiti. Nouméa, Nouvelle-Calédonie: Commission du Pacifique Sud,
Document Technique No. 7.
Bhattacharya, A. and Kumar, P. (2010). ‘Water hyacinth as a potential
biofuel crop’. Electronic Journal of Environmental, Agricultural and
Food Chemistry 9(1): 112–122.
Blanvillain, C., Salducci, J.-M., Tuturai, G. and Maeura, M. (2003).
‘Impacts of introduced birds on the recovery of the Tahiti flycatcher
(Pomarea nigra)’. Biological Conservation 109(2): 197–205.
Bremner, A. and Park, K. (2007). ‘Public attitudes to the management of
invasive non-native species in Scotland’. Biological Conservation 139:
306–314.
Dalla Bernardina, S. (2010). ‘Les Invasions Biologiques sous le Regard
des Sciences de l’Homme’. In: R. Barbault and M. Atramentowicz
(coord.) Les Invasions Biologiques, une Question de Natures et de
Sociétés, pp. 65–108. Versailles, France: Editions Quae.
Estévez, R.A., Anderson, C.B., Pizzaro, J.C. and Burgman, M.A. (2014).
‘Claryfying values, risk perceptions, and attitudes to resolve or avoid
social conflicts in invasive species management’. Conservation Biology
29(1): 19–30.
Fischer, A., Selge, S., van der Wal, R. and Larson, B.M.H. (2014). ‘The
public and professionals reason similarly about the management of nonnative invasive species: A quantitative investigation of the relationship
between beliefs and attitudes’. PLOS One 9(8): e105495.
Fitzgerald, G., Fitzgerald, N. and Davidson, C. (2007). Public Attitudes
Towards Invasive Animals and Their Impacts. Canberra, Australia:
Invasive Animals Cooperative Research Centre.
Fourdrigniez, M. and Meyer, J.-Y. (2008). Liste et Caractéristiques
des Plantes Introduites Naturalisées et Envahissantes en Polynésie
Française. Papeete, Tahiti: Contribution à la Biodiversité de Polynésie
française No. 17, Délégation à la Recherche, Papeete (Unpublished
report).
Fourdrigniez, M., Taputuarai, R., Vivier, R. and Homo, D. (2014). Guide
de Protection des Îles contre les Espèces Envahissantes en Polynésie
française. Papeete, Tahiti: STP Multipress.
Fraser, W. (2001). Introduced Wildlife in New Zealand: A Survey of
General Public Views. Landcare Research Science Series No. 23.
Canterbury, New Zealand: Manaaki Whenua Press.
Garcia-Lorente, M., Martin-Lopez, B., Gonzalez, J.A., Alcorlo, P. and
Montes, C. (2008). ‘Social perceptions of the impacts and benefits
of invasive alien species: Implications for management’. Biological
Conservation 141: 2969–2983.
516
Haami, B. (1994). ‘The Kiore Rat in Aotearoa: a Maori Perspective’. In:
J. Morrison, P. Geraghty and L. Crowl (eds.) Fauna, Flora, Food and
Medecine, Science of Pacific Island People, Volume 3, pp. 65–76. Suva,
Fidji: Institute of Pacific Studies.
Hart, P.S. and Larson, B.M.H. (2014). ‘Communicating about invasive
species: How “driver” and “passenger” model influence public
willingness to take action?’ Conservation Letters 7(6): 545–552.
Hill, R.L. (1989). ‘Environmental Protection Procedures and the
Biological Control Programme against Gorse in New Zealand’. In: E.S.
Delfosse (ed.) Proceedings of the VIIth International Symposium on
Biological Control of Weeds, 6–11 March 1988, pp. 127–133. Rome,
Italy.
IEOM (Institut d’Emission d’Outre-Mer) (2017). Polynésie française.
Rapport Annuel 2016. Papeete, Tahiti: STP Multipress.
ISPF (Institut de la Statistique de la Polynésie Française) (2016). Polynésie
Française en Bref 2016. French Polynesia at a Glance <www.ispf.pf>
Accessed 10 June 2017.
Jacquier, H. (1960). ‘Enumération des plantes introduites à Tahiti
depuis la découverte jusqu’en 1885’. Bulletin de la Société des Etudes
Océaniennes 130: 117–146.
Johnston, M.J. and Marks, C.A. (1997). Attitudinal Survey on Vertebrate
Pest Management in Victoria. Melbourne, Australia: Report Series No.
3.
Kueffer, C., Daehler, C.C., Torres-Santana, C.W., Lavergne, C., Meyer,
J.-Y., Otto, R. and Silva, L. (2010). ‘A global comparison of plant
invasions on oceanic islands’. Perspectives in Plant Ecology, Evolution
and Systematics 12: 145–161.
Larrue, S., Meyer, J.-Y. and Chiron, T. (2010). ‘Anthropogenic vegetation
contributions to Polynesia's social heritage: the legacy of candlenut
tree (Aleurites moluccana) forests and bamboo (Schizostachyum
glaucifolium) groves on the island of Tahiti’. Economic Botany 64(4):
329–339.
Lorence, D.H., Wood, K.R., Perlman, S.P. and Meyer, J.-Y. (2016).
‘Flore Vasculaire et Végétation des Îles Marquises: Caractéristiques,
Originalités et Vulnérabilité’. In: R. Galzin, S.-D. Duron and J.-Y. Meyer
(eds.) Biodiversité terrestre et marine des îles Marquises, Polynésie
française, pp. 311–336. Paris, France: Société Française d’Ichtyologie.
Lowe, S., Browne, M., Boudjelas, S., and De Poorter, M. (2000). 100 of
the World’s Worst Invasive Alien Species. A Selection from the Global
Invasive Species Database. The Invasive Specialist Group (ISSG) of the
Species Survival Commission (SSC) of the World Conservation Union.
IUCN.
Marshall, N.A., Friedel, M., van Klinken, R.D. and Grice, A.C. (2011).
‘Considering the social dimension of invasive species: The case of
buffel grass’. Environmental Management 14: 327–338.
McNeely, J.A. (ed.). (2000). The Great Reshuffling: Human Dimensions
of Invasive Alien Species. Gland, Switzerland: IUCN.
Meyer, J.-Y. (2003). ‘French Polynesia’. In: C. Shine, J.K. Reaser and
A.T. Gutierrez (eds.) Invasive Alien Species in the Austral-Pacific
Region, pp. 22–34. Cape Town, South Africa: National Reports and
Directory of Resources. The Global Invasive Species Programme.
Richardson, D.M., Pysek, P., Rejmanek, M., Barbour, M.G., Panetta, F.D.
and West, C.J. (2000). ‘Naturalization and invasion by alien plants:
Concepts and definitions’. Diversity and Distribution 6: 93–107.
Russell, J.C. (2014). ‘A comparison of attitudes towards introduced
wildlife in New Zealand in 1994 and 2012’. Journal of the Royal
Society of New Zealand 44(4): 136–151.
Selge, S., Fischer, A. and van der Wal, R. (2011). ‘Public and professional
views on invasive non-native species – a qualitative social scientific
investigation’. Biological Conservation 144(12): 3089–3097.
Smith, C.W. (1985). ‘Impacts of Alien Plants on Hawaii’s Native Biota’.
In: C.P. Stone and J.J. Scott (eds.) Hawaii’s Terrestrial Ecosystems:
Preservation and Management, pp. 180–250. Honolulu, Hawaii, USA:
University of Hawaii Press.
Van Driesche, R.G. and Simberloff, D. (2016). ‘Benefit-Risk Assessment
of Biological Control in Wildlands’. In: R.G. Van Driesche, D.
Simberloff, B. Blossey, C. Causton, M.S. Hoddle, D.L. Wagner, C.O.
Marks, K.M. Heinz and K.D. Warne (eds.) Integrating Biological
Control into Conservation Practice, pp. 80–104. Chichester, UK: John
Wiley and Sons, Ltd.
Veitch, C.R. and Clout, M.N. (2000). ‘Human dimensions in the
management of invasive species in New Zealand’. In: J.A. McNeely
(ed.) 2000. The Great Reshuffling: Human Dimensions of Invasive Alien
Species, pp. 63–71. Gland, Switzerland: IUCN.
Warner K.D. and Kinslow, F. (2013). ‘Manipulating risk communication:
Value predispositions shape public understanding of invasive species
science in Hawaii’. Public Understanding of Science 22: 203–216.
Whistler, W.A. (2009). Plants of the Canoe People: An Ethnobotanical
Voyage through Polynesia. Honolulu, Hawaii, USA: Island Press.
C. Stringer, S. Boudjelas, K. Broome, S. Cranwell, E. Hagen, G. Howald, J. Kelly, J. Millett, K. Springer and K. Varnham
Stringer, C.; S. Boudjelas, K. Broome, S. Cranwell, E. Hagen, G. Howald, J. Kelly, J. Millett, K. Springer and K. Varnham.
Married bliss and shotgun weddings: effective partnerships for island restoration
Married bliss and shotgun weddings: effective partnerships
for island restoration
C. Stringer1, S. Boudjelas2, K. Broome3, S. Cranwell4, E. Hagen5, G. Howald5, J. Kelly1, J. Millett6,
K. Springer1 and K. Varnham1
Royal Society for the Protection of Birds, The Lodge, Potton Road, Sandy, Bedfordshire, SG19 2DL, United
Kingdom. <clare.a.stringer@gmail.com>. 2 Pacific Invasives Initiative, School of Biological Sciences, The University
of Auckland, Private Bag 92019, Auckland 1142, New Zealand. 3 Department of Conservation, Private Bag 3072,
Hamilton 3240, New Zealand. 4 BirdLife International, Pacific, 10 MacGregor Road, Suva, Fiji. 5 Island Conservation,
2100 Delaware Ave., Suite 1 Santa Cruz, CA 95060 USA. 6 Independent practitioner, C/o Programme Coordination
Unit, PO Box 310, Victoria, Mahe, Seychelles.
1
Abstract Island restoration is expanding as a tool for enhancing conservation outcomes. The ability of conservation
managers to eradicate multiple invasive species over increasingly complex or large areas is steadily improving, but
progress inevitably presents new challenges. There is always a larger, even more complex operation ahead, and islands
with human populations present their own suite of problems and opportunities. The majority of the large and/or complex
island restoration projects to-date have been carried out with a high level of government commitment including funding
support (e.g. New Zealand, Australia, USA). However, there is increasing interest in applying this methodology to islands
in jurisdictions where there is less central government support. This can be further complicated by regulatory systems and
implementation logistics. Non-governmental organisations are now taking lead roles in many projects to restore islands
worldwide, working collaboratively to share the financial, logistical and regulatory challenges and share in the outcomes.
If we are to succeed in truly “turning the tide” on invasive species it will be necessary for governmental and nongovernmental organisations to partner even more effectively in order to expand the capacity for such conservation actions
worldwide. Choosing the right partners, clarifying shared values, programme goals, responsibilities and definitions of
success is needed for NGOs, governments and other partners to work effectively and make the progress that is necessary
to continue achieving good conservation outcomes in the future.
Keywords: collaboration, eradication, organisational structure, planning, project management
INTRODUCTION
“Cross sector partnerships ... are far from commonplace.
True partnerships are the stuff of legends. Think of the
Fellowship of the Ring...” Tennyson (2011).
The eradication of invasive vertebrates, especially
rodents, as a key component of island restoration has an
extensive history spanning more than 50 years (e.g. Towns
& Broome, 2003; Howald, et al., 2007; Towns, et al., 2013;
Russell & Broome, 2016). Techniques for carrying out
this work have developed over this time, and eradication
projects are now often highly complex and specialised
operations using equipment and people from all over the
world.
Until the last fifteen years, the agencies with the
resources to undertake the largest projects were generally
government conservation agencies (GCAs) in developed
countries such as New Zealand, Australia, the USA and
Canada. Projects included the consultation of stakeholders,
but early operations were often led and managed by single
organisations on government land (Towns & Broome,
2003).
As the case for carrying out island restoration projects
has become more established (Courchamp, et al., 2003;
Bellingham, et al., 2010; Jones, et al., 2016), new
organisations, especially non-governmental organisations
(NGOs) and for-profit enterprises (FPEs) have become
increasingly involved in island restoration. There are now
several NGOs and initiatives worldwide that are entirely
dedicated to the restoration of islands through the removal
of invasive species (e.g. Island Conservation (USA);
Grupo de Ecología y Conservación de Islas (Mexico);
Predator Free New Zealand (New Zealand)), and this work
is gaining in prominence within the wider conservation
NGO community (e.g. recent work by the Royal Society
for the Protection of Birds (UK) and the South Georgia
Heritage Trust). In addition, new commitments to carry out
this work have been made by international organisations
and through international agencies, e.g. the International
Union for the Conservation of Nature (IUCN)’s Honolulu
Challenge (IUCN, 2016), BirdLife International’s Invasive
Species Programme (BirdLife International, 2017).
The increasing challenge of partnerships
In order to continue to obtain the conservation benefits
available from the eradication of invasive species on islands,
projects will need to be carried out in even more complex
conditions. In light of current technology and experience,
we could describe a project where area is less than 10,000
ha; there is a single owner; country jurisdiction is clear; and
a single funding source is available as “simple”. In some
countries all, or the majority of these projects have now
been tackled, or the need for them has not arisen (Howald,
et al., 2007; Dawson, et al., 2015; Parkes, et al., 2017;
Stanbury, et al., 2017). In less developed countries where
conservation funding is much scarcer, the idea of national
governments supporting island restoration projects is often
not well established and finding funds for any project of
this sort is difficult. In many countries, islands without
human habitation or regular use are extremely uncommon.
Islands with significant human populations, complex
and challenging topography, and/or located in extremely
remote parts of the world are thus becoming a higher
priority (Oppel, et al., 2011, Dawson, et al., 2015; Parkes,
et al., 2017; Stanbury, et al., 2017). For example, the Royal
Society for the Protection of Birds (RSPB) is currently
developing a project to restore Gough Island (Tristan
da Cunha) through the eradication of house mice (Mus
musculus).
The island is extremely remote and is located in a UK
Overseas Territory with a small human population and
insufficient financial resources to support the operation.
The project partnership will include no fewer than six
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 517–521. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
517
Island invasives: scaling up to meet the challenge. Ch 3B Strategy: Collaboration
project partners from at least three countries, including
government agencies, NGOs, and the local community.
The DIISE (2017) records 25 rodent eradication
attempts that have been made on islands greater than one
square kilometre in area since 2010. Of these, the majority
(15 of the 25) were not undertaken solely by government
agencies, and even where operations were government-led,
some sort of partnership was needed (e.g. between State
and Federal government agencies).
Howald, et al. (2011) explored the advantages and
challenges of different organisational structures conducting
island restoration projects. The authors found that there
were clear advantages and disadvantages attached to GCAs,
NGOs and FPEs in conducting eradication campaigns, but
concluded that the potential advantages of collaboration
were often greater than the challenges. In this paper, we
consider local community groups separately from NGOs
as another type of organisation which is increasingly
proposing and supporting new island restoration projects.
As well as type, the size and culture of organisations
also has significant importance and impacts on internal
bureaucracy, speed of decision-making and level of
tolerance for risk.
Whilst it is apparent that partnerships provide
opportunities to capitalise on the strengths and compensate
for the weaknesses of different types of organisations,
partnerships can be complicated to establish and maintain.
The same people who have significant strengths and
experience in designing and implementing island
eradication projects do not always have a similar level
of experience or expertise in developing or maintaining
organisational partnerships, especially when organisations’
cultural aspects can be highly variable. Staff turnover can
be an issue, as partnerships are effectively formed between
individuals as well as organisations, and some organisations
have higher turnover than others. It is important for the
organisations that are planning and managing eradication
projects to recognise the importance of consistency of
staffing in these projects, and endeavour to provide this,
as well as supporting training in partnership-working for
technical staff wherever possible.
This paper assesses some of the factors that may
be influential in making partnerships work. There is no
way to carry out a scientific analysis of how to create a
strong partnership that will lead to a successful project
outcome: partnerships (like marriages) are not a scientific
construct. However, the authors of this paper have been
involved in a wide range of projects with partners from
government, NGOs, and local communities. From our
combined experience, the main elements needed, in our
opinion, are presented, along with some of the common
pitfalls. Sharing our experiences may enable other project
managers to analyse their own potential partnerships, and
hopefully use these principles to enhance the likelihood of
project success.
What partnerships are, and when they should be
established
According to Wilcox (1998) a partnership is an
agreement between two or more individuals or groups
to work together to achieve common aims. Sterne, et al.
(2009) identified nineteen characteristics of partnerships,
including mutual trust and respect, clearly identified roles
and responsibilities, transparency of decision-making,
and a process for adjudicating disputes. The Nature
Conservancy (2017) suggests there are six stages to most
partnerships, and Tennyson (2011) defines twelve stages.
The main points are:
● Prepare: define the need for partners.
● Select: choose the best partner(s) to work with.
518
● Negotiate: create agreement to inspire action and
reduce the potential for conflict.
● Manage: implement joint work.
● Measure: monitor and evaluate the partnership.
● Conclude: adapt, improve or conclude the partnership.
During the first stage, it is important for project
managers to consider carefully whether forming a
partnership is the best choice in their individual situation.
Reasons to establish a project partnership include the
desire to increase capacity amongst other organisations and
stakeholders; the need to access a new decision-making
authority or constituents; the opportunity to share costs; and
the ability to make projects more sustainable and resilient.
One method of comparing these potential benefits with the
potential costs would be a SWOT (Strengths, Weaknesses,
Opportunities, Threats) analysis.
It is advisable to establish partnerships early in the
project planning process, if indeed it is considered that a
partnership would be beneficial. The Nature Conservancy
(TNC) recommends a partner scoping exercise at the
start of planning any conservation project (The Nature
Conservancy, 2017). However, this step is not currently
included in resources specific to island restoration, such as
the Pacific Invasives Initiative (PII) Resource Kit (Pacific
Invasives Initiative, 2011). In particular, there appear to be
significant benefits from involving community partners,
including landowners, at the initial stages of project
planning (Varnham, 2011; McClelland, et al., 2011), not
least because their local knowledge can add value to
planning and their drive to succeed can assist in motivating
the rest of the partnership.
Potential partners could be identified in a stakeholder
analysis which may be carried out as part of the feasibility
study for an operation, e.g. step 2.1 of the PII Resource
Kit (Pacific Invasives Initiative, 2011), or through a
scoping exercise (The Nature Conservancy, 2017; Flora
and Fauna International, 2009). However, it is likely
that if external funds are to be sought for an eradication,
partners may need to be identified even sooner than this.
As part of considering the composition of a partnership, it
can be useful to consider the implications of excluding a
particular organisation or group and how this could affect
the outcome and the other partners. For example, excluding
local people from a partnership could lead to mistrust from
funders and external agencies as well as the community
themselves, or even prevent the project from going ahead.
Often, there may be little or no choice about who to
work with, for example it is often necessary to work with
a local government agency, or the island owner. In some
cases, they could be reluctant partners at first, but may
become more engaged when they see the benefits of the
relationship. This engagement could take a long time to
achieve, and in some cases may never be possible. In other
cases, partners may be willing, but there may be high costs
connected to their involvement. Thinking ahead about
the costs and benefits will help in considering whether a
partnership is appropriate, and in minimising the costs and
maximising the benefits (Flora and Fauna International,
2009).
The level of intensity of partnership that is desired
should also be considered. Some partnerships are shortterm, and relatively informal relationships, whereas others
may develop into strategic long-term, organisation-wide
relationships. It is also possible for any partnership to break
down before its objectives have been achieved. A plan for
how to deal with such a break should be included in the
partnership agreement or memorandum of understanding
(see discussion below).
Stringer, et al.: Effective partnerships for island restoration
Key elements of strong partnerships
Although there is a diverse range of organisations
involved in island restoration work worldwide, the
elements required for partnerships are the same. It has been
suggested that the key principles of equity, transparency,
and mutual benefit should apply to most partnerships
(Tennyson, 2011). These principles provide a foundation
on which the partnership can operate. In order to build
this foundation, those forming new partnerships should
consider the following in particular: good communication;
clearly defined roles; appropriate leadership, staffing and
personalities; clear, shared vision and expectations; and
clarity over funding and resource issues (Wilson, 2005;
Tennyson, 2011; Ozarski, 2015).
One basic tool that most partnerships use is a written
partnership agreement (also referred to in some cases as
a memorandum of understanding or memorandum of
partnership). Partnership agreements can vary widely in
their level of detail depending on the complexity and aims
of the partnership and the degree of formality but should, at
a bare minimum, define the boundaries of the partnership.
Partnership agreements should cover the areas set out
below and may include others, depending on the specific
needs of the project.
As previously noted, one of the main characteristics
of partnerships is that the partners are working towards
a common aim. Where a group or individual has an
interest in delivery of an island restoration project but is
not as committed to the same goal as others, a partnership
should not be formed – this could lead to confusion and
frustration. The project team should seek to maintain a
good relationship with such stakeholders but should not
force an inappropriate partnership.
Good communication
When partnerships do not work, poor communication is
often blamed. Effective communication is essential to move
projects forward, especially in partnerships where partners
may have different motivations and expectations. Setting
out a shared communication strategy is recommended
within partnerships. General principles of communication
for partnerships (after The Nature Conservancy, 2017)
include:
● be timely in communication;
● brainstorm new issues;
● be consultative, not dictatorial;
● be flexible;
● document agreements and plans, and revisit, adjust
and adapt as the situation changes;
● a policy of “no surprises”.
In addition to these principles, it is important to respect
cultural and organisational differences and challenges
when communicating. Communication methods should be
adapted to suit each partner organisation’s strengths and
weaknesses, e.g. emailing high resolution newsletters to
communities with limited internet access is not effective
communication.
Partnerships may be formed for many reasons. These
may include the development of fundraising support,
advocacy, avoiding bureaucracy, the need for landowner
and resident buy-in and support, resource sharing, research,
provision of expertise, and to enable different organisations
to gain project experience, perhaps building towards
their own projects in the future. All of these reasons are
legitimate; however, it is important that all partners are
clear about their own and other partners’ motivation, and
the scope of each partner’s involvement. It is only ongoing
communication that will enable this clarity. At times, some
partners may also need to operate transparently outside
the scope of a partnership, for example, government
agencies which may also have a regulatory role. If the
scope of each partnership has been clearly established and
communication is clear, this should be possible.
It is important to include positives in communication.
Even though project planning and implementation is
challenging, project teams benefit from celebrating
successes, recognising achievements, and saying “thanks”.
It is important to ensure dispersed partners are all able to
take part in celebrations and reflect on the achievements
and progress being made.
It is also important to consider the way in which a
project will communicate itself externally. Publicity and
“branding” can often be stumbling blocks in partnerships,
and many projects have developed their own brands,
independent of the partner organisations (e.g. the Isles
of Scilly Seabird Recovery Project). Early decisions on
shared messages and how to acknowledge partners and
supporters can help to avoid issues later on.
Clearly defined roles and responsibilities
It is extremely important that the partners in any island
restoration project understand their respective roles and
what is expected of them. The roles and responsibilities
of different organisations will vary over the course of the
project and it is important that all partners understand how
their roles and those of other partners will change over
time. This is particularly important in the planning and
post-operational phases of the project where roles may be
less obvious.
Communities, local and non-local NGOs, and local
and non-local government agencies may all have a role.
In federal systems, it may also be necessary to involve
different levels of government (e.g. the Macquarie Island
pest eradication project involved both the Tasmanian State
Government and the Australian Federal Government,
or in the UK Overseas Territories where local and UK
governments may play a role). In the Macquarie Island
pest eradication project, the funding agreement between
the Tasmanian and Australian Federal governments
outlined that funding was joint, but that implementation
was a Tasmanian government responsibility. Based on
that agreement there was no confusion over operational
roles. Without such clarity, multiple partners and
stakeholders may perceive themselves to have decisionmaking authority leading to confusion and potentially to
operational difficulties.
In order to minimise this sort of confusion, most
projects develop some sort of partnership agreement or
Memorandum of Understanding. This may be legally
binding in some partnerships. However, even if roles are
clearly set out in writing, it should not be assumed that all
project partner staff who are participating in meetings or in
project teams are aware of these roles, and they may need
to be reiterated and revisited many times. Templates for
developing project partnership agreements are available
from Flora and Fauna International (2009), The Nature
Conservancy (2017) and in the Partnering toolbook
(Tennyson, 2011). None of these are specific to island
restoration projects, and it would be useful if practitioners
could develop and share resources in this area in the future.
Sometimes the project plan can be effective as
a partnership agreement. All projects should have a
clear plan which clearly describes the agreed roles and
responsibilities of all partners. A good reason to have a
partnership agreement is when a partnership is likely to go
beyond the scope of a single project.
It is also important to set out the roles and responsibilities
for the various advisory and steering groups that will be
519
Island invasives: scaling up to meet the challenge. Ch 3B Strategy: Collaboration
developed within most partnership projects, each of which
may involve a subset of the project partners. Terms of
reference for these groups have been developed by many
projects but, as with partnership agreements, they are not
yet generally shared. Developing such a resource would be
of great use to future projects, and this could be hosted on
websites such as those of the Pacific Invasives Initiative
(section now in development) or the Great Britain Nonnative Species Secretariat.
Leadership, staffing and personalities
As with any endeavour, the people involved are a
major factor that will lead to the work being enjoyable and
effective, and the success of any island eradication project
will largely depend on the skills, dedication, and attitude
of the team involved. If a project is being carried out by
a single agency, then that agency can recruit a team made
up of the most suitable/experienced people available and
although there may be some personality issues during the
project implementation period, it should be possible to
manage these as part of normal business practices.
However, when working in partnership, there can
be pressure to include representatives from different
organisations in teams despite substantial differences in
experience and culture. In addition, the representatives
of the partner organisations who should meet regularly to
discuss project planning and implementation may have
differing views and experiences. If one organisation/
individual has been leading/driving the project for some
time prior to implementation, they may have particularly
strong opinions on how things should be done, and this
could lead to clashes with others.
Personality clashes and power struggles can derail
an operation, and lead to breakdowns in organisational
relationships if left unaddressed. Whilst large organisations
may have the capacity to replace staff that are causing
difficulties within a partnership, smaller organisations
may not have this ability. Individuals who may be skilled
at motivating others and providing leadership within a
project may not always be best suited to developing and
maintaining complex project partnerships and vice versa.
Most partnerships require more management time
than anticipated. If a particularly wide partnership is
necessary, it may be necessary to bring in new staff
whose job is primarily to service the partnership in terms
of communications and logistics. Organisations that
are planning large projects with complex partnerships
should consider recruiting personnel who have skills and
experience in this area and can complement the technical
skills of the operational management team.
If possible, organisations partnering in island restoration
projects should aim to involve more than one staff member
in each project so that there is a chance to review decisions
and assess how the partnership is developing, and to assist
continuity in case of staff turnover. If problems arise, each
partner organisation should have a clear understanding
of how they can raise concerns and address them at an
early enough stage to avoid a complete breakdown of the
partnership and potentially of the whole project.
Clear, shared expectations
As discussed above, organisations may enter into
partnerships for a variety of reasons and with a variety of
expected benefits. Additionally, motivations for wanting
to be involved may be very different, even if the ultimate
goal is shared. At the start of the partnership relationship,
organisations should work to establish their shared goals
and vision for the work to be undertaken. They should
develop a project plan or agree the process that will be
used to develop this. They should set out guidelines for
decision-making and what will happen in the event of
520
disagreements. It is also helpful to agree on a formal
grievance process before a dispute emerges.
One area where there seems to be particular potential
for a mis-match of expectations is in relation to preeradication preparation. It is important to make clear plans
regarding who will make an island “ready” for eradication,
e.g. track cutting, removal of waste, informing residents,
leading on any research, etc., as well as who will fund
this work. Sometimes preparation can take a long time,
and it can be difficult maintaining enthusiasm and energy
throughout this phase.
There must also be clear expectations about how the
partnership will move on, in the event of either success or
failure of the planned eradication operation. Partners should
be clear on: who will assume responsibility for reviewing
the project, and for a repeat attempt if necessary; how
failure will be communicated and who will lead on this and
whether the partnership will be expected to remain in place
until a repeat operation is planned and concluded. It is good
practice to build in a review point for partnerships at a key
milestone (e.g. once an eradication operation is completed)
to assess how well the partnership has progressed, whether
all partners have met their commitments, etc. The outputs
of this process could inform the organisations if they are
considering extending the partnership to cover further
projects.
Project plans for eradication projects often conclude
two years after on-island operations cease, or when the
island is declared officially ‘pest-free’. However, in many
situations, site managers or residents will need to remain
engaged with projects in the longer-term, for monitoring,
biosecurity, or to begin more intensive restoration efforts
such as the reintroduction of threatened species. Partners
should make their plans clear as soon as they can, as if some
partners plan to withdraw from working on the island posteradication, it may be necessary to introduce new partners
to assist in post-project site management. In particular,
it is extremely important to be clear about who will be
responsible for maintaining monitoring and biosecurity
arrangements after the eradication project is complete, and
who will respond in the event of a pest incursion (either of
the pest that was eradicated, or something entirely new).
Clarity over funding and resource issues
It is good practice for partners to share information on
their planned contribution to a project, including cash and
non-cash (in-kind) contributions. Some contributions may
be invisible to those partners who are not directly involved.
For example, a partner who is taking the lead on drafting
legal contracts may be spending time and money on
expensive services, but another partner may be completely
unaware of this activity. Project budgets should ideally take
account of in-kind contributions of time from all project
partners, as well as cash contributions from donors. A clear
project plan (as discussed above), and a designated project
manager are important to ensure all partners know what
contribution they are expected to make. Clear governance
of a partnership is also important so that the project
manager is given clear accountability and responsibility
for deliverables and is not told to do different things by
different partners.
Joint fundraising can be a problematic and difficult area
for many partnerships. Partnerships should aim to develop
a collaborative fundraising agenda with mutually agreed
messaging. There should be clarity over which partners
have responsibility for donor cultivation and management,
and how donors will be managed after the project has
concluded to avoid perceptions of “donor poaching” (The
Nature Conservancy, 2017).
Partners need to be clear on what financial disbursement
or opportunities they expect from the project. For example,
Stringer, et al.: Effective partnerships for island restoration
if local communities wish for project supplies to be
purchased from local outlets, this should be flagged early
so that any increased costs associated with this can be
dealt with. If government agencies expect to charge fees to
the project (or conversely will waive standard fees) these
expectations should be raised and addressed early on to
avoid budgetary shocks.
CONCLUSIONS
The days of single agencies implementing island
restoration projects may be waning, as islands with human
populations, mixed tenure, complex legal status and
multiple stakeholders are now high on international priority
lists for future operations (Oppel, et al., 2011; Dawson, et
al., 2015; Stanbury, et al., 2017). Operational managers
and organisations committing to carry out these operations
in future need to be aware of the skills and the level of
time needed to maintain partnerships, and the potential
pitfalls. Recent experiences have served to illustrate that
partnerships can deliver immense gains for conservation
– without them, we would not have seen recent operations
on Vahanga, Antipodes, Palmyra, Desecheo and the Isles
of Scilly (as just a few examples).
Lessons from particular partnerships for island
restoration projects may be captured in grey literature
such as project review documents, but there is little openly
accessible material available about best practice in this
area. It would be very helpful to future project managers
if review documents and templates for partnership
agreements could be shared openly with others. Websites
such as those for the Pacific Invasives Initiative or the Great
Britain Non-Native Species Secretariat could usefully host
this sort of information.
Although working in partnerships can be very
challenging at times, it is apparent that island restoration
will only continue to deliver benefits worldwide if its
practitioners are able to work together and draw in new
organisations. By considering the elements of partnerships
early in the process, we hope that more operations will be
“matches made in heaven” and a shotgun will seldom need
to be drawn from the figurative cupboard.
ACKNOWLEDGEMENTS
We would like to acknowledge the organisers of the
Island Invasives Conference 2017 for including this paper
in the conference programme. Thanks to the audience at
the conference, and those who took the time to comment
on the presentation. Two anonymous reviewers made
comments on the paper which improved it substantially.
Finally, thanks to all in the island eradications community
for your interest in developing and improving the way we
work in future partnerships.
REFERENCES
Bellingham, P.J., Towns, D.R., Cameron, E.K., Davis, J.J., Wardle, D.A.,
Wilmshurst, J.M. and Mulder, C.P.H. (2010). ‘New Zealand island
restoration: Seabirds, predators, and the importance of history’. New
Zealand Journal of Ecology 34: 115–136.
BirdLife International. (2017).’ Invasive Alien Species’. <www.birdlife.
org/worldwide/programme-additional-info/invasive-alien-speciesprogramme-more-info>. Accessed 2 July 2017.
Courchamp, F., Chapuis, J.L. and Pascal, M. (2003). ‘Mammal invaders
on islands: Impact, control and control impact’. Biological Reviews 78:
347–383.
Dawson, J., Oppel, S., Cuthbert, R.J., Holmes, N., Bird, J.P., Butchart,
S.H.M., Spatz, D.R. and Tershy, B. (2015). ‘Prioritizing islands for the
eradication of invasive vertebrates in the United Kingdom overseas
territories’. Conservation Biology 29: 143–153.
DIISE. (2017). ‘The Database of Island Invasive Species Eradications’.
Developed by Island Conservation, Coastal Conservation Action
Laboratory UCSC, IUCN SSC Invasive Species Specialist Group,
University of Auckland and Landcare Research New Zealand, Auckland.
<http://diise.islandconservation.org>. Accessed 26 June 2017.
Flora and Fauna International. (2009). Guidance for Working with Other
Organisations. Report produced as part of the USAID-funded project:
‘Life on the Edge – Conserving Biodiversity and Rebuilding Livelihoods
in the Wake of Natural Disaster and Human Conflict’. <www.faunaflora.org/wp-content/uploads/Partnership_guidance_September2009.
pdf>. Accessed 3 July 2017.
Howald, G., Donlan, C.J., Galván, J.P., Russell, J., Parkes, J., Samaniego,
A., Wang, Y., Veitch, D., Genovesi, P., Pascal, M., Saunders, A. and
Tershy, B. (2007). ‘Invasive rodent eradication on islands’. Conservation
Biology 21: 1258–1268.
Howald, G.R., Donlan, C.J., McClelland, P., Macdonald, N. and Campbell,
K.J. (2011). ‘Advantages and challenges of government, non-profit and
for-profit approaches to eradications: leveraging synergies by working
together’. In: C.R. Veitch, M.N. Clout and D.R. Towns (eds.) Island
invasives: eradication and management, pp. 432–436. Occassional
Paper SSC no. 42. Gland, Switzerland: IUCN and Auckland, New
Zealand: CBB.
IUCN. (2016). ‘The Honolulu Challenge on Invasive Alien Species’.
<www.iucn.org/theme/species/our-work/invasive-species/honoluluchallenge-invasive-alien-species>. Accessed 2 July 2017.
Jones, H.P., Holmes, N.D., Butchart, S.H.M., Tershy, B.R., Kappes,
P.J., Corkery, I., Aguirre-Munoz, A., Armstrong, D.P., Bonnaud, E.,
Burbidge, A.A., Campbell, K., Courchamp, F., Cowan, P.E., Cuthbert,
R.J., Ebbert, S., Genovesi, P., Howald, G.R., Keitt, B.S., Kress, S.W.,
Miskelly, C.M., Oppel, S., Poncet, S., Rauzon, M.J., Rocamora, G.,
Russell, J.C., Samaniego-Herrera, A., Seddon, P.J., Spatz, D.R., Towns,
D.R. and Croll, D.A. (2016). ‘Invasive mammal eradication on islands
results in substantial conservation gains’. Proceedings of the National
Academy of Sciences of the United States of America 113. 15: 4033–
4038.
McClelland, P.J., Coote, R., Trow, M., Hutchins, P., Nevins, H.M.,
Adams, J., Newman, J. and Moller, H. (2011). ‘The Rakiura Titi
Islands restoration project: community action to eradicate Rattus rattus
and Rattus exulans for ecological restoration and cultural wellbeing’.
In: C.R. Veitch, M.N. Clout and D.R. Towns (eds.) Island invasives:
eradication and management, pp. 451–454. Occassional Paper SSC no.
42. Gland, Switzerland: IUCN and Auckland, New Zealand: CBB.
Oppel, S., Beaven, B.M., Bolton, M., Vickery, J. and Bodey, T.W. (2011).
‘Eradication of invasive mammals on islands inhabited by humans and
domestic animals’. Conservation Biology 25: 232–240.
Ozarski, J. (2015). ‘Cooperation for Mutual Benefit: Opportunities for
Primary Industry and the New Zealand Department of Conservation’.
Fulbright
New
Zealand.
<www.fulbright.org.nz/wp-content/
uploads/2015/08/OZARSKI-Jill-final-report-new-PDF.pdf>. Accessed
2 July 2017.
Pacific Invasives Initiative. (2011). ‘Resource Kit for Rodent and Cat
Eradication’.
<www.pacificinvasivesinitiative.org/rk/index.html>.
Accessed 2 July 2017.
Parkes, J.P., Byrom, A.E. and Edge K-A. (2017). ‘Eradicating mammals
on New Zealand island reserves: What is left to do?’ New Zealand
Journal of Ecology 41: 263–270.
Russell, J.C. and Broome, K.G. (2016). ‘Fifty years of rodent eradications
in New Zealand: Another decade of advances’. New Zealand Journal of
Ecology 40: 197–204.
Stanbury, A., Thomas, S., Aegerter, J., Brown, A., Bullock, D., Eaton, M.,
Lock, L., Luxmoore, R., Roy, S., Whitaker, S., and Oppel, S. (2017).
‘Prioritising islands in the United Kingdom and crown dependencies
for the eradication of invasive alien vertebrates and rodent biosecurity’.
European Journal of Wildlife Research 63: 31.
Sterne, R., Heaney, D. and Britton, B. (2009). ‘The Partnership Toolbox’.
World Wildlife Fund. <http://assets.wwf.org.uk/downloads/wwf_
parthershiptoolboxartweb.pdf>. Accessed 2 July 2017.
Tennyson, R. (2011). ‘The Partnering Toolbook.’ The Partnering
Initiative.
<https://thepartneringinitiative.org/publications/toolbookseries/the-partnering-toolbook/>. Accessed 22 April 2018.
The Nature Conservancy (TNC). (2017). ‘Partnering for Conservation’.
<www.conservationgateway.org/ConservationPlanning/partnering/
Pages/default.aspx>. Accessed 2 July 2017.
Towns, D.R. and Broome K.G. (2003). ‘From small Maria to massive
Campbell: Forty years of rat eradications from New Zealand islands’.
New Zealand Journal of Zoology 30: 377–398.
Towns, D.R., West, C.J., and Broome, K.G. (2013). ‘Purposes, outcomes
and challenges of eradicating invasive mammals from New Zealand
islands: An historical perspective’. Wildlife Research. 40(2): 94–107.
Varnham, K. (2011). ‘Involving the community in rodent eradication
on Tristan da Cunha’. In: C.R. Veitch, M.N. Clout and D.R. Towns
(eds.) Island invasives: eradication and management, pp. 504–507.
Occassional Paper SSC no. 42. Gland, Switzerland: IUCN and
Auckland, New Zealand: CBB.
Wilcox, D. (1998). ‘Guide to Development Trusts and Partnerships, The
Partnerships Handbook’. <www.partnerships.org.uk>. Accessed 29
June 2017.
Wilson, C. (2005). ‘Developing Effective Partnerships Between the
Department of Conservation and Community Groups’. Science
for Conservation 248. Wellington, New Zealand: Department of
Conservation.
521
A. Walsh, A. Wilson, H. Bower, P. McClelland and J. Pearson
Walsh, A.; A. Wilson, H. Bower, P. McClelland and J. Pearson. Winning the hearts and minds –
proceeding to implementation of the Lord Howe Island rodent eradication project: a case study
Winning the hearts and minds – proceeding to implementation of the
Lord Howe Island rodent eradication project: a case study
A. Walsh1, A. Wilson1, H. Bower1, P. McClelland2 and J. Pearson3
Lord Howe Island Board, P.O. Box 5, Lord Howe Island, NSW Australia, 2898. <andrew.walsh@lhib.nsw.gov.au>.
Pete McClelland Conservation Services, 237 Kennington – Roslyn Bush Road, Invercargill 9872, New Zealand.3Royal
Society for the Protection of Birds, The Lodge, Sandy, Bedfordshire SG19 2DL United Kingdom.
1
2
Abstract Lord Howe Island (LHI) is a World Heritage listed volcanic remnant island, 600 km from mainland Australia.
Home to many threatened endemic species, and an important nesting site for many migratory seabirds, LHI is also home to
an established community of 350 people. The island economy relies heavily on tourism, based around the World Heritage
landscapes, and terrestrial and marine biodiversity. The impact of rats and mice on the world heritage values of the island
are well documented. A feasibility study in 2001 considered the eradication of both rats and mice on the Lord Howe Island
Group (LHIG) in a single operation as feasible and achievable. Since then, numerous studies and community consultations
have been undertaken; and the methodology, risks and benefits of an eradication effort have been carefully considered
and evaluated. The project has long been topical in the community, with both ardent support and opposition. Though
funding for the project was secured in 2012, implementation of the project has had repeated delays, and consequently,
additional consultation efforts, studies and assessments have been required to address lingering community concerns. This
paper describes a wide range of techniques that were used in an effort to gain community support and acceptance of the
project, and ultimately, to enable its implementation. The paper also outlines those aspects of the social process that have
led to greater support and those which led to less support and, in some cases, opposition. Within these lessons learnt, the
paper presents some insights around how communities engage with and respond to scientific information and pest control
(including perceived influence on livelihoods, conservation, health and the legacy to be left for future generations). These
insights are explained in the context of psychological processes such as emotional responses (fear/trust) and personal
values (economic, environmental, social). Given that the eradication of invasive predators on larger inhabited islands
is the next logical step for island conservation, projects such as the LHI Rodent Eradication Project are of particular
value as real life hard-earned lessons. We hope that these findings can help facilitate future successful island eradication
programmes that work with the community.
Keywords: community, economic, engagement, human health, inhabited, tourism
INTRODUCTION
Purpose
The purpose of this paper is to present a case study of
eradication planning on an inhabited island, illustrating the
importance of social impact considerations and presenting
some guiding principles and lessons learnt.
The island
Lord Howe Island (LHI) is located 570 km east of
Australia (Fig. 1). It covers 1,455 ha, is 12 km long, and
1.0–2.8 km wide. The LHIG was listed as a World Heritage
Area in 1982 and is located within the Lord Howe Island
Marine Park (NSW).
LHI is part of the State of New South Wales and is
administered by the Lord Howe Island Board (LHIB),
which comprises four locally elected islanders and three
ministerially appointed mainland members. The LHIB
(and its administrative arm) are directly responsible for
the care, control and management of the island’s natural
values and the affairs and trade of the island and carry out
all local government functions on behalf of approximately
350 island residents.
The settlement area covers about 15% of the island (400
ha) and is used predominantly for residential, pastoral/
agricultural and commercial purposes. Tourism is the most
significant industry and major source of income on the
island and is heavily focused around the world heritage
values of both the marine and terrestrial environments
(Lord Howe Island Tourism Association, 2015).
Current impacts of rodents on LHI
Ecological impacts
The devastating ecological impacts of introduced rodents
on offshore islands around the world are well documented
(Groombridge, 1992; Towns, et al., 2006; Jones, et al.,
2008). Similar impacts on LHI have been observed since
the arrival of mice (Mus musculus) in approximately the
1860s and ship rats (Rattus rattus) via a shipwreck in 1918.
The Lord Howe Island Biodiversity Management Plan
(DECC, 2007a) summarises the immediate and ongoing
impact the introduction of rodents has had on flora and
fauna species on LHI, including extinction of several
species of birds, plants and invertebrates.
Socio-economic impacts
From the perspective of the human population, rats
and mice are major domestic pests. They infest residences,
destroy foodstuffs and vegetable gardens; and contaminate
homes with excrement. They are also a known health risk to
humans as they harbour and transmit diseases and parasites
such leptospirosis and rat lungworm disease (Shiels, et al.,
2014).
From an economic perspective, rats cause considerable
economic loss to the island's kentia palm industry, with
predation of seed as high as 30% (Parkes, et al., 2004)
severely reducing seed production (Pickard, 1983; Billing,
1999; further detail in Wilkinson & Priddell, 2011).
Tourism, the LHIG's main industry, is based on the
islands' unique biodiversity and World Heritage values.
These values are significantly threatened by rodents
(IUCN, 2017) therefore reducing the visitor experience
offered by the island.
History of rodent control on LHI
Islanders and the LHIB have been involved in the control
of rodents (rats and mice) on Lord Howe Island since about
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
522
up to meet the challenge, pp. 522–530. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Walsh, et al.: Lord Howe Island rodent eradication
be eradication of the masked owl (Tyto novaehollandiae),
which was deliberately introduced to LHI in the 1920s and
1930s to control rats. The species now also preys on on
many island birds (Milledge, 2010).
The one-off eradication proposes to distribute a
cereal-based bait pellet (Pestoff 20R) containing the
toxicant brodifacoum across the LHIG via dispersal from
helicopters to the uninhabited parts of the island, and by
a combination of hand broadcasting and the placement
of bait in bait stations in the settlement area (for more
information, see LHIB, 2016). Post eradication, rodent
prevention and surveillance monitoring would be ongoing
to prevent reinvasion and would need a high level of
community vigilance.
The LHIB received significant funding (AUS$9.5M)
in 2012 for planning and implementation of the REP
from the Federal Government’s former Caring for Our
Country programme (now National Landcare program)
AU$4,500,000 and the NSW Environment Trust
AU$4,542,442.
PROJECT DEVELOPMENT AND EARLY
CONSULTATION
During the turn of the century, successes in island
eradications undertaken primarily in New Zealand
(summarised in Russell & Broome (2016)) were gaining
international attention. At the same time there was growing
recognition by government in Australia (state and federal)
of the impacts of rodents on LHI (summarised in NSW
Scientific Committee, 2000; TSSC, 2006 and DECC,
2007a).
Fig. 1 Lord Howe Island showing the Permanent Park
Preserve, airstrip and roads.
1920, highlighting both the long-recognised impacts of
rodents and difficulty in achieving meaningful outcomes
through ongoing control on the island (Saunders & Brown
2001). Previous control methods included a bounty on rat
tails, hunting with dogs and shooting, introduction of owls
and the use of various poisons including barium chloride,
diphacinone, warfarin and brodifacoum (ibid). Control
baiting on the island has undergone several reviews over
time with Billing (1999) the most recent, resulting in a
current pulse baiting schedule delivering over 4.5 tonnes
a year of Ratex (coumatetralyl) and Roban (difenacoum)
over approximately 10% of the island. While the use of
brodifacoum has been discontinued by the LHIB in the
lead up to the eradication, many Island residents continue
to use brodifacoum-based rodenticides such as Talon™ and
Tomcat™ to control rats and mice around their properties
and inside dwellings. The above is intended to highlight the
long-term and extensive use of rodenticides on the island,
which could be avoided through a one off eradication, but
also to highlight that the use of poison on the island is not
new and residents are quite familiar with it.
The project
The LHIB is proposing to undertake the Lord Howe
Island Rodent Eradication Project (LHI REP) to eradicate
introduced rodents from LHI. A secondary outcome would
In 2001, a LHIB staff member and a community
member attended an international eradication conference
and subsequently submitted a proposal to the LHIB seeking
support for research into feasibility of eradicating rodents
on LHI. This led to commissioning of a feasibility study
(Saunders & Brown, 2001) jointly funded by the LHIB and
WWF. Saunders and Brown concluded that eradication on
LHI was feasible using a combination of aerial broadcast,
hand broadcast and bait stations with a brodifacoum based
product. Based on some initial consultation undertaken
with community on the island during the study, it was
considered that the socio-political environment on LHI
was conducive to supporting a possible eradication. The
study identified potential ecological and social risks and
gaps that needed to be further explored and recommended
key next steps including:
● a cost benefit analysis
● additional field trials on rodent densities, bait uptake
and non-target species impacts
● that a process be established to allow the community
to be kept informed and be able to influence
decisions
● establishment of a Taskforce to drive implementation
● the feasibility report be made available to the
community and briefing sessions provided.
Following the feasibility study, a Rodent Eradication
Taskforce was established by the LHIB through an
“expression of interest” process in the community, meeting
initially in 2002.
Receipt of a donation (ca. AUD$34,000) from the
Foundation for National Parks and Wildlife allowed
commissioning of a Cost Benefit Analysis (Parkes, et al.,
2004). The study looked at additional feasibility, risks
and benefits of eradication on LHI and again confirmed
that eradication was feasible and highly beneficial,
provided risks could be appropriately managed and
523
Island invasives: scaling up to meet the challenge. Ch 3B Strategy: Collaboration
funding and approvals obtained. The analysis identified
social constraints suggesting that people may oppose any
proposed eradication attempt at two levels: some may
not see the need to attempt it, while others may disagree
with the methods required. A process to involve the LHI
community in considering options and an active process to
seek agreement for baiting on properties and in houses was
recommended in the study, but not detailed.
For several more years the LHIB sought funding to
deliver facilitated community engagement and on island
trials prior to implementation of the REP. Further funding
(ca. AUD$ 200,000) was secured in 2008 from the
Australian Government Caring for our Country Program to
continue planning and trials, community consultation and
engagement of a project manager. Consultation in 2008
included several rounds of public hall meetings including:
● specific sessions for livestock, poultry and dog
owners
● for residents on the fringe of the settlement area
where special consideration are required
● follow up sessions to address concerns previously
raised.
A theme emerged that the meetings quickly
disintegrated as some community members become vocal
and dominating, particularly those opposed. Over several
meetings, this led to a reduction in numbers of attendees.
It was also unfortunately stated by a scientist at one of the
public hall meetings at the time that “the eradication would
not go ahead unless there was 100% support”. This created
an undeliverable promise still haunting the project today.
Ad-hoc consultation also occurred through development
of fact sheets, individual meetings and through briefing
papers and updates at Board meeting open sessions.
Based on recommendation in Saunders & Brown
(2001) and Parkes, et al. (2004), additional studies were
also conducted on the LHI currawong (Carlile & Priddel,
2006) and non-toxic field trials (DECC, 2007b) that
examined rodent and non-target species uptake of the bait
pellets, bait breakdown in the environment and spread of
the bait using helicopter.
In July 2009, locally elected Board members went door
to door surveying residents on their views and concerns
regarding the REP based on set questions. In total 125
residents were interviewed and detailed responses showed
that there was sufficient community support to proceed.
The survey results were used together with results of field
trials to develop the Draft LHI Rodent Eradication Plan
(LHIB, 2009) which then underwent external peer review.
The Draft Plan was peer reviewed by the Island Eradication
Advisory Group (IEAG) of the New Zealand Department of
Conservation; the Invasive Species Specialist Group of the
Species Survival Commission of the World Conservation
Union; the Worldwide Fund for Nature (WWF), Australia;
Birds Australia; Landcare Research, New Zealand; CSIRO
and Professor Tim Flannery. All peer reviews of the Plan
were supportive.
Public comment on the Draft Plan was sought in
November 2009. Of the 83 submissions received, 39
submissions opposed the Plan, 33 supported it, four gave
in-principle support, while one submission was undecided.
All 39 submissions opposing the Plan were from LHI
residents, organisations acting on their behalf, or had
strong links with LHI community members. Of the 37
submissions supporting the proposal, 11 originated from
LHI, and four came directly from scientists or scientific
groups with experience in rodent eradication. All four
submissions giving in-principle support originated from
LHI as did the single undecided submission. It should be
524
noted that 84 submissions is considered a very high level
of response on LHI.
The most frequently raised issue (25 submissions)
was concern about non-target impacts during the
proposed eradication operation. Other dominant issues
included concerns about possible impacts on the health
of the community, the tourism industry and the marine
environment, as well as the need for improved consultation.
The most commonly mentioned issue supporting the Draft
Plan was that eradication would deliver clear environmental
benefits to LHI (13 submissions) (see Table 1).
The submissions provided a valuable snapshot of
opinion, setting the direction for future studies and
consultation for the project once additional funding could
be secured. A submission analysis report was prepared
but unfortunately due to not being able to secure funding
for consultation, was not released to the community until
2013. A revised Draft Plan was also prepared addressing
submissions but was never released, for reasons unknown.
One study that was progressed immediately was a
Human Health Risk Assessment (HHRA) undertaken
by a toxicology consultant Toxikos (2010). The HHRA
considered all potential pathways related to direct and
indirect contact with the poison from the REP (e.g. ingestion,
inhalation, skin contact, ingestion of contaminated
water and food) and found there to be no significant risk
associated with any pathway with the proposed mitigation
in place. The HHRA was presented to community and
helped to satisfy some community concerns. However,
although the LHIB undertook a competitive tender process
to select the consultant Toxikos, some members of the
community criticised the independence of the study and
therefore disregarded the results. Subsequent third-party
reviews of Toxikos’ HRRA by New South Wales Health,
South Australian Health and Pacific Environment Pty Ltd
(outlined in LHIB, 2016) did little to change the perception
of some community members.
From 2010 to 2012 the LHIB continued to seek staged
funding for progression of the REP. In May of 2012 it was
recognised by the Board, that despite information provided
to date it was clear that community concerns remained, and
further work was required to address these. To enhance
community awareness of the benefits that eradication
would deliver for the environment, tourism and public
health, it was recommended that a professional facilitator
be engaged to consult with the LHI community. It was also
recommended that a Community Liaison Group (CLG)
was created. In June 2012 the project manager at the time
resigned.
In July 2012, the LHIB received funding of AU$9.5M
to implement the REP in full from the New South Wales
Government’s Environment Trust and the Australian
Government’s Caring for Our Country programme. The
project was divided into three stages:
● Stage 1 – to complete all planning and preparations
for the eradication operation
● Stage 2 – to implement the baiting strategy including
captive management and post baiting monitoring
● Stage 3 – to monitor the environmental outcomes of
the baiting operation.
Again, unfortunately, receipt of the full funding at once
led to some perception in the community that the REP
was a fait accompli and no longer open for community
discussion.
In late 2012, a selection process for engaging a
community consultation facilitator was undertaken. This
included involvement of Board members and 13 community
Walsh, et al.: Lord Howe Island rodent eradication
panel members to choose an applicant that would be able
to “connect” with islanders. Two shortlisted consultants
travelled to the island and were interviewed individually
by the 13-member community panel. Consultants ‘Make
Stuff Happen’ were selected and contracted to establish
the CLG and, together with the CLG, to develop a draft
Communication and Community Engagement Strategy.
The CLG (12 members from 17 nominations) held their
first meeting on Friday 8 February 2013. Terms of reference
were to:
● Review REP information so it is clear, correct and
relevant
● Identify ways to communicate with the community
about implementing the REP
● Discuss issues and concerns about the REP
Additional meetings were held in March, April and
June 2013, facilitated by the consultants. These meetings,
whilst unpicking community issues also brought to light
that community support: opposition was approximately
50:50. A Community Engagement Report and Plan (Make
Stuff Happen, 2013) was developed recommending:
● Providing sufficient resources to restore trust and
information flow
● Maintaining the momentum of the CLG and
relationship with the LHIB
● Creating a compelling case by focussing on key
drivers of change
● Providing content information at different levels
● Providing a variety of options for how information is
received focusing on small scale approaches
● Demonstrating respect for community concerns and
local knowledge.
Further detail can be found in the report (Make Stuff
Happen, 2013).
As a recommendation of the report was to try and build
support, the consultant facilitated an “open house” bringing
experts relevant to the eradication to LHI in Aug 2013. This
included a toxicologist, a medical doctor who has worked
on eradications (Macquarie and South Georgia), an animal
husbandry expert and project staff. Experts were available
separately at small tables where community members
could sit and ask questions of them over two days. This
was held in the neutral ground of the museum and was
attended by about 65 residents. Some individuals, however
were known to actively boycott the event.
Additional activities in 2013 included:
● Meeting with Tourism Association to discuss risk
analysis
● Targeted discussions with specific businesses related
to tourism issues
● Key messages refined and communicated through a
variety of means.
Following the open house, ‘Make Stuff Happen’
provided additional recommendations and proposed a
Stage 2 engagement strategy to the LHIB. However, given
the value of the contract, it had to be retendered on the
open market.
Table 1 Key issues raised in submissions to the 2009 Draft Rodent Eradication Plan.
Issue
Non-target impacts
Human health concerns
More consultation required
Tourism impacts
Marine impacts
Economic impacts
Eradication will deliver environmental benefits
Proposed eradication too risky
Children’s health concerns
Threat posed by negative media associated with eradication
Question rodent impacts
Feasibility – it won’t work!
Captive management issues
High cost of operation
Use of divers to remove bait
Rodents have significant impacts
Don’t support aerial baiting
Peer review process flawed
Quarantine efficacy – new protocols to prevent reinvasion
Expand current control programme
The eradication is an experiment
Need to work with community to gain support
Distrust of Board
Number of
submissions
% of
submissions
25
18
18
16
14
13
13
10
9
9
8
8
7
7
7
7
6
6
6
5
5
5
5
30.1
21.7
21.7
19.3
16.9
15.7
15.7
12.0
10.8
10.8
9.6
9.6
8.4
8.4
8.4
8.4
7.2
7.2
7.2
6.0
6.0
6.0
6.0
525
Island invasives: scaling up to meet the challenge. Ch 3B Strategy: Collaboration
In recognition of the differing views within the
community putting successful implementation at risk, the
LHIB decided in early 2014 to put the proposed eradication
on hold, and to go back to the community and to discuss the
available options. The Board made the decision to divide
the project into two separate but linked stages.
Stage One: community engagement and consultation
which would go back to basics and ask what the
community wants in relation to the eradication of rodents
so that they can make an informed decision on the future
of the project. This included the consequences of not doing
the eradication. At the completion of that process an
assessment was to be made of the level of support to gauge
whether it is sufficient to progress to Stage Two.
Stage Two: operational implementation, which would
commence in June 2015, but would only take place if there
was sufficient community support for the project following
the consultation process.
The tender process to select a consultant for the
additional community engagement was undertaken with
the assistance of the Department of Premier and Cabinet
in early 2014 with Elton Consulting selected. Between
July 2014 and February 2015, Elton Consulting undertook
a series of community consultation visits to Lord Howe
Island. They spoke on a one-on-one basis, through
personal visits or open sessions at the public hall, to many
island residents, (on a number of occasions) concerning
the issue of rodent control and potential eradication on
the island. They implemented an incremental approach to
consultation to unpack the complexity of the community
response to the previous rodent eradication process, and to
identify what it would take for the community to actively
engage in the evaluation of alternatives and options, with
the aim of obtaining community support or endorsement of
one of the options.
A Community Working Group (CWG) was established,
based on residents who indicated a willingness to participate,
along with an open call for nomination / involvement,
put out through a newsletter to community residents. In
working towards a solution, the CWG identified many
issues (particularly regarding human health, potential
impacts to business and tourism and potential impact to
the environment) and considered a range of options. The
option to “do nothing” was generally not considered as
an alternative, as there was broad agreement that rats and
mice are a problem, and that Lord Howe Island would be
better off with no rodents.
together with an anonymous survey (Elton Consulting,
2015b) to allow the community to choose between:
Option 1 – Retain and expand the current management
programme
Option 2 – moving to the planning and approvals stage of
an eradication programme.
The survey also asked for level of agreement on
whether the rodent problem needed to be addressed and
ranking of areas concerns for both options.
A total of 212 respondents (71% of the 299 people on
the electoral roll) participated in the survey. 208 survey
responses were received before the closing time. A
consensus was reached that the rodent problem on Lord
Howe Island needs to be addressed with the majority of
respondents (91%) agreeing (38%) or strongly agreeing
(53%). A marginal majority 52 of the respondents expressed
a preference for Option 2, while 48% of respondents
expressed a preference for Option 1.
In line with the agreed Process for Resolution (Fig. 2),
the LHIB responded to the majority view and on 19 May
2015 made the decision to proceed to the Planning and
Approvals Phase (Option 2).
PROGRESSING TO IMPLEMENTATION
Since 2015, the project team focussed jointly on
progressing the necessary approvals and operational
planning; and continuing to increase community support
for and/or acceptance of the REP, recognising that some
people who may not support the REP would however
accept it. The latter is critical to ensuring that baiting can
be conducted on every property on the island. Residual
community issues and how we are attempting to resolve
them are detailed below.
Human health
Safety of people has always been a priority for the
LHIB and the community when considering the LHI REP.
Given the criticism of the independence of the original
HHRA described previously, the community suggested
that a further additional study be undertaken.
The NSW Office of the Chief Scientist and Engineer
(OCSE) was identified by the community / CWG as an
agency with a high level of independence and credibility
and was subsequently requested to oversee an additional
Two scenarios were therefore further investigated and
discussed:
1. Ongoing control through the existing baiting program,
and the potential to expand this.
2. An eradication programme as previously proposed or
modified where possible to address island residents’
concerns.
The CWG agreed to develop and implement a
community survey to test community support for these
scenarios, whilst recognising that many of the community
concerns with the proposed eradication could be addressed
during the Planning and Approvals Phase. The CWG also
agreed that an additional independent HHRA was needed
and should be progressed.
In May of 2015, an options paper (Elton Consulting,
2015a), providing detailed explanation of options and
answers to key questions, was disseminated to all people
registered on the electoral roll for Lord Howe Island,
526
Fig. 2 The agreed process for resolution.
Walsh, et al.: Lord Howe Island rodent eradication
independent HHRA for the project in line with the agreed
Process for Resolution (Fig. 2). The OCSE was requested
by the NSW Minister for the Environment to convene an
Expert Panel in 2016 to:
● Provide advice to the Board on processes for
commissioning the HHRA including identification
of suitable experts and scope of the request for
proposal. (The CWG endorsed the Scope.)
● Review proposals to undertake the HHRA and select
a preferred candidate; review project plans and
methodologies, and review draft and final reports of
the HHRA as required. (Proposed membership of the
panel was endorsed by the CWG.)
● Provide advice to the Minister for the Environment
on the HHRA.
● Respond to media enquires as they relate to the
Terms of Reference for the Expert Panel.
The Expert Panel (with the assistance of two nominated
CWG members) selected Ramboll Environ Pty Ltd. to
undertake the HHRA. This work, overseen by the OCSE,
concluded that estimates of exposure from all potential
sources associated with the REP are below those likely to
result in adverse health effects for residents and visitors.
(Ramboll Environ, 2017).
The outcomes from this additional HHRA and expert
panel review concur with the results of previous HHRA
undertaken by Toxikos Pty Ltd (2010), which found that,
with the proposed mitigation in place, the REP would be
safe for the community and visitors. Whilst the outcome
was the same as the previous study, it should be noted the
process undertaken by the OCSE to select Ramboll was
very different to how Toxikos were originally selected,
giving the report substantially more credibility within the
community.
Economics, tourism and livelihoods
Ongoing concern from some elements of the community
regarding potential impacts to tourism (specifically
reduction in visitors) before, during or post eradication
led the LHIB to commission an economic evaluation of
the project in 2016. Community input was sought for the
development of the scope of the evaluation and a prominent
local business owner was included on the tender selection
panel.
A study by Gillespie & Bennett (2017) looked at the
costs and benefits (market and non-market services) of
not proceeding with the REP compared to the costs and
benefits of proceeding (i.e. a Cost-Benefit Analysis) as well
as the distribution of said costs and benefits. Though all
costs and benefits were considered, particular focus was
placed on the potential impacts or benefits to biodiversity
(non-market services) and tourism (market service) as the
major contributors. This is based on the fact that a key
motivation for visiting LHI is to experience the natural,
undeveloped and unspoilt surroundings (Lord Howe Island
Tourism Association, 2015), some of which are under
threat from rodents.
Using choice modelling undertaken for other relevant
studies, Gillespie & Bennett (2017) applied the benefit
transfer technique to provide an economic value estimate
for the biodiversity value of protecting species from
extinction. Considering the economic importance of
tourism to LHI, tourism impacts or benefits were modelled
using supply and demand data in peak and off-peak
tourist periods, before, during and after the REP. The
study showed that accommodation providers on the island
would be the biggest beneficiaries during the REP as the
workforce required for the project would more than offset
any temporary reduction in tourism. It also showed that
tour and accommodation providers would be the major
beneficiaries of increased tourism after the eradication.
The REP was demonstrated to have a Benefit to Cost
ratio of 17:1, resulting in an estimated net social benefit
of AU$142M, with AU$58M of that returning directly to
LHI residents. Hence, the REP was justified on economic
efficiency grounds.
Overall, the cost-benefit analysis was considered an
important tool for the REP in overcoming some residents’
concerns about tourism and the economy. Others, however,
will not be convinced until they actually see the visitors on
island during or after the REP.
The eradication method
There has been considerable debate in the community
about the method proposed for the eradication and, in
particular, the aerial distribution delivery method.
A range of alternatives for eradicating rodents were
considered for LHI including alternative techniques and
mortality agents. Many were considered to have fatal
flaws and were unsuitable for use for eradication on LHI
because: a) the technique was not suited to the terrain or
size of the island, b) it did not ensure that all individuals
would be killed or c) was too experimental. However, early
in the project these flaws in alternative methods were not
well communicated with the community. The only method
identified as capable of removing every rat and mouse on
LHI was aerial distribution, in conjunction with minimal
hand broadcast and bait stations where required (i.e. the
settlement area), of highly palatable bait containing an
effective toxicant.
To overcome concerns in the community relating to the
method, the project team has recently taken the community
through the process of looking at all the options and
ruling out options with fatal flaws and therefore not
suitable for deployment on LHI (see Table 2). In addition,
the project team have undertaken one on one property
management plans with all residents to agree the particular
baiting method to be used on their property considering
their concerns. These two tools have led to a greater
understanding in the community of the methods proposed
and why other methods were unsuitable.
Environmental and non-target impacts
Some members of the community have been concerned
about environmental and non-target impacts. We have used
a combination of methods to help allay concerns including:
● Undertaking monitoring on the island to provide
evidence of rodent damage to a range of species on
LHI and communicated those results back to the
community regularly. We have found that a single
photo of a rat taking a seabird chick or egg at a local
nesting ground to be much more powerful than
scientific reports of images of rodent damage locally
or images from other locations around the world.
● Conducting a range of studies on the island to
determine locally at-risk species and those not at risk
(detail in Wilkinson & Priddell, 2011) and repeatedly
communicating the results to community.
● Engaging world experts in captive management to
manage the captive management programmes on
LHI for the two high risk species and conducted
trials on-island to show the community how the
species can be managed without harm (Taronga
Conservation Society, 2014).
527
Island invasives: scaling up to meet the challenge. Ch 3B Strategy: Collaboration
Table 2 Assessment of eradication options.
Eradication
technique
Suitable for
eradication
Feasible for
eradication on LHI
Justification
No suitable pathogen yet developed that could eliminate all
individuals.
Disease
No
No
Trapping
Yes
No
May be feasible for eradication on small islands, however
may cause individuals to become trap shy. Size and
inaccessible terrain of LHI makes this option unfeasible.
Biological
No
No
Currently experimental. Likely to fail to completely
eradicate the target species.
Fertility Control
No
No
No suitable fertility control yet developed that could
eliminate all individuals.
Toxicant – bait
station / hand
broadcast only
Yes
No
May be feasible for eradication on small islands.
Size and inaccessible terrain of LHI makes this option
unfeasible.
Toxicant – aerial
broadcast only
Yes
No
Highly successful on uninhabited islands. Socially
unacceptable on LHI. Problematic with the number and
nature of buildings.
Toxicant –
combination of
aerial and hand
broadcast / bait
stations
Yes
Yes
Allows for bait to be made readily available to all individual
rodents. Brodifacoum in the form of Pestoff 20R has been
selected as the preferred toxicant on LHI considering proven
success, efficacy and non-target impacts.
● Talking to the community about the differences
(particularly in species) between LHI and other
islands where non-target impacts have been
observed. This included talking through differences
in ecology and feeding behaviour leading to different
risk profiles.
● Being open and up front with people about our
expectations of non-target impacts, how we have
formed those expectations and how those impacts
compare to current impacts from rodents.
● Conveying the thoroughness, scientific rigour, and
independence behind the environmental assessments
undertaken by the various regulatory agencies that
have assessment and approval roles on the project.
● Highlighting recovery of species and ecosystems
from eradications around the world, through sharing
recovery stories, science and media on other
eradications.
● Developing appropriate mitigation plans for
domestic animals and livestock at an individual
level.
Lack of trust
Lack of trust of new people, new technologies, and the
LHIB in general, was perhaps the most difficult issue to
address. It manifests as suspicion of non-island experts
and scientific reports, unwillingness to accept change,
spreading of misinformation and criticism of LHIB
decisions and communications. It stems partially from a
sense of resentment by islanders of Government control
of the island. It also stems partially from a history of poor
communication and follow through by the LHIB on many
issues unrelated directly to the REP. Trust is essential to
being able to communicate all aspects of the REP, including
new information risk, benefits, mitigation and for people
to feel comfortable expressing their true concerns. Trust
is not easily given and has to be earnt over a long time
through listening, demonstrating genuine interest in all
aspects of the community (not just those related to rodent
528
eradication), doing what you say you will and following
through on commitments.
The most important mechanism we have found to build
trust is to have our core project staff living on the island
and living in the community for as long as possible. Our
Project Manager and Assistant Project Manager have both
been resident on LHI for at least two years. Wherever
possible we engaged locals in the project including in
communications roles, in advocacy, in adding local
knowledge, in brainstorming and in any other aspect where
they are willing and able.
CURRENT STATUS
The Planning and Approval phase was completed in
2017. At the Sept 2017 Board meeting, the LHIB made the
final decision to proceed with the eradication based on the
technical, social and financial feasibility of the project as
per the agreed process for resolution that was an outcome
of ongoing community consultation in 2015 including:
● the status of approvals
● level of community support
● recommendations from an additional independent
HHRA.
In March 2018, a decision was made to delay
implementation of the REP until winter 2019 due to not
having received one of the permits (previously received
and surrendered due to a technical administrative flaw).
LESSONS LEARNT
Planning a rodent eradication project on an inhabited
island over a long period of time has given us many chances
to reflect on what has worked well and what hasn’t and to
adapt our strategies over time.
Walsh, et al.: Lord Howe Island rodent eradication
Least effective tools
On reflection, the tools we have found least effective
were:
● Having a predetermined solution (eradication by
the methods proposed) with little opportunity for
genuine community input and influence on the
decision-making process at the start. This has left
some members of the community disenfranchised.
● Flooding the community with more and more
technical, scientific information. While science plays
a critical role in providing information and answers,
it needs to be provided in a way that key concepts
and results can be easily understood, are relatable
and align with people’s values. Many scientific
reports are too technical and too detailed for many
people to understand, easily, or relate to. More
information is not always better.
● Public Town Hall meetings. We have found that
these have generally been dominated by a select
minority of people (either supporters or opposition),
which leaves many people unheard or losing interest.
● Scientists and ecologists are not always the best at
community engagement. Often pure science does not
address the emotional issues particularly when they
concern people’s children or livelihoods.
● Mainland consultants and experts. We have found
generally (although not always) poor engagement
outcomes where we have flown specialists from the
mainland to help with various technical aspects.
Often the consultants are there for very short
timeframes and may not return. As a result, the
required depth of understanding of community
issues is often overlooked and, consequently, there
is little opportunity to build rapport and trust. While
we recognise that subject matter and consultation
experts are required, they may be best deployed
behind the scenes providing the right advice to the
core project team.
● Presenting information from other sites – each
community considers their island to be different from
all others.
Most effective tools
We have found the most effective tools for
communicating with our community to be:
● Having key members of the project team based
on the island for as long as possible before
implementation. This gives the community the
opportunity to get to know the team and start to build
trust. An open-door policy allows people to come to
the project team at any time to discuss any concerns
they have with the project. Being based on LHI also
allowed the team to understand the broader issues
that face the community and to interact with the
community outside of work.
● One-on-one consultation with every resident,
repeatedly and as many times as necessary on their
properties. This has allowed us to identify individual
concerns (and underlying motivations) and work
with residents to address them. Working through
with residents about exactly how they would like
baiting undertaken on their properties considering
their concerns (i.e. vegetable gardens, pets, children,
etc.) has been critical for getting people comfortable
with the project.
● The economic evaluation was an important piece of
work as it converted biodiversity values (negative
outcomes without eradication and positive outcomes
with eradication) into economic (tourism) terms
that our community could relate to in a meaningful
way (livelihoods) and understand key concepts and
implications.
● Independence and credibility of the OCSE in
undertaking the HHRA.
● Engaging locals on as many aspects of the project as
possible where skills allowed.
● Being patient, passionate, resilient and willing to go
the extra mile to succeed. These personality traits of
individuals in the core project team are essential to
gaining trust in the community.
DISCUSSION
The LHI REP has been a long time in planning and
community consultation. Implementation of the project
was likely drawn out most due to the fact that early
planning and consultation was not done as effectively as
it could have been. Additionally, funding was received in
full which could have sent the message to the community
that the project was a fait accompli. This did not engender
the initial community support that was essential and has
led to a long road to recovery. Though many of the tools
from the emerging Strategic Environmental Assessment/
Social Impact Assessment toolbox (Russell, et al, 2018)
were eventually employed on the project, these were often
reactive, not integrated and used too late in the process.
The project would have likely encountered much
less community resistance if these tools were used much
earlier and in an integrated and methodological fashion. It
is likely that the plan that was taken to community was
too far developed down a particular path and was therefore
considered not open to community input. This meant that
it wouldn’t have allowed sufficient opportunity for the
community to adapt the plan to their needs or to feel a sense
of ownership of the plan. Although extensive community
engagement has since identified issues, and these were
addressed, it would have been much more effective if
this was done at the start of the planning process. Early
community engagement (not information sharing) to gain
support needs to be the top-priority for future eradications.
ACKNOWLEDGEMENTS
The LHI REP is funded by the Australian Government’s
National Landcare programme and the New South Wales
Environmental Trust. This paper provides an update to a
previous paper published by Wilkinson & Priddell (2011)
and builds on a paper by Hutton, et al. (2016) at the Island
Arks 2016 conference.
Thanks to previous and current staff involved in the
project. This paper has been prepared with the benefit of
hindsight and is by no means criticism of those previously
involved.
REFERENCES
Billing, J. (1999). The Management of Introduced Rodents on Lord Howe
Island. Unpublished report. Lord Howe Island. Lord Howe Island
Board.
Carlile, N. and Priddel, D. (2006). Population Size and Distribution of
the Lord Howe Currawong Stepera graulina crissalis. Unpublished
Report prepared for the Lord Howe Island Board. NSW: Department of
Environment and Conservation.
529
Island invasives: scaling up to meet the challenge. Ch 3B Strategy: Collaboration
DECC (2007a) Lord Howe Island Biodiversity Management Plan.
Hurstville: Department of Environment and Climate Change.
DECC (2007b). Report on Non-toxic Bait Trials Lord Howe Island –
August 2007. Unpublished report for the Lord Howe Island Board.
Hurstville: Department of Environment and Climate Change.
Elton Consulting. 2015a. Rodent Control on Lord Howe Island – Towards
a Solution. Unpublished report for the Lord Howe Island Board. Elton
Consulting Pty Ltd.
Elton Consulting. 2015b. Lord Howe Island Rodent Consultation
Community Survey. Unpublished report for the Lord Howe Island
Board. Elton Consulting Pty Ltd.
Taronga Conservation Society Australia, (2014). Captive Management
for Woodhen and LHI Currawong Associated with the Lord Howe
Island Rodent Eradication Project. Unpublished report for the Lord
Howe Island Board. Sydney: Taronga Conservation Society.
Towns, D.R., Atkinson, I.A.E. and Daugherty, C.H. (2006). ‘Have
the harmful effects of introduced rats on islands been exaggerated?’
Biological Invasions 8: 863–891.
Toxikos (2010). Human Health Risk Assessment for the use of Brodifacoum
for the Lord Howe Island Rodent Eradication Plan. Unpublished Report
for the Lord Howe Island Board. Caulfield East, Vic., Australia: Toxikos
Pty Ltd.
Gillespie, R. and Bennet, J. (2017). ‘Costs and benefits of rodent
eradication on Lord Howe Island, Australia’. Ecological Economics
140: 215–224.
TSSC (2006) Commonwealth Listing Advice on Predation by Exotic
Rats on Australian Offshore Islands of Less than 1000 km2 (100,000
ha). Threatened Species Scientific Committee. Canberra: Australian
Government Department of the Environment.
Groombridge, B. (1992). Global Biodiversity: Status of the Earth’s Living
Resources: A Report. Report prepared for the World Conservation
Monitoring Centre, British Museum (Natural History), and International
Union for Conservation of Nature and Natural Resources. London:
Chapman and Hall.
Wilkinson, I.S. and Priddel, D. (2011). ‘Rodent eradication on Lord Howe
Island: challenges posed by people, livestock, and threatened endemics’.
In: C.R. Veitch, M.N. Clout and D.R. Towns (eds.) Island invasives:
eradication and management, pp. 508–514. Occassional Paper SSC no.
42. Gland, Switzerland: IUCN and Auckland, New Zealand: CBB.
Hutton, I., Brice, L., Kelly, D. and Murray, C. (2016) The Social Aspect
to Successful Island Conservation Programs: Lessons Learnt from the
Proposed Rodent Eradication on Lord Howe Island. Paper Presented at
Island Arks IV Conference, Norfolk Island, 2016.
IUCN (2017). ‘World Heritage Outlook Lord Howe Island Group’.
<https://www.worldheritageoutlook.iucn.org/node/990>.
Jones, H.P., Tershy, B.R., Zavaleta, E.S., Croll, D.A., Keitt, B.S.,
Finkelstein, M.E. and Howald, G.R. (2008). ‘Severity of the effects of
invasive rats on seabirds: a global review’. Conservation Biology 22:
16–26.
LHIB. (2009). Draft Lord Howe Island Rodent Eradication Plan. NSW:
Lord Howe Island Board.
LHIB. (2016). Lord Howe Island Rodent Eradication Project: Public
Environment Report. EPBC 2016/7703. Prepared by Lord Howe Island
Board. NSW: Lord Howe Island Board. <http://www.lhib.nsw.gov.
au/sites/lordhowe/files/public/images/documents/lhib/Environment/
Rodent%20Eradication/LHI%20REP%20Final%20Public%20
Environment%20Report_21Dec2016.pdf>.
Lord Howe Island Tourism Association, 2015. Lord Howe Island
Destination Management Plan 2014–2017. Unpublished report for
Destination NSW. Lord Howe Island: Lord Howe Island Tourism
Association.
Make Stuff Happen. 2013. Community Engagement Report and Plan.
Unpublished report for the Lord Howe Island Board. Annandale, New
South Wales: Make Stuff Happen Pty Ltd.
Milledge, D. (2010). Research to Inform the Eradication of the Introduced
Masked Owl Population on Lord Howe Island. Unpublished report.
Landmark Ecological Services.
NSW Scientific Committee, (2000). Final Determination to List the
Predation by the Ship Rat Rattus rattus on Lord Howe Island as a KEY
THREATENING PROCESS. <http://www.environment.nsw.gov.au/
determinations/ShipRatLordHoweKTPListing.htm>. Accessed 8 June
2016.
Parkes, J., Ruscoe, W., Fisher, P. and Thomas, B. (2004). Benefits,
Constraints, Risks and Costs of Rodent Control Options on Lord Howe
Island. Unpublished report to the Lord Howe Island Board. Lincoln,
New Zealand: Landcare Research.
Pickard, J. (1983). ‘Vegetation of Lord Howe Island’. Cunninghamia 1:
133–265.
Ramboll Environ (2017), Human Health Risk Assessment – Proposed
Lord Howe Island Rodent Eradication Program. Unpublished report for
the NSW Office of the Chief Scientist. The Junction, NSW, Australia:
Ramboll Environ Pty Ltd.
Russell, J.C. and Broome, K.G. (2016) ‘Fifty years of rodent eradications
in New Zealand: Another decade of advances’. New Zealand Journal of
Ecology 40(2): 197–204.
Russell, J., Taylor, C.N. and Aley, J. (2018). ‘Social assessment of inhabited
islands for wildlife management and eradication’. Australasian Journal
of Environmental Management 25(1): 24–42.
Saunders, A. and Brown, D. (2001). An Assessment of the Feasibility of
Eradicating Rodents from the Lord Howe Island Group. Unpublished
report to the Lord Howe Island Board. Hamilton, New Zealand:
Endangered Species Recovery Council.
Shiels, A.B., Pitt W.C., Sugihara R.T., Witmer G.W. (2014). ‘Biology
and impacts of Pacific island invasive species 11. The black rat, Rattus
rattus (Rodentia, Muridae)’. Pacific Science 68(2): 145–184.
530
Y. Bedolla-Guzmán, F. Méndez-Sánchez, A. Aguirre-Muñoz, M. Félix-Lizárraga, A. Fabila-Blanco, E. Bravo-Hernández, A. Hernández-Ríos, M. Corrales-Sauceda, A. Aguilar-Vargas, A. AztorgaOrnelas, F. Solís-Carlos, F. Torres-García, L. Luna-Mendoza, A. Ortiz-Alcaraz, J. Hernández-Montoya, M. Latofski-Robles, E. Rojas-Mayoral and A. Cárdenas-Tapia
Bedolla-Guzmán, Y.; F. Méndez-Sánchez, A. Aguirre-Muñoz, M. Félix-Lizárraga, A. Fabila-Blanco, E. Bravo-Hernández, A. Hernández-Ríos, M. Corrales-Sauceda, A. Aguilar-Vargas, A. AztorgaOrnelas, F. Solís-Carlos, F. Torres-García, L. Luna-Mendoza, A. Ortiz-Alcaraz, J. Hernández-Montoya, M. Latofski-Robles, E. Rojas-Mayoral and A. Cárdenas-Tapia. Recovery and current status of
seabirds on the Baja California Pacific Islands, Mexico, following restoration actions
Recovery and current status of seabirds on the Baja California Pacific
Islands, Mexico, following restoration actions
Y. Bedolla-Guzmán, F. Méndez-Sánchez, A. Aguirre-Muñoz, M. Félix-Lizárraga, A. Fabila-Blanco,
E. Bravo-Hernández, A. Hernández-Ríos, M. Corrales-Sauceda, A. Aguilar-Vargas, A. Aztorga-Ornelas,
F. Solís-Carlos, F. Torres-García, L. Luna-Mendoza, A. Ortiz-Alcaraz, J. Hernández-Montoya,
M. Latofski-Robles, E. Rojas-Mayoral and A. Cárdenas-Tapia
Grupo de Ecología y Conservación de Islas, A.C., Moctezuma 836, Zona Centro, 22800, Ensenada, Baja California,
Mexico. <yuliana.bedolla@islas.org.mx>.
Abstract The Baja California Pacific Islands, Mexico, are globally important breeding sites for 22 seabird species and
subspecies. In the past, several populations were extirpated or reduced due to invasive mammals, human disturbance,
and contaminants. Over the past two decades, we have removed invasive predators and, for the last decade, we have
been implementing a Seabird Restoration Programme on eight groups of islands: Coronado, Todos Santos, San Martín,
San Jerónimo, San Benito, Natividad, San Roque, and Asunción. This programme includes monitoring; social attraction
techniques; removal of invasive vegetation; reducing human disturbance; and an environmental learning and biosecurity
programme. Here, we summarise historical extirpations and recolonisations during the last two decades of restoration
actions, and we update the status of breeding species after more than a decade. To date, from 27 historically extirpated
populations, 80% have returned since the first eradication in 1995. Social attraction techniques were key in recolonisations
of Cassin’s auklet (Ptychoramphus aleuticus), royal tern (Thalasseus maximus), and elegant tern (T. elegans). A total of
19 species breed on these islands, four more species than a decade ago, including 12 new records. The most abundant
seabirds, black-vented shearwater (Puffinus opisthomelas), Cassin’s auklet, western gull (Larus occidentalis), and
Brandt’s cormorant (Phalacrocorax penicillatus), have shown a remarkable population increase. Current threats include
the potential reintroduction of invasive mammals, guano mining, recreational activities, pollution, and commercial
fisheries. To maintain these conservation gains in the long-term it is necessary to continue implementing restoration
actions and reinforcing protection on these important natural protected areas.
Keywords: conservation, invasive mammal eradications, population status, seabird recovery, social attraction techniques,
threats
INTRODUCTION
Mexican islands and their surrounding waters are key
breeding and foraging sites for one-third of all seabird
species worldwide, placing Mexico as the third most
diverse country and the second in terms of endemism
(Croxall, et al., 2012). In particular, the Baja California
Pacific Islands (Fig. 1), influenced by the productive waters
of the California Current System, support more than a
million breeding pairs of 22 seabird species and subspecies
(Wolf, et al., 2006). Unfortunately, on these islands at
least 18 seabird populations were extirpated, several
more diminished from their former abundances, and the
Guadalupe storm-petrel (Oceanodroma macrodactyla) is
presumed extinct due to the presence of invasive mammals,
human disturbance, and contaminants that affected their
breeding grounds during the last two centuries (Everett &
Anderson, 1991; McChesney & Tershy, 1998; Wolf, et al.,
2006).
et al., 2006, Whitworth pers. comm., 2007). In order to attract
birds back and improve recolonisation rates, we initiated,
in 2008, a Seabird Restoration Programme that includes
monitoring, implementing social attraction techniques used
successfully elsewhere (Jones & Kress 2012), removal of
introduced vegetation for habitat enhancement, and an
environmental learning and biosecurity programme with
local communities (Aguirre-Muñoz, et al., 2011, 2016).
Over the last decade, we have recorded several positive
Over the past two decades, we have removed 60
populations of invasive mammals from 39 islands in
Mexico, in collaboration with government agencies,
academic institutions, fishing cooperatives and a donor
network (Aguirre-Muñoz, et al., 2011, 2016, 2018). In the
Baja California Pacific, 12 islands smaller than 1,000 ha
are now free of invasive predators; 24 populations of cats
(Felis catus), goats (Capra hircus), rabbits (Oryctolagus
cuniculus), donkeys (Equus asinus), dogs (Canis lupus
familiaris), ship rats (Rattus rattus), and deer mice
(Peromyscus eremicus cedrosensis) were eradicated
between 1995–2004 and 2013 (Aguirre-Muñoz, et al.,
2011, 2016). Only a small population of house mice (Mus
musculus) remains on Coronado Sur Island, and whitetailed antelope squirrels (Ammospermophilus leucurus) on
Natividad Island (Aguirre-Muñoz, et al., 2016).
Seabird surveys, carried out on some of these islands a
few years after the eradications, recorded a low number of
natural recolonisations (Palacios pers. comm., 2003, Wolf,
Fig. 1 The Baja California Pacific islands where seabird
restoration actions have been done.
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 531–538. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
531
Island invasives: scaling up to meet the challenge. Ch 3C Strategy: Outcomes
outcomes after the implementation of these restoration
actions that have not been documented yet. Moreover, the
last comprehensive compilation of the status of seabird
breeding populations on these islands was made more than
10 years ago and needs to be updated (Wolf, et al., 2006).
Here, we summarise historical seabird extirpations
for each island and subsequent recolonisations after the
implementation of restoration actions during the last two
decades and update the status of all breeding species on
eight islands groups: Coronado, Todos Santos, San Martín,
San Jerónimo, San Benito, Natividad, San Roque, and
Asunción.
METHODS
We used historical records of breeding seabirds from
published and grey literature to determine the number of
extirpated populations and compare the status of seabird
populations after the implementation of restoration
actions (invasive mammal eradication, social attraction
techniques). Current information derives from our own
seabird censuses and estimations conducted in 2008–2017
on San Roque and Asunción islands, in 2013–2017 on
Coronado, Todos Santos, San Martín, San Jerónimo, and
Natividad, and in 2016–2017 on San Benito islands. For
surface-nesting species, we surveyed active nests from
land-based vantage points, complemented with boat counts
and searches around the islands, every 15 days during the
whole breeding season. All colonies were mapped and
divided into sub-colony sites to increase count accuracy.
For burrow-nesting species, we conducted a continuous
exhaustive and intensive search of active nests in all
potential breeding sites. On islands with accessible nesting
sites, we conducted a census all around and across the
whole island and checked nest content using a hand-lamp
or a borescope. For those species with high nest density
such as western gull (Larus occidentalis) on Todos Santos
Islands, Cassin’s auklet (Ptychoramphus aleuticus) on San
Jerónimo Island, and black-vented shearwater (Puffinus
opisthomelas) on Natividad Island, we estimated nest
densities during peak incubation, counting nests within
circular and square plots randomly distributed and
georeferenced. Burrow occupancy was determined by
recording apparently occupied burrows, i.e. with signs
of activity such as guano, feathers, clear entrances, and
footprints (Walsh, et al., 1995). Population size was
calculated through Bayesian statistics using the total
number of nests and occupied burrows (McCarthy, 2007).
We included in our counts pairs nesting within artificial
colonies installed on all the islands (Table 1).
We analysed the number of recolonisations of extirpated
colonies by island and seabird group (surface-nesting
species and burrow-nesting species). A recolonisation rate
was not possible to estimate as post-eradication surveys
were not systematic on many islands until we started our
monitoring in 2008. We also present a brief account for
each currently breeding species, where we include the
maximum number of breeding pairs estimated during our
own survey period on each island, except for storm-petrels
on San Benito Islands.
Table 1 Social attraction techniques implemented on seabird populations on the Baja California islands, Mexico, from
2008 to 2017. Y = Yes, N = No. *Mirrors were used from 2008–2011.
Species
Heermann’s gull
Elegant tern
Island
San Roque
San Roque
Asunción
Brandt’s cormorant
Coronado Norte
Coronado Sur
Todos Santos Sur
Double-crested cormorant Coronado Norte
Coronado Sur
Todos Santos Sur
Pelagic cormorant
Todos Santos Sur
Cassin’s auklet
Coronado Norte
Coronado Sur
Todos Santos Sur
Todos Santos Norte
San Martín
San Jerónimo
Natividad
San Roque
Asunción
Scripps´s murrelet
Todos Santos
Year
2008–2017
2008–2017
2008–2017
2014–2017
2014–2017
2014–2017
2014–2017
2014–2017
2014–2017
2016–2017
2015–2017
2014–2017
2014–2017
2016–2017
2014–2017
2014–2017
2014–2017
2014–2017
2014–2017
2016–2017
Social attraction techniques
Decoys, acoustic playbacks
Decoys, acoustic playbacks, mirrors*
Decoys, acoustic playbacks, mirrors*
Decoys, acoustic playbacks
Decoys, acoustic playbacks
Decoys, acoustic playbacks
Decoys, acoustic playbacks
Decoys, acoustic playbacks
Decoys, acoustic playbacks
Decoys
Artificial burrows, acoustic playbacks
Artificial burrows, acoustic playbacks
Artificial burrows, acoustic playbacks
Artificial burrows, acoustic playbacks
Artificial burrows, acoustic playbacks
Artificial burrows, acoustic playbacks
Artificial burrows, acoustic playbacks
Artificial burrows, acoustic playbacks
Artificial burrows, acoustic playbacks
Artificial burrows, acoustic playbacks
Successful
Y
Y
N
N
N
Y
N
N
Y
N
N
Y
Y
Y
N
Y
Y
Y
Y
Y
Black storm-petrel
Coronado Norte
2015–2017
Artificial burrows, acoustic playbacks
N
Coronado Sur
2015–2017
Artificial burrows, acoustic playbacks
N
Coronado Norte
2017–2017
Artificial burrows, acoustic playbacks
N
Ashy storm-petrel
532
Coronado Sur
2015–2017
Acoustic playbacks
N
Todos Santos Sur
2016–2017
Artificial burrows, acoustic playbacks
N
Bedolla-Guzmán,et al.: Seabirds on Baja California islands
Study area
The eight island groups are located on the continental
shelf off the west coast of the Baja California Peninsula,
Mexico within 66 km of the coast (Fig. 1). Their climate
is Mediterranean to desert-like. The northern islands are
characterised by subarctic waters throughout the year while
a tropical-subtropical domain persists during summer and
autumn in the southern islands (Durazo & Baumgartner,
2002; Durazo 2009, 2015). Natividad (736 ha), San Roque
(35 ha), and Asunción (41 ha) islands were designated
as part of the El Vizcaíno Biosphere Reserve in 1988
(CONANP, 2000) while Coronado (173 ha), Todos Santos
(118 ha), San Martín (265 ha), San Jerónimo (48 ha) and
San Benito (541 ha) were recently included in the Islas
del Pacífico de la Península de Baja California Biosphere
Reserve in 2016 (DOF, 2016).
RESULTS AND DISSCUSION
Fig. 2 Number of seabird colonies historically extirpated
and restored after two decades of restoration actions on
the Baja California Pacific islands, Mexico.
Recovery and status
In total, according to historical records, 27 seabird
populations were extirpated from the 12 coastal Baja
California Pacific islands where restoration actions were
conducted; Todos Santos, San Martín, and San Roque
islands were the most affected islands, with between five
to six taxa extirpated on each island. In contrast, San
Benito Islands have no historical record of any extirpation
(Table 2, Fig. 2). Extirpated species included five burrownesting species: Leach’s storm-petrel (Oceanodroma
leucorhoa), black storm-petrel (O. melania), Scripps’s
murrelet (Synthliboramphus scrippsi), Craveri’s murrelet
(S. craveri) and Cassin’s auklet; and five surface-nesting
species: brown pelican (Pelecanus occidentalis), doublecrested cormorant (Phalacrocorax auritus), Brandt’s
cormorant (Phalacrocorax penicillatus), royal tern
(Thalasseus maximus), and elegant tern (T. elegans) (Table
2). Burrow-nesting species lost 15 breeding populations
of which 9 colonies corresponded to Cassin’s auklet and
Scripps’s murrelet that were extirpated from almost all
their historical breeding sites in Mexico (Table 2, Fig.
3). Similarly, 12 colonies of surface-nesting species
were extirpated, with brown pelican and double-crested
cormorant being the most impacted species (Fig. 3).
Fig. 3 Number of colonies historically extirpated and
restored after two decades of restoration actions of each
seabird species on the Baja California Pacific islands,
Mexico.
After two decades of restoration actions, in total, 22
colonies of extirpated seabirds have returned to breed
to these islands, which represent 80% of all extirpated
colonies. San Martín and San Roque islands are the islands
that have benefited the most as currently all extirpated
species are breeding again on these islands. Likewise, the
species with more colonies extirpated are now breeding
on almost all their historic sites (Table 2, Fig. 3). Social
attraction techniques were key in recolonisations of
Cassin’s auklet on Natividad Island; and royal tern and
elegant tern on San Roque Island (Table 1).
Moreover, we have recorded 12 new colonisations
during the last decade that have never been recorded before,
five of them on San Jerónimo Island (Table 2). These
new records together with recolonisations have increased
considerably the number of breeding taxa on many islands
in comparison with the last comprehensive compilation
(Wolf, et al., 2006). For instance, San Jerónimo Island with
only four species recorded last decade now supports 12
breeding species (Fig. 4). Currently, breeding seabirds on
these 12 islands comprise 19 species, four more species
than the last record (Wolf, et al., 2006).
The most abundant seabird is the black-vented
shearwater, which has a total population an order of
magnitude higher than all other species, but is restricted to
three breeding sites (Natividad, San Benito and Guadalupe
islands). Cassin’s auklet, western gull, Brandt’s cormorant,
Fig. 4 Number of seabird species recorded a decade ago
in the last comprehensive compilation (Wolf, et al., 2006)
and during our monitoring on the Baja California Pacific
islands, Mexico.
and double-crested cormorant are relatively abundant and
have a wide distribution. In contrast, elegant tern and bluefooted booby (Sula nebouxii) only have one pair nesting in
one site (Table 2).
533
534
e
E
8
2
Total breeding taxa
Total extirpated taxa
g
7501,250$f
225
Cassin's auklet
Craveri's murrelet
Guadalupe murrelet
Scripps's murrelet
Elegant tern
Royal tern
Caspian tern
Western gull
Heermann's gull
2
7
E
15
#
133
e
1
6
B
e
135
2d
0
8
14*
22
364
3
2
4
f
B
B
PEfi
150
e
20e
0Beh
2
6
12
19
2,248–
3,691
15
53
1
5
B
PEfj
25125$PEfi
1500
e
336e
0
8
20
90
#
5,846–
9,598
13
732
e
1
6
0
10
#
PB
PB
186*
1382
335
791
500– 136
2,500Efl
1m
25–
125PEfi
300
Ee
520Eel
376
PB*%
%
20f
L
f
f
SJ
299*
PB*
C
PB
9
#
80*
49*
2,442
1*
833
1
4
0
12
30,000 50,000110,000
50–
250PEfi
250
Ef
7d
150d 164
238
204
C
PB*
SM
Eek
L
Pelagic cormorant
32
93e
20*
C
1,200
TSS
100d
E
E
e
L
Brandt's cormorant
Ee
E
55*
C
TSN
150
230d 187
E
e
L
168d
30
1
C
328
CS
Double-crested
cormorant
33
d
Ec
L
1*
203
78
PB
PE
C
Blue-footed booby
Brown pelican
d
CN
1,818
Bc
Black storm-petrel
Least storm-petrel
PB
Ashy storm-petrel
b
Ea
L
Leach’s storm-petrel
Black-vented shearwater
Species
SB
0
13
37,667
PBf
125500f&
575
f
100f
79f
63f
197
f
f
135,000f
260,000f
510,000f
125–
600f
L
#
0
13
B
B
174&
1,010
109
18
40
101
B
B
B
124
C
n
NA
C
2
5
E
o
PEm
2,500–
5,000f
750f
57f
37–75f
0
7
10
1*
2,746
3,504
434
366
76,570 110,000120,000
L
SR
113
99
C
4
5
E
o
BPEm
El
El
B
f
34f
0
9
1,659
4
1
870
1,749
42
<100f 5,802
Ef
<10f
L
AS
131
270
C
3
4
E
o
BPEm
El
Bf
1
6
2,128
1
E
1,373
25–50f 5,200
10f
Ef
L
Table 2 Last and current status of the seabird breeding populations on the Baja California Pacific Islands, Mexico. CN: Coronado Norte, CS: Coronado Sur, TSN:
Todos Santos Norte, TSS: Todos Santos Sur, SM: San Martín, SJ: San Jerónimo, SB: San Benito, NA: Natividad, SR: San Roque, AS: Asunción. L = mostly
estimates from Wolf, et al., 2006; C = our own surveys: maximum numbers recorded during the monitoring period on each island. Estimates are number of breeding
pairs. Historical extirpations are indicated. B: Breeder, PB: Probable breeder, E: Extirpated, PE: Possibly extirpated, U: Unknown. *New record, #Intensive nest
search on the island, $Estimation for the whole archipelago with spotlight survey, %Adults with fully developed brood patch were captured with mist-nest, &Scripps’s
and Guadalupe murrelets were considered together, a) Van Rossem, 1915; b) Carter, et al., pers. comm. 1996; c) Ainley & Everett, 2001; d) Whitworth, et al., pers.
comm.,2007; e) Palacios, et al., pers. comm. 2003; f) Wolf, et al., 2006; g) Jehl, 1977; h) Van Denburgh, 1924; i) Jehl & Bond, 1975; j) Everett, 1989; k) Anderson &
Keitt, 1980; l) Everett & Anderson, 1991; m) Whitworth et al. 2018. n) Keitt, et al., 2003; o) McChesney & Tershy, 1998.
Island invasives: scaling up to meet the challenge. Ch 3C Strategy: Outcomes
Bedolla-Guzmán,et al.: Seabirds on Baja California islands
Black-vented shearwater (Puffinus opisthomelas)
In the past, the breeding population of black-vented
shearwater declined on Natividad Island, its main breeding
colony worldwide due to predation by feral cats and habitat
destruction (Keitt, et al., 2002, 2003). At present, we
estimate a population of 110,000–120,000 breeding pairs,
which indicates almost a twofold increase in relation to the
last estimation two decades ago (Keitt, et al., 2003). The
small population on San Benito Islands of around 100 pairs
remains almost unchanged since its last record (Wolf, et
al., 2006, Table 1).
Leach’s storm-petrel (Oceanodroma leucorhoa)
Leach’s storm-petrel, considered the most abundant
species in the region, currently breeds only on Islote Medio
in Coronado Islands and on San Benito Islands (Wolf, et
al., 2006), however, its population estimate has not been
updated yet. In the last century, it was extirpated from
Coronado Norte Island by feral cats (Grinnell & Daggett,
1903, van Rossem, 1915), our surveys indicate that the
species is possibly extirpated. In 2016, we captured adults
with brood patches using mist nets on San Jerónimo Island,
thus, we consider this species as a probable breeder but we
have not found active nests.
Ashy storm-petrel (Oceanodroma homochroa)
The Coronado Islands are considered the southernmost
breeding range of the ashy storm-petrel and the only
breeding site in Mexico (Ainley, 1995). This species was
considered as a probable breeder on Coronado Norte Island
(Jehl, 1977) and at present, our surveys indicate the same,
although we have not found an active nest. In the last
decade, Islote Medio, an islet historically pest-free, was
the only confirmed site with a small breeding population
(Wolf et al., 2006, Carter, pers. comm. 2006). However,
we recently confirmed this species breeding on Todos
Santos Islands. In 2014, we found the first active nest on
Todos Santos Sur Island, and we corroborated species
identification by measuring adults captured using mist-nets
at night; broadcasting responses in the nest; and carrying
out genetic analyses (GECI unpubl. data). We captured
five adults with brood patches on San Martín Island, which
indicates that the breeding range of the ashy storm-petrel
is probably expanding or is wider that was recorded before
(Table 2).
Black storm-petrel (Oceanodroma melania)
In the Mexican Pacific, black storm-petrels nest
exclusively on Coronado and San Benito islands (Ainley
& Everett, 2001). On Coronado Islands, this species
is recorded as breeding on Coronado Norte, Coronado
Medio, and Islote Medio, and as extirpated on Coronado
Sur, with a total estimated population of 100–150 breeding
pairs (Grinnell & Daggett, 1903; Osburn, 1909; Sowls,
et al., 1980; Ainley & Everett, 2001; Carter, pers. comm.
2006). In 2016, we found 120 breeding pairs breeding
on Coronado Norte and Islote Medio, and one nest on
Coronado Sur, which indicates this species recolonisation
of this island (Table 2). On San Martín Island, we captured
adults with fully developed brood patches in 2017, thus,
we consider this species as a probable breeder, but we have
not found an active nest yet. Its breeding population size on
San Benito Islands has not been updated yet since the last
estimate in 1999 (Wolf, et al., 2006).
Brown pelican (Pelecanus occidentalis)
Currently, the brown pelican breeds on all its historical
breeding sites on these islands, except on Todos Santos
Norte Island, and 40% of its breeding population is
concentrated on Todos Santos Sur Island (Table 2). San
Jerónimo Island, previously not recorded as a breeding
site (Wolf et al., 2006), recently supports a population of
around 300 pairs (Table 2). This species was one of the
most affected by organochlorines and human disturbance;
colonies of thousands or hundreds of pairs recorded in
the last century were dramatically reduced on Coronado
(ca. 5,000 pairs; Jehl, 1973, Gress, 1995), and San Benito
islands (~1,000 pairs, Everett & Anderson, 1991; Wolf, et
al., 2006), and was extirpated on Todos Santos (Everett
& Anderson, 1991; Palacios & Mellink, 2000; Palacios,
pers. comm. 2003), San Martín (ca. 1000 pairs; Jehl,
1973; Anderson & Keith 1980), and Asunción islands
(Anthony, 1925; Wolf, et al., 2006). All colonies, except on
Todos Santos Islands, recovered considerably after these
threats were mitigated (Palacios & Mellink, 2000; Wolf,
et al., 2006; Whitworth, pers. comm. 2007). At present,
no declined colony has reached its historical numbers as
the species’ population size remains in the hundreds of
pairs, except on Todos Santos Sur Island that supports a
population of more than 1,000 pairs (Table 2).
Blue-footed booby (Sula nebouxii)
We recorded one nest of blue-footed booby with two
chicks on San Jerónimo Island in September 2016. This
record represents the first on the Baja California Pacific
islands and the northernmost breeding range for this
species that was previously considered on Midriff Islands,
in the Gulf of California (Hernández Díaz & Salazar
Gómez, 2011).
Double-crested cormorant (Phalacrocorax auritus)
During the last century, the double-crested cormorant
was extirpated from Todos Santos Norte (Van Denburgh,
1924; Palacios pers. comm. 2002), San Martín (Everett
& Anderson, 1991), and San Roque islands (Wolf, et al.,
2006). Currently, this species breeds on all 12 islands,
except on Todos Santos Norte Island where it remains
extirpated (Table 2). In the past, San Martín Island
supported the largest colony in North America with
hundreds of thousands of pairs (Wright, 1913; Gress, et
al., 1973; Carter, et al., 1995); after the main threats were
removed, the colony increased from zero to around 600
pairs (Palacios & Mellink, 2000, Palacios, pers. comm.
2003) and, at present, this island sustains the biggest
population in the region with about 800 pairs (Table 2).
The breeding colony on the Coronado Islands declined
from thousands to hundreds of pairs (Howell, 1917; Gress,
et al., 1973, Carter, et al., 1995), and on San Roque Island
from thousands to zero pairs (Townsend, 1923; Huey,
1927; Wolf, et al., 2006). We recorded double-crested
cormorant recolonisation on San Roque in 2008. These
colonies have remained in the hundreds of pairs during the
last two decades (Carter, pers. comm. 1996; Palacios, pers.
comm. 2003; Whitworth, comm. pers. 2007; Table 2).
The small colony recorded on Natividad in 2000
of around 60 nests (Wolf, et al., 2006), at present, has a
sevenfold increment in population size (Table 2). On Todos
Santos Sur Island the colony doubled its size and social
attraction techniques were successful with a record of eight
nests within an artificial colony.
Brandt’s cormorant (Phalacrocorax penicillatus)
Brandt’s cormorant returned to nest at all its historical
breeding sites on the Baja California Pacific Islands (Table
2). In the past, it was extirpated from San Martín Island,
the main breeding site with several thousand nests (Wright,
1913; Everett & Anderson, 1991; Palacios & Mellink,
2000; Palacios, pers. comm. 2003), and from San Jerónimo
535
Island invasives: scaling up to meet the challenge. Ch 3C Strategy: Outcomes
Island (Wolf, 2002). At present, both islands maintain
colonies of hundreds of pairs, and the largest colonies
with > 3,000 pairs are located on the southernmost islands,
Natividad, San Roque, and Asunción that before had low
numbers of pairs (Table 2; Wolf, et al. 2006). In total, the
breeding population of Brandt’s cormorant has increased
nine times more than the last decade from around >1,000
pairs to > 10,000 pairs (Table 2; Wolf, et al. 2006).
Pelagic cormorant (Phalacrocorax pelagicus)
Coronado and Todos Santos islands are considered the
southernmost breeding range for the pelagic cormorant
(Hobson, 2013). On these islands, small breeding
populations have been previously reported on all four
Coronado Islands and Todos Santos Norte Island (Palacios,
pers. comm. 2003; Carter, pers. comm. 2006, Whitworth,
pers. comm. 2007). During our monitoring in 2013–2017,
we have recorded nests on Coronado Sur, Coronado
Medio, both Todos Santos islands, and, for the first time,
one nest on San Jerónimo Island in 2017, which represents
an expansion of its breeding range to the south (Table 2).
Heermann’s gull (Larus heermanni)
Heermann’s gull breeds in small colonies (ca. 50–
100 pairs) on San Benito Medio and San Roque islands
(Table 2). On San Benito, the colony has increased from
nine nests (Jehl, 1976) to more than 100 pairs that has not
changed during the last decade (Wolf, et al., 2006, Table
2). On San Roque, previous surveys showed a population
of 35–42 pairs (Huey, 1927; Mellink, 2001). In 2008, when
we started implementing social attraction techniques, we
recorded 23 nests within the artificial colony and also
during all subsequent years. The colony has reached its
maximum number in 2017 with 42 nests (Table 2).
Western gull (Larus occidentalis)
The western gull is a species widely distributed on the
Baja California Pacific islands. There are no historical
records of extirpated colonies. It breeds on all 12 islands and
is one of the most abundant species, with a total population
estimate of approximately 20,000 breeding pairs (Table 2).
The Todos Santos Islands concentrate around 50% of the
current population. Estimates 10 years ago, showed a total
population three times smaller than today. This increase
has been very remarkable mainly on Todos Santos Sur, but
also on Todos Santos Norte, San Martín, and San Jerónimo
where colony sizes range from about 1,000 to 8,000 pairs
(Table 2).
Caspian tern (Hydroprogne caspia)
In 2013, we recorded 15 nests of Caspian tern on San
Jerónimo Island, the first record on the Baja California
Pacific islands. In 2014, the population increased to 49
pairs and we also recorded nests on San Martín Island (89
pairs). The colony on San Jerónimo decreased to 11 pairs
the last year because it was established on a California sea
lion (Zalophus californianus) resting area. In contrast, the
colony on San Martín found suitable habitat on the island
and has increased to almost 200 pairs. (Table 2).
Royal tern (Thalasseus maximus)
This species was extirpated from San Roque Island and
its last breeding record was 90 years ago (Bancroft, 1927;
Everett & Anderson, 1991). In 2017, after eight years of
the implementation of social attraction techniques, 870
breeding pairs were recorded nesting within the artificial
colony installed for elegant tern (Table 2). For the first time,
we found 80 nests of royal tern on San Jerónimo Island in
2013 but their numbers decreased rapidly to seven in 2017
(Table 2).
536
Elegant tern (Thalasseus elegans)
In the past, the elegant tern bred on San Roque and
Asunción islands (Anthony, 1925) but was extirpated
from both islands (Everett & Anderson, 1991). Currently,
in 2017, we found one pair nesting on San Roque Island
within the colony of royal tern associated with the
artificial colony. On Asunción Island, this species remains
extirpated.
Scripps’s murrelet (Synthliboramphus scrippsi)
Scripps’s murrelet was presumably extirpated from all
historical breeding sites, except Coronado and San Benito
islands (Table 2; Jehl & Bond, 1975; Everett & Anderson,
1991; Drost & Lewis, 1995; Wolf et al., 2006). At present,
it has returned to breed on all the islands where was
extirpated (Table 2). Our nest census on the islands shows
lower population sizes from Coronado to San Benito
in comparison to nocturnal surveys at-sea conducted a
decade ago (Wolf, et al., 2006; Carter, pers. comm. 2015).
We considered the last monitoring overestimated the
population size. We found four breeding pairs nesting in
artificial burrows on Todos Santos Sur Island (Table 1).
Guadalupe murrelet (Synthliboramphus hypoleucus)
Guadalupe murrelet breeds on the San Benito islands
in a low proportion in comparison with Scripps’s murrelet,
but we do not have a precise number as identification
in the nest is complicated. Colonies extirpated from
Natividad, San Roque and Asunción islands mentioned
by Wolf, et al. (2006) may have been a misidentification
of Synthliboramphus craveri (Keitt, 2005, Carter pers.
comm. 2015; Whiworth, et al., 2018). We recorded a pair
(one individual was Guadalupe and the other Scripps’s
murrelet) on a trap-camera on San Jerónimo Island but the
species’ breeding status is not confirmed yet.
Craveri’s murrelet (Synthliboramphus craveri)
Craveri’s murrelet is currently breeding in low numbers
on San Benito, Natividad, San Roque, and Asunción
islands (Table 2). Whitworth, et al. (2018) previously
found breeding evidence on San Martín, San Roque, and
Asunción islands in 2007.
Cassin’s auklet (Ptychoramphus aleuticus)
Cassin’s auklet was also extirpated from almost all of
its historical breeding sites, except from San Jerónimo and
San Benito islands (Table 2). Currently, it is breeding again
on all islands but Coronado Norte, an island that supported
a population of thousands of breeding pairs (Osborn,
1909) extirpated by cat predation (Jehl, 1977). Extirpated
colonies from San Roque and Asunción islands have been
growing from about 100 pairs in 2008, when we recorded
their recolonisation, to around 2,000 pairs in 2017 (Table
2).
Social attraction systems were key for the colonisation
on Coronado Sur and recolonisation on Natividad Island.
In 2017, we recorded 20 breeding pairs nesting in artificial
burrows on Coronado Sur Island, which represents the first
record for this island, and for the archipelago, as the last
record was on Islote Medio, a pest-free islet, three decades
ago (Everett & Anderson, 1991). After more than a century
since the last breeding record (Kaeding, 1905), in 2016, we
found five breeding pairs on Natividad Island, including
one pair inside an artificial burrow close to a sound system
(Table 1). We recorded one nest on Todos Santos Sur in
2014, and currently, more than 40 pairs are nesting on the
archipelago, 17 of them in artificial burrows (Tables 1 and
2).
Bedolla-Guzmán,et al.: Seabirds on Baja California islands
THREATS
Major historic threats to seabird colonies included
predation by invasive mammals, habitat modification,
and direct disturbance (Everett & Anderson, 1991; Wolf,
et al., 2006). Current threats are similar to a decade ago
but less extensive: 1) potential reintroduction of invasive
mammals; 2) invasive plants that reduce nesting habitat,
3) habitat modification by guano mining, 4) exploitation
of eggs, 5) disturbance by human activities including
recreation and inadequate waste management; 6) fisheries
impacts, and 7) pollution.
All these islands are now free of invasive predators
(Aguirre-Muñoz, et al., 2016), however, potential
reintroduction is high on all islands due to constant
movement from the continent, as temporal and permanent
fishermen’s camps are established on all of them, except
on the Coronado Islands. Natividad Island, inhabited by
a fishing community of about 300 people, is the most
susceptible to this threat: the reintroduction of a few pets
has been recently recorded (CONANP, comm. pers.). The
impacts of house mice that remain on Coronado Sur Island
and white-tailed antelope squirrels on Natividad Island
need to be evaluated.
Invasive plants such as ice-plant (Mesembryanthemum
crystallinum) are widely distributed on many islands
(Rebman et al., 2016), and are displacing native flora and
affecting their associated fauna. Brandt’s cormorants that
nest on clean areas along the coast line, and Cassin’s auklets
and black-vented shearwaters that breed in subterranean
burrows might be the most affected species.
Guano mining caused severe damage on San Jerónimo,
San Roque, and Asunción islands during the last century
(Everett & Anderson, 1991; Wolf, et al., 2006). At present,
this activity continues on San Jerónimo Island at least
since 2015 and is causing disturbance and destruction of
the Cassin’s auklet colony (GECI unplub. data). Human
exploitation of western gull eggs persists on San Benito
and Natividad islands. However, the impact on these
populations is unknown. In 2016, the harvesting of
Heermann’s gull eggs on the small colony of San Benito
Medio Island caused low productivity (GECI unpubl. data).
Recreation activities (surfing, kayaking, and fishing) are
a threat on Todos Santos, San Benito Oeste, and Natividad
islands. We have recorded tourists and residents walking
close to breeding colonies which could cause temporal
abandonment of nests and increase gull predation, and also
damage to the burrows of nocturnal species. On Natividad
Island, a metal fence around the town landfill built in
front of the breeding colony in 2014, severely impacted
the population. In 2016, the structure was removed by
the fishing cooperative, however, inadequate waste
management is a serious problem for the black-vented
shearwater.
Commercial fishing is one of the most important
economic activities along the Mexican Pacific coast
(CONAPESCA, 2014). Although no fisheries bycatch
impacts have yet been evaluated, information from the
Gulf of California, shows that 17 species of seabirds and
aquatic birds, mainly brown pelicans and blue-footed
boobies, are incidentally caught in nets during fishing
operations (Comunidad y Biodiversidad, A.C. pers.
comm.). The most abundant seabird species in the region
forage on small pelagic fish, thus, probable competition for
food with commercial fisheries represents a potential threat
that should be studied.
Light pollution in fishermen’s camps created an
important impact on nocturnal species decades ago (Wolf,
et al., 2006) – currently this threat has been mitigated
(GECI unpub. data) but still it is necessary to evaluate it
on fishing boats around the islands. Current information
about pollution-related threats is scarce. Oil spill hazards
are a potential threat due to the region being an important
transportation route and having fuel reception facilities.
However, there is no action plan or personnel trained to
manage oiled seabirds, and fauna in general.
Plastic consumption for variety of seabird species
is increasing worldwide (Wilcox, et al., 2015). We have
found evidence of plastic consumption in black-vented
shearwater breeding on Guadalupe Island (GECI unpubl.
data), thus, the impact of microplastics on seabirds in the
region requires evaluation.
CONSERVATION
Although 65% of all seabirds recorded breeding on
the islands are listed in the IUCN Red List of threatened
species and are protected in Mexico in the Norma Oficial
Mexicana NOM-059-SEMARNAT-2010, it is essential
to update the status of several species. The majority of
protected natural areas fully protect breeding and foraging
areas for coastal species, however, foraging sites of pelagic
seabirds are not included as most are located hundreds of
kilometres from colonies. To address this issue, we are
developing a Marine Important Bird Areas (Marine IBAs)
proposal in collaboration with government agencies,
national and international NGOs and seabird experts.
We are also in the process of publishing an action plan
for endemic seabirds that delineates the next actions to
improve their conservation status. In the short-term this
plan will incorporate all breeding seabirds in Mexico.
Regulations and surveillance enforcement to prevent
introduction of invasive species, mitigate disturbance
to colonies and impacts of fisheries are primordial.
An important step is the National Island Biosecurity
Programme, recently initiated on several islands including
San Benito Islands, which aims to involve all key
stakeholders in the protection of island environments
from invasive alien species (Latofski-Robles et al., 2019).
Moreover, it is necessary to continue working together
with local communities to raise awareness about the threats
seabirds are facing.
We consider research priorities to improve management
decisions are: 1) to continue monitoring of populations
and obtain data on productivity; 2) to obtain accurate
population estimates for nocturnal seabirds, especially for
storm-petrels; 3) to evaluate the impact of threats such as
fisheries interactions and microplastics.
ACKNOWLEDGEMENTS
We thank organisations and agencies that have
collaborated for several years in this Programme: SEGOB,
SEMARNAT, CONANP, CONABIO, SEMAR, National
Audubon Society (Steve W. Kress, Paula Shannon, Susan
Schubel), Cornell Lab of Ornithology (Eduardo ÍñigoElías), CICESE, UABC, Coop. Buzos y Pescadores de
la Baja California, Coop. Ensenada, Coop. California de
San Ignacio, Coop. Pescadores Nacionales de Abulón,
FEDECOOP. This Programme is supported by the
Montrose and Luckenbach Trustee Councils, the Alianza
WWF-Fundación Carlos Slim, FMCN, the National Fish
and Wildlife Foundation, the David and Lucile Packard
Foundation, and the Marisla Foundation.
537
Island invasives: scaling up to meet the challenge. Ch 3C Strategy: Outcomes
REFERENCES
Aguirre-Muñoz, A., Samaniego-Herrera, L., Luna-Mendoza, L., OrtizAlcaraz, A., Rodríguez-Malagón, M., Méndez-Sánchez, F., FélixLizárraga, M., Hernández-Montoya, J.C., González-Gómez, R., TorresGarcía, F., Barredo-Barberena, J.M. and Latofski-Robles, M. (2011).
‘Island restoration in Mexico: ecological outcomes after systematic
eradications of invasive mammals’. In: C.R. Veitch, M.N. Clout and
D.R. Towns (eds.) Island invasives: eradication and management, pp.
250–258. Occassional Paper SSC no. 42. Gland, Switzerland: IUCN
and Auckland, New Zealand: CBB.
Aguirre Muñoz, A., Samaniego Herrera, A., Luna Mendoza, L., Ortiz
Alcaraz, A., Méndez Sánchez, F. and Hernández Montoya, J. (2016). ‘La
Restauración Ambiental Exitosa de las Islas de México: Una Reflexión
sobre los Avances a la Fecha y los Retos por Venir’. In: E. Ceccon and
C. Martínez-Garza (eds.) Experiencias Mexicanas en la Restauración
de los Ecosistemas, pp. 487–514. Ciudad de México: CRIM-UNAM,
Universidad Autónoma del Estado de Morelos, CONABIO.
Aguirre-Muñoz, A., Bedolla-Guzmán, Y., Hernández-Montoya, J.,
Latofski-Robles, M., Luna-Mendoza, L., Méndez-Sánchez, F., OrtizAlcaraz, A., Rojas-Mayoral and Samaniego-Herrera, A. (2018). ‘The
Conservation and Restoration of the Mexican Islands, a Successful
Comprehensive and Collaborative Approach Relevant for Global
Biodiversity’. In: A. Ortega-Rubio (ed.) Mexican Natural Resources
Management and Biodiversity Conservation. pp.177–192. Springer
International Publishing AG, part of Springer Nature.
Ainley, D.G. (1995). ‘Ashy Storm-petrel (Oceanodroma homochroa)’.
In: A. Poole and F. Gill (eds.) The Birds of North America, No. 185.
Philadelphia, PA: Academy of Natural Sciences and Washington, DC:
American Ornithologists’ Union.
Ainley, D.G. and Everett, W.T. (2001). ‘Black Storm-Petrel (Oceanodroma
melania), version 2.0’. In: A.F. Poole and F.B. Gill (Eds.). The Birds of
North America. Ithaca, NY, USA: Cornell Lab of Ornithology.
Anderson, D.W. and Keith, J.O. (1980). ‘The human influence on seabird
nesting success: Conservation implications’. Biological Conservation
18: 65–80.
Anthony, A.W. (1925). ‘The expedition to Guadalupe Island, Mexico, in
1922. The birds and mammals’. Proceedings of the California Academy
of Sciences 14(4): 277–320.
Bancroft, G. (1927). ‘Notes on the breeding coastal and insular birds of
coastal Lower California’. Condor 29: 188–195.
Carter, H.R., Sowls, A.L., Rodway, M.S., Wilson, U.W., Lowe, R.W.,
McChesney, G.J., Gress, F. and Anderson, D.W. (1995). ‘Population
size, trends, and conservation problems of the double-crested cormorant
on the Pacific Coast of North America’. Colonial Waterbirds 18: 189–
207.
CONANP (Comisión Nacional de Áreas Naturales Protegidas). (2000).
Programa de Manejo de la Reserva de la Biósfera El Vizcaíno. Mexico
City: SEMARNAP.
CONAPESCA (Comisión Nacional de Acuacultura y Pesca). (2014).
‘Anuario estadístico de acuacultura y pesca 2014’. <https://www.
gob.mx/conapesca/documentos/anuario-estadistico-de-acuacultura-ypesca>.
Croxall, J.P., Butchart, S.H.M., Lascelles, B., Stattersfield, A.J., Sullivan,
B., Symes, A. and Taylor, P. (2012). ‘Seabird conservation status,
threats ad priority actions: a global assessment’. Bird Conservation
International 22(1): 1–34.
Van Denburgh, J. (1924). ‘The birds of Todos Santos Islands’. Condor
26: 67–71.
DOF. Diario Oficial de la Federación. (2016). Decreto por el que se
declara Área Natural Protegida, con el carácter de reserva de la
biosfera, la región conocida como Islas del Pacífico de la Península de
Baja California, Diario Oficial de la Federación. <http://www.dof.gob.
mx/nota_detalle.php?codigo=5464451&fecha=07/12/2016>.
Drost, C.A. and Lewis, D.B. (1995). ‘Xantu’s Murrelet (Synthliboramphus
hypoleucus)’. In: A.F. Poole and F.B. Gill (eds.) The Birds of North
America, No. 164. Philadelphia, PA: Academy of Natural Sciences and
Washington, DC: American Ornithologists’ Union.
Everett, W.T. and Anderson, D.W. (1991). ‘Status and conservation of the
breeding seabirds on offshore Pacific islands of Baja California and the
Gulf of California’. ICBP Technical Plublication 11: 115–139.
Gress, F. (1995). ‘Organochlorines, Eggshell Thinning, and Productivity
Relationships in Brown Pelicans Breeding in the Southern California
Bight’. PhD thesis. Davis, California: University of California.
Gress, F., Risebrough, R.W., Anderson, D.W., Kiff, L.F. and Jehl, J.R.
(1973). ‘Reproductive failures of double-crested cormorants in southern
California and Baja California’. Wilson Bulletin 85: 197–208.
Grinnell, J. and Daggett, F. (1903). ‘An ornithological visit to Los
Coronados Islands, Lower California’. Auk 20: 27–37.
Hernández Díaz, J.A. and Salazar Gómez, E.N. (2011). ‘Blue-footed
Booby (Sula nebouxii), Version 1.0’. In: T.S. Schulenberg (ed.)
Neotropical Birds Online. Ithaca: Cornell Lab of Ornithology. <https://
doi.org/10.2173/nb.bfoboo.01>.
538
Hobson, K.A. (2013). ‘Pelagic Cormorant (Phalacrocorax pelagicus)’.
In: P.G. Rodewald (ed.) The Birds of North America. Ithaca: Cornell
Lab of Ornithology. < https://birdsna.org/Species-Account/bna/species/
pelcor/introduction>.
Howell, A.B. (1917). ‘Birds of the islands off the coast of southern
California’. Pacific Coast Avifauna 12: 1–127.
Huey, L.M. (1927). ‘Northernmost breeding station of the Heermann Gull
on the Pacific Ocean, and other notes from San Roque Island, Lower
California’. Condor 29: 205–206.
Jehl, J.R. (1973). ‘Studies of a declining population of brown pelicans in
northwestern Baja California’. Condor 75(1): 69–79.
Jehl, J.R. (1976). ‘The northernmost colony of Heermann’s gull’. Western
Birds 7: 25–26.
Jehl, J.R. (1977). ‘An annotated list of birds of Islas los Coronados, Baja
California, and adjacent waters’. Western Birds 8: 91–101.
Jehl, J.R. and Bond, S.I. (1975). ‘Morphological variation and species
limits in murrelets of the genus Endomychura’. Transactions of the San
Diego Society of Natural History 18: 9–23.
Jones, H.P. and Kress, S.W. (2012). ‘A review of the world’s active seabird
restoration projects’. Journal of Wildlife Management 76(1): 2–9.
Kaeding, H.B. (1905). ‘Birds from the west coast of Lower California and
adjacent islands’. Condor 7: 105–111.
Keitt, B.S., Wilcox, C., Tershy, B.R, Croll, D.A. and Donlan, C.J. (2002).
‘The effect of feral cats on the population viability of black-vented
shearwaters (Puffinus opisthomelas) on Natividad Island, Mexico’.
Animal Conservation 5: 217–223.
Keitt, B., Tershy, B.R. and Croll, D.A. (2003). ‘Breeding biology and
conservation of the black-vented shearwater Puffinus opisthomelas’.
Ibis, 145: 673-680.
Keitt, B. (2005). ‘Status of Xantus’s murrelet and its nesting habitat in
Baja California, Mexico’. Marine Ornithology 33: 105–114.
Latofski-Robles, M., Méndez-Sánchez, F., Aguirre-Muñoz, A., JáureguiGarcía, C., Koleff-Osorio, P., González-Martínez, A.I., Born-Schmidt,
G., Bernal-Stoopen, J. and Rendón-Hernández, E. (2019). ‘Mexico’s
island biosecurity programme: collaborative formulation and
implementation’. In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell
and C.J. West (eds.) Island invasives: scaling up to meet the challenge,
pp. 484–488. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
McCarthy, M. A. (2007). Bayesian Methods for Ecology. New York,
USA: Cambridge University Press.
McChesney, G. and Tershy, B. (1998). ‘History and status of introduced
mammals and impacts to breeding seabirds on the California channel
and northwestern Baja California Islands’. Colonial Waterbirds 21(3):
335–347.
Mellink, E. (2001). ‘History and status of colonies of Heermann’s gull in
Mexico’. Waterbirds: The International Journal of Waterbird Biology
24(2): 188–194.
Osburn, P.I. (1909). ‘Notes on the birds of Los Coronados Islands, Lower
California’. Condor 11: 134–138.
Palacios, E. and Mellink, E. (2000). ‘Nesting waterbirds on Islas San
Martín and Todos Santos, Baja California’. Western Birds 31: 184–189.
Sowls, A.L., DeGange, A.R., Nelson, J.W. and Lester, G.S. (1980).
Catalog of California Seabird Colonies, Biological Services Program.
Washington, D.C.: U.S. Fish and Wildlife Service (FWS/OBS-80/37).
Townsend, C.H. (1923). ‘Birds collected in Lower California’. Bulletin of
the American Museum of Natural History 28: 1–26.
Van Rossem, A. (1915). ‘Notes on murrelets and petrels’. Condor 17:
74–78.
Walsh, P. M., Halley, D. J., Harris, M. P., del Nevo, A., Sim, I. M. W. and
Tasker, M. L. (1995). Seabird Monitoring Handbook for Britain and
Ireland. Norfolk, United Kingdom: The Seabird Group.
Wilcox, C., Van Sebille, E. and Hardesty, B. D. (2015). ‘Threat of plastic
pollution to seabirds is global, pervasive, and increasing’. Proceedings
of the National Academy of Sciences 112 (38): 11899–11904.
Wolf, S. (2002). ‘The Relative Status and Conservation of Island Breeding
Seabirds in California and Northwest Mexico’. MSc thesis. Santa Cruz:
University of California.
Wolf, S., Keitt, B., Aguirre-Muñoz, A., Tershy, B., Palacios, E. and Croll,
D. (2006). ‘Transboundary seabird conservation in an important North
American marine ecoregion’, Environmental Conservation, 33(4):
294–305.
Whitworth, D.L., Carter, H.R., Palacios, E. and Gress, F. (2018). Breeding
of Craveri’s Murrelet Synthliboramphus craveri at four islands off westcentral Baja California, México. Marine Ornithology 46: 117-124.
Wright, H.W. (1913). ‘The birds of San Martín Island, Lower California’.
Condor 15: 207–210.
E.A. Bell, M.D. Bell, G. Morgan and L. Morgan
Bell, E.A.; M.D. Bell, G. Morgan and L. Morgan. The recovery of seabird populations on Ramsey Island, Pembrokeshire, Wales, following the 1999/2000 rat eradication
The recovery of seabird populations on Ramsey Island, Pembrokeshire,
Wales, following the 1999/2000 rat eradication
E.A. Bell1, M.D. Bell2, G. Morgan3 and L. Morgan3
Wildlife Management International Ltd, PO Box 607, Blenheim 7201, New Zealand, <biz@wmil.co.nz>. 2Wildlife
Management International Ltd, PO Box 607, Blenheim 7201, New Zealand. 3Royal Society for the Protection of Birds,
Ramsey Island, c/o Trefeiddan Farm, St David’s, Pembrokeshire, SA62 6PY, United Kingdom.
1
Abstract Ramsey Island, 259 ha, is ca. 1 km off the Pembrokeshire coast, south-west Wales. The eradication of brown
rats (Rattus norvegicus) was successfully completed in the winter 1999/2000 using a ground-based bait station operation.
The pre-eradication survey using tape playback estimated the Manx shearwater (Puffinus puffinus) population to be
849 pairs. These surveys were repeated in 2007, 2012 and 2016. Each survey showed the Manx shearwater population
had increased, reaching 4,796 pairs in 2016 with birds spreading from previously known breeding locations. European
storm petrels (Hydrobates pelagicus) were first recorded breeding on Ramsey Island in 2008 with up to 12 pairs in 2016
(a minimum estimate based on accessible survey areas). Other species have also shown improvements to population
estimates and range since the rat eradication. This evidence shows that there can be little doubt that the presence of brown
rats on Ramsey played a significant role in suppressing breeding numbers and limiting the breeding range of seabirds on
the island and the positive results following the successful eradication are now being seen.
Keywords: brown rat, European storm petrel, Hydrobates pelagicus, Manx shearwater, Puffinus puffinus, Rattus
norvegicus, survey
INTRODUCTION
STUDY AREA AND METHODS
Rats are known to have devastating effects on seabird
and land bird populations by predation of eggs, chicks
and adults which reduces breeding success, recruitment,
population size and distribution. They have caused
extinctions of birds on numerous islands throughout the
world (Moors & Atkinson, 1984, Atkinson, 1985, Towns,
et al., 2006; Jones, et al., 2008). Smaller burrowing seabirds
are recognised as the species most affected by invasive rats
(Jones, et al., 2008; Towns, et al., 2011). The eradication of
rats from seabird islands is recognised as one of the most
important tools in avian conservation in recent times, with
significant long-term restoration benefits such as increased
productivity and populations sizes and establishment of
new, or return of previously locally extinct, seabird species
being achieved (Atkinson, 1985; Moors, et al.,1992;
Lock, 2006; Ratcliffe, et al., 2009; Booker & Price,
2010, Bourgeois, et al., 2013; Le Corre, et al., 2015).The
protection and enhancement of UK seabird breeding habitat
has been recognised as an important conservation priority,
including under international conservation agreements
(Brooke, et al., 2007; Ratcliffe, et al., 2009; Dawson, et
al., 2015; Thomas, et al., 2017). Over 400 islands around
the world have been successfully cleared of rats, including
twelve in the United Kingdom, with a subsequent increase
in bird populations (Thomas & Taylor, 2002; Towns &
Broome, 2003; Jones, et al., 2008; Howald, et al., 2007,
DIISE, 2015, Thomas, et al., 2017).
Study site
A feasibility study of eradicating brown rats (Rattus
norvegicus) from Ramsey Island was completed in 1998
and led to the ground-based eradication in autumn 1999.
Documenting the recovery of bird species on islands
that have had invasive mammals removed is becoming
increasingly important. RSPB has been monitoring
bird populations on Ramsey Island since 1992. Due to
difficulty in accessing natural burrows, between 2013
and 2016, RSPB constructed a man-made seabird habitat
using artificial burrows with the aim to establish a Manx
shearwater (Puffinus puffinus) colony that could be used
to monitor productivity, recruitment and adult survival.
This paper details the changes to the Manx shearwater
population on Ramsey Island, including within the manmade habitat, and the subsequent colonisation of the
island by European storm petrel (Hydrobates pelagicus)
following the eradication of brown rats.
Ramsey Island, 259 ha (5°20′W, 51°51′N), is located
about 1 km off the Pembrokeshire coast, south-west Wales
(Fig. 1). It is a nature reserve owned and managed by the
RSPB. Ramsey Island is approximately 3.2 km long and
1.6 km across at its widest point and is surrounded by
coastal cliffs which are particularly high and steep on the
western side of the island. There is also a number of small
islets (including a chain of islets from the southern end)
and caves around the coastline. The coastline of the island
is made up of exposed rocky shores with a small number of
sandy coves. The top of the island is gently rolling and is
dominated by two prominent peaks (Carn Ysgubor 101 m
and Carn Llundian 136 m). The island supports three main
habitats; acid grassland, bracken-dominated grassland and
coastal heathland (Doncaster, 1981; Hurford & Evans,
2006; CCW, 2008). The heathland and maritime grassland
communities are of conservation importance in Wales
(JNCC, 2001; Hurdford & Evans, 2006; CCW, 2008).
The rush-pasture fields are grazed by rabbits (Oryctolgus
cuniculus), ponies (Equus caballus) and sheep (Ovis aries)
as part of the management to support wildlife, particularly
choughs (Pyrrhocorax pyrrhocorax) (Doncaster, 1981;
Long, 2003). The bank vole (Myodes glareolus) and
common shrew (Sorex araneus) are also present on the
island.
Fig. 1 Location of Ramsey Island, Wales.
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 539–544. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
539
Island invasives: scaling up to meet the challenge. Ch 3C Strategy: Outcomes
The island is part of the Pembrokeshire Coast National
Park and has a range of designations including as a Site of
Special Scientific Interest (SSSI), National Nature Reserve
(NNR), Important Bird Area (IBA), Special Protection
Area (SPA) and Marine Special Area of Conservation
(MSAC) (JNCC, 2001; CCW, 2008; Hayhow, et al.,
2016). Ramsey holds important breeding populations of
razorbill (Alca torda), guillemot (Uria aalge), kittiwake
(Rissa tridactyla), Manx shearwater, chough and wheatear
(Oenanthe oenanthe) (JNCC, 2001; Johnstone, et al.,
2011). Ramsey Island is also an important breeding site for
Atlantic grey seals (Halichoerus grypus).
The island is popular with visitors who are interested
in the seabirds, land birds, flora and history. These visitors
travel to the island on small passenger vessels from St
David’s between April and October. There is a jetty and
several buildings on the island, including the warden’s
home and information shelter.
Manx shearwaters have been recorded on Ramsey
Island since the 18th century (Mathew, 1894; Holloway,
2010; Lovegrove, et al., 2010). Historical reports and more
recent seabird monitoring on Ramsey recorded declines
in Manx shearwaters and other seabirds (Mathew, 1894;
Humpridge & Bullock, 1999; Lovegrove, et al., 2010).
These declines were attributed in part to the presence of
brown rats and predation on eggs and chicks (Humpridge
& Bullock, 1999; Lovegrove, et al., 2010).
It is not known when brown rats became established on
Ramsey Island; but this was likely to have occurred more
than two hundred years ago from an early shipwreck.
Brown rat eradication
The eradication of brown rats was completed as a
ground-based operation using rodenticide cereal blocks in
protective bait stations to reduce risk to non-target species.
A 50 m ×50 m grid was established in autumn 1999. Bait
stations were made from 500 mm lengths of corrugated
plastic drainage pipe staked into position using wire.
A total of 1,260 stations were placed on the main island
and offshore stacks. The poisoning operation ran from 11
January 2000 to 10 March 2000. Two 24 g blocks of cerealbased rodenticide bait (Neosorexa™, active ingredient
0.005% difenacoum, manufactured by Sorex Ltd) were
placed in each station on the main island and ten blocks
on the offshore stack throughout the poisoning programme
and replaced as required when eaten by rats, non-target
species and/or damaged by weather.
The stations on the main island were checked daily,
but stations on the offshore stacks were checked when sea
conditions allowed. Bait take was recorded by bait station
number and the species believed to have consumed or
removed the bait.
From 1 March to 15 March 2000, monitoring stations
were established around the island next to and in-between
the bait stations. Chew sticks, chocolate blocks and small
pieces of candle were used. Sand and mud areas on the
island were checked for rat foot prints and burrows and
rat runs were checked for fresh activity. All monitoring
points were individually numbered and any evidence of
activity (i.e. teeth marks or foot prints) was recorded by
station number and the species believed to have consumed
or marked the monitoring item. Each monitoring site was
checked regularly, either separately or together with the
poisoning bait station grid. Any rat and non-target species
sign found on detection devices was recorded.
Manx shearwater breeding population survey
Earlier surveys on Ramsey Island had shown that Manx
shearwaters only occur in a narrow strip around the coast
540
of the island on hills Carn Llundain and Carn Ysgubor
(I.D. Bullock, unpublished data; Perkins, et al., 2017). For
burrow counting and sampling purposes, Ramsey Island
was divided into 42 sub-areas by topographical features.
A full count of suitable burrows (i.e. more than 0.7 m in
length and not doubling back to the surface) in these 42
sub-areas was completed in 1999, 2007, 2012 and 2016.
Estimation of the numbers of Manx shearwaters
on Ramsey Island was based on playback of recorded
calls (Brooke, 1978; Smith, et al, 2001; Perkins, et al.,
2017). This method relies on the fact that if a male Manx
shearwater call is played down a sample of burrows
during the incubation period, most incubating males, but
no incubating females will respond to that call (Smith, et
al., 2001). For a given number of breeding pairs, it is then
possible to establish the number of males that respond to
recorded calls. From this, using the following formula it
is possible to estimate the number of breeding pairs in the
burrows on the island.
The response rate for Manx shearwaters was calculated
by Bullock in 1999 (0.409) and was based on a study set
of 13 burrows (Humpridge & Bullock 1999). Alternative
response rates were available from Skomer (0.43, Smith,
et al., 2001) and Skokholm (0.505, Brooke, 1978) or the
seabird monitoring handbook (0.505, Walsh, et al., 1995).
The Ramsey response rate of 0.409 was used in 2007 as
it allows direct comparison to the earlier survey on the
island. The response rates were recalculated for the 2012
survey (to 0.4625) using methods developed by Murray, et
al. (2003), Newton, et al. (2004) and Perrins, et al. (2012).
The response rates were recalculated for the 2016 survey
(to 0.845) which used dual-sex calls which had been shown
to give a more reliable correction factor (Perkins, et al.,
2017).
A recording of male Manx shearwater calls was played
down 20% of burrows in each sub-area during the main
incubation period unless the sub-area contained fewer than
50 burrows up to 2016 and then duetting male and female
calls were used for 2016 (Perkins, et al., 2017). In those
cases, the recording was played in all burrows. Recordings
were played at natural volumes (‘normal’ Manx shearwater
call volumes as heard from the burrows that were set ‘by
ear’ before 2016 and by a decibel reader in 2016) within 30
cm of the burrow entrance for up to 25 seconds. Playback
of calls was carried out in the day and responses, or lack
thereof, were recorded. Playback was undertaken between
the end of May and mid-June at a time when all eggs laid
should be being incubated by one adult (Brooke, 1990).
Between 2013 and 2016 nearly 100 artificial nest
boxes were established on the island. These burrows are
the same design as those developed in New Zealand for
fluttering shearwaters (Puffinus gavia) by Bell (1995) and
recommended for burrow-nesting petrel and shearwater
species (Gummer, et al., 2014). These artificial burrows
were put in place to provide easily accessible study
burrows for tracking studies and to determine productivity
and population parameters such as survival and recruitment
(Morgan, 2012; Kirk, et al., 2013).
European storm petrel breeding population survey
Surveys of suitable storm petrel habitat (i.e. stone
walls, rock tumbles and scree) on Ramsey Island were
undertaken using playback in 2008, 2010, 2012 and 2016.
A recording of a male European storm petrel was played
close to a suspected site and a reply listened for (Ratcliffe,
et al., 1998; Gilbert, et al., 1999; Mayhew, et al., 2000).
Burrow entrances that had responses were mapped using
GPS.
Bell, et al.: Seabirds on Ramsey Island after rat eradication
RESULTS
Brown rat eradication
Bait acceptance was good, with rats accounting for 165
kg of bait consumed. As the LD50 for a 250-g brown rat is
9 g and the mean (± SE) bait take by rats was 81.6 ± 0.7 g
(3.4 ± 0.02 blocks, range 0–30 blocks) the rat population
on Ramsey Island was estimated to be between 1,850 and
5,400 rats). The bait take pattern was typical of other rat
eradication operations; very high in the immediate five to
ten days after original baiting and dropping to a relatively
low level 21 days after original baiting. Bait take dropped
to zero by day 41 after the original baiting (Fig. 2). The rats
were widely distributed across the island, but the density
was not even, as shown by the distribution of bait take (Fig.
3). Rats were present in all coastal areas and in highest
numbers within the central and northern areas of the island.
Rabbits interfered with the bait stations between days
12 and 26 of the operation, with a number of carcasses
being collected. Bait stations were modified by halving the
entrance size to prevent access by rabbits and this greatly
reduced their interference levels. Carrion crows (Corvus
corone), ravens (Corvus corax) and herring gulls (Larus
argentatus) also interfered with the stations from day 25
after the birds had learnt to reach into the stations to get
access to the bait. Eight crows and three raven carcasses
were located but no herring gull deaths were recorded.
The bait stations were further modified by extending the
length from 500 mm to 750 mm which reduced crow, raven
and gull interference to almost nil. Crows were observed
working in pairs to remove wires; one pulling the wire
out while the other stood on top of the station to hold it
in place, to get access to the bait. Access to the bait by
the bank voles could not be prevented and 30 dead voles
were found. A small captive population was maintained
during the eradication as a precaution and was released
after the poison had been removed. Voles and vole sign on
monitoring tools were recorded throughout the eradication.
Monitoring for rat presence continued for two years
after the end of the poisoning operation. No rats or sign
were detected. The rat-free status for Ramsey Island was
declared in March 2002.
The distribution of Manx shearwaters remained largely
unchanged between 1999 and 2007 censuses, but the
range spread between the 2007, 2012 and 2016 censuses
(Fig. 4). There have also been significant increases in the
population within the distribution with new areas recorded
in 2007, 2102 and 2016 that previously had no responses
recorded in 1999 (Fig. 4).
Burrow density is greatest along the west, north and
north-east coasts and on the hills (Fig. 4). Interestingly
in a section at the northern end of Ramsey Island there
was no response to the recordings despite a high number
of suitable burrows available for Manx shearwaters (n =
2,247) in 1999 or 2007. This area showed a low level of
response in 2012 and higher in 2016.
A prospecting pair of Manx shearwaters was recorded
in one of the artificial burrows in 2015. Two pairs nested
successfully in the artificial burrows in 2016 and seven
pairs were recorded incubating eggs in April 2017.
Manx shearwater breeding population survey
The number of Manx shearwater burrows on Ramsey
Island totalled 13,800 burrows in 1999, 14,970 burrows
in 2007, 12,302 burrows in 2012 and 12,319 in 2016
(Humpridge & Bullock, 1999; Morgan & Morgan, 2008;
Morgan & Morgan, 2013; Morgan & Morgan, 2017).
The Manx shearwater breeding population size
increased 3-fold and 5-fold, 8 and 17 years after the rat
eradication respectively (Table 1).
Fig. 2 Bait take by rats during the brown rat (Rattus
norvegicus) eradication on Ramsey Island, Wales,
1999/2000.
Fig. 3 Distribution of bait take during the brown rat (Rattus
norvegicus) eradication on Ramsey Island, Wales,
1999/2000.
541
Island invasives: scaling up to meet the challenge. Ch 3C Strategy: Outcomes
Table 1 The total number of burrows, response rate used, total number of burrows sampled using playback,
total number of responses and total number of breeding pairs of Manx shearwater Puffinus puffinus on
Ramsey Island between 1999 and 2016.
Year
1999
2007
2012
2016
Total number
of burrows
13,800
14,970
12,302
12,319
Response
rate used
0.409
0.409
0.4625
0.845
Total number sampled
using playback
2,760
3,190
2,788
2,860
Total number
of responses
74
208
402
941
Total number of
breeding pairs
905
2,387
3,835
4,796
Fig. 5 Location of European storm petrel breeding sites (in
black circle) on Ramsey Island, Wales, 2016.
Fig. 4 Distribution and density of Manx shearwaters
(Puffinus puffinus) from the full surveys in 1999, 2007,
2012 and 2016 on Ramsey Island, Wales.
European storm petrel breeding population survey
The first storm petrel breeding burrows were detected
in 2008 (4 pairs). By 2016, the numbers had increased to
12 breeding burrows (Table 2).
DISCUSSION
The Ramsey Island brown rat eradication was one of the
first eradications undertaken by the RSPB and a number of
important lessons were learnt that helped with the planning
and implementation of later eradications on Lundy Island,
St Agnes and Gugh, Isles of Scilly and the Shiant Isles.
542
Bait station design was adapted to suit local conditions and
local non-target species.
The Ramsey Island operation was the first islandwide eradication that used a rodenticide containing the
active ingredient difenacoum proving that this rodenticide
could be used to successfully target all brown rats on an
island. This toxin has since been used on a number of
island eradication operations in the UK and around the
world (Howald, et al., 2007; Jones, et al., 2008). Since the
eradication of rats, there have been biosecurity protocols
put in place to prevent a re-incursion on Ramsey Island and
to outline how to respond if rats are ever detected on the
island. It is important that these measures are maintained
indefinitely.
Ramsey Island has seen dramatic changes since
the removal of brown rats, not least the increase in the
distribution and density of Manx shearwaters. The number
of Manx shearwaters has multiplied by five times between
Bell, et al.: Seabirds on Ramsey Island after rat eradication
Table 2 The number of burrows of European storm petrels
Hydrobates pelagicus on Ramsey Island between 2008
and 2016.
Year
2008
2009
2010
2011
2012
2013
2014
2015
2016
Number of breeding burrows
4
Not surveyed
6
Not surveyed
5
5
5
8
12
1998 and 2016 to almost 5,000 breeding pairs, representing
a 560% increase. This proves that the brown rats were
having a significant role in suppressing the number of
breeding pairs on the island and their range across the
island and provides more evidence that invasive rats have
significant impacts on seabird populations on islands
(Atkinson & Moors, 1984; Atkinson, 1985; Towns, et al.,
2006; Jones, et al., 2008). A similar pattern was observed
on Lundy Island following the rat eradication operation in
2004 (Brown, et al., 2011). It is suspected that although
increased productivity will have occurred on Ramsey
Island, given that the Manx shearwater does not breed
until five or six years of age (Brooke, 1990), much of this
increase may be due to immigration from the extremely
large neighbouring colonies on Skomer and Skokholm.
This theory was confirmed by the capture of an adult that
had been ringed as a chick on Skomer in 1993 which was
on its way to feed a chick on Ramsey Island in 2017 (GM,
pers. obs.).
The greatest increases have occurred within the existing
sub-colonies, but there has also been expansion into new
areas. Nine sections that showed nil response in 1999
and 2007 contained breeding birds in 2012 and a further
five new sections were occupied in 2016. There is limited
habitat available on Ramsey Island away from the coastal
areas. However, restoration of drystone walls, former rabbit
warrens and artificial burrows have all provided more
nest-sites. However, the presence of rabbits may affect
the distribution of Manx shearwaters on Ramsey Island.
Competition for burrows with a small number of birds
may account for restricted range and densities in specific
locations on the island. The development of an artificial
study colony on Ramsey Island has proved successful
with up to seven birds nesting in the man-made burrows in
2017, of which five successfully fledged chicks.
European storm petrels have also started breeding on
the island for the first time since records began. Although
storm petrels are known to breed on two offshore islands,
the Bishops and Clerks (163 apparent occupied sites in
2017; G.M., pers. obs.), it was not until 2008 that they
were confirmed on Ramsey Island itself. Six birds were
recorded breeding in 2012 and this increased to 12 pairs
in 2016. It is important to note that these estimates are the
minimum number of storm petrels present on the island as
not all adults may respond to the recorded calls. This has
been shown to be the case in a number of other studies
(Insley, et al., 2002; Brown, 2006; Hounsome, et al., 2006)
and, as these studies have also shown that correction
factors for storm petrels are known to be highly variable
between sites and even between years, the use of recorded
calls and corrections have not been used to estimate the
current Ramsey Island storm petrel population. As the
population increases, an island-specific correction factor
will be calculated for Ramsey Island and used to estimate
the population size in the future. Currently, the minimum
estimate is used (i.e. the known response to taped calls).
However, the basic playback-response method is
widely used, standardised and is comparable between years
and across sites. It is also a low-impact method, completed
during the day, and provides spatial information on
breeding burrows. The storm petrel population on Ramsey
Island is likely to increase into a range of available habitat
including drystone walls, rabbit burrows and rock tumbles.
The success on Ramsey Island provided valuable
information and techniques for later eradication operations
in the UK, particularly those with important non-target
species. It also showed that ground-based eradication
techniques developed in New Zealand could be adapted
and used on islands in the UK, and Ramsey Island serves
as a good example of the significant long-term benefits that
can be achieved through short-term investment.
ACKNOWLEDGEMENTS
This project was carried out with funding from RSPB.
Noesorexa™ bait was donated by Sorex Ltd. We would like
to thank all the volunteers who helped with all stages of the
eradication and subsequent Manx shearwater and European
storm petrel monitoring. Kelvin Floyd (WMIL) provided
Figs 1 and 3. Two anonymous referees commented on this
paper.
REFERENCES
Atkinson, I.A.E. (1985). ‘The Spread of Commensal Species of Rattus
to Oceanic Islands and their Effects on Island Avifaunas’. In: P.J.
Moors (ed.) Conservation of Island Birds, pp. 35–81. ICBP Technical
Publication 3. Cambridge, UK: International Council for Bird
Preservation (ICBP).
Bell, B. (1995). ‘Translocation of Fluttering Shearwaters: Developing
a Method to Re–establish Seabird Populations’. In: M. Serena (ed.)
Reintroduction Biology of Australian and New Zealand Fauna, pp.
143–156. Australia: Surrey Beatty and Sons.
Booker, H. and Price, D. (2010). ‘Manx shearwaters on Lundy: A study of
population and distribution change from 2001 to 2008’. Journal of the
Lundy Field Society 2: 105–112.
Bourgeois, K., Ouni, R., Pascal, M., Dromzée, S., Fourcy, D., and Abiadh,
A. (2013). ‘Dramatic increase in the Zembretta Yelkouan shearwater
breeding population following ship rat eradication spurs interest in
managing a 1500-year old invasion’. Biological Invasions 15: 475–482.
Brooke, M. (1978). ‘Sexual differences in the voice and individual
vocal recognition in the Manx shearwater (Puffinus puffinus)’. Animal
Behaviour 26: 622–629.
Brooke, M. (1990). The Manx Shearwater. London: Poyser.
Brooke, M., Hilton, G. and Martins, T. (2007). ‘Prioritising the world’s
islands for vertebrate-eradication programmes’. Animal Conservation
10: 380–390.
Brown, A., Price, D., Slader, P., Booker, H., Lock, L. and Deveney, D.
(2011). ‘Seabirds on Lundy: their current status, recent history and
prospects for the restoration of a once important bird area’. British Birds
104: 139–158.
Brown, J. (2006). ‘Census of European storm-petrels Hydrobates
pelagicus on Skomer Island’. Atlantic Seabirds 8: 21–30.
CCW (Countryside Council for Wales). (2008). Core Management Plan
Including Conservation Objectives for St David’s Special Area of
Conservation. Bangor, UK: Countryside Council for Wales.
DIISE. (2015). ‘The Database of Island Invasive Species Eradications,
developed by Island Conservation, Coastal Conservation Action
Laboratory UCSC, IUCN SSC Invasive Species Specialist Group,
University of Auckland and Landcare Research New Zealand’. <http://
diise.islandconservation.org>.
Dawson, J. Oppel, S., Cuthbert, R., Homes, N., Bird, J., Butchart, S.,
Spatz, D. and Tershey, B. (2015). ‘Prioritizing islands for the eradication
of invasive vertebrates in the United Kingdom overseas territories’.
Conservation Biology 29: 143–153.
543
Island invasives: scaling up to meet the challenge. Ch 3C Strategy: Outcomes
Doncaster, C. (1981). ‘The spatial distribution of ants' nests on Ramsey
Island, South Wales’. Journal of Animal Ecology 50: 195–218.
Gilbert, G., Gibbons, D. and Evans, J. (1999). Bird Monitoring Methods
B: A Manual of Techniques for Key UK Species. Sandy, Bedfordshire,
UK: RSPB, BTO, JNCC, ITE & The Seabird Group.
Gummer, H., Taylor, G. and Collen, R. (2014). Field Guidelines for
Burrow-nesting Petrel and Shearwater Translocations—a Companion
Guide to the Seabird Translocation Best Practice Documents.
Wellington: Department of Conservation.
Hayhow, D., Burns, F., Eaton, M., Al Fulaij, N., August, T., Babey,
L., Bacon, L., Bingham, C., Boswell, J., Boughey, K., Brereton, T.,
Brookman, E., Brooks, D., Bullock, D., Burke, O., Collis, M., Corbet, L.,
Cornish, N., De Massimi, S., Densham, J., Dunn, E., Elliott, S., Gent, T.,
Godber, J., Hamilton, S., Havery, S., Hawkins, S., Henney, J., Holmes,
K., Hutchinson, N., Isaac, N., Johns, D., Macadam, C., Matthews, F.,
Nicolet, P., Noble, D., Outhwaite, C., Powney, G., Richardson, P., Roy,
D., Sims, D., Smart, S., Stevenson, K., Stroud, R., Walker, K., Webb, J.,
Webb, T., Wynde, R. and Gregory, R. (2016). State of Nature 2016. UK:
The State of Nature partnership.
Holloway, S. (2010). The Historical Atlas of Breeding Birds in Britain
and Ireland 1875–1900. London, UK: Poyser Series, A&C Black.
Hounsome, M., Insley, H., Elliott, S., Graham, K. and Mayhew, P.
(2006). ‘Monitoring European storm-petrels Hydrobates pelagicus: a
comparison of the results provided by mark/recapture and tape response
methods’. Atlantic Seabirds 8: 5–20.
Howald, G., Donlan, C., Galvan, J., Russell, J., Parkes, J., Samaniego, A.,
Wang, Y., Veitch, D., Genovesi, P., Pascal, M., Saunders, A. and Tershey,
B. (2007). ‘Invasive rodent eradication on islands’. Conservation
Biology 21: 1258–1268.
Humpridge, R. and Bullock, I. (1999). ‘Estimation of the breeding Manx
shearwater population on Ramsey Island’. RSPB Ramsey Island Annual
Report 1999–2000. St Davids, Pembrokeshire, UK: RSPB.
Hurdford, C. and Evans, S. (2006). ‘Monitoring Coastal Heaths at St
David’s’. In: C. Hurford and M. Schnieder (eds.) Monitoring Nature
Conservation in Cultural Habitats: A Practical Guide and Case Studies,
pp. 157–168. Netherlands: Springer.
Insley, H., Hounsome, M., Mayhew, P. and Chisholm, K. (2002). ‘A
preliminary comparison of population estimation by tape response and
mark/recapture methods for storm petrels Hydrobates pelagicus in two
contrasting Scottish colonies’. Ringing & Migration 21: 75–79.
JNCC (Joint Nature Conservation Committee). (2001). Ramsey and
St David’s Peninsula Coast: SPA Description. Special Protection
Areas under the EC Birds Directive. <http://jncc.defra.gov.uk/default.
aspx?page=2063>. Accessed 16 June 2017.
Johnstone, I., Thorpe, R. and Noble, D. (2011). The State of Birds in
Wales 2011. Cardiff, Wales: RSPB Cymru.
Jones, H., Tershey, B., Zavaleta, E., Croll, D., Keitt, B., Finkelstein, M.
and Howard, G. (2008). ‘Review of the global severity of the effects of
invasive rats on seabirds’. Biological Conservation 22: 16–26.
Kirk, H., Morgan, G. and Guilford, T. (2013). ‘Ramsey Island Manx
shearwater geolocation study: Field report 2012’. Birds in Wales 10:
42–47.
Le Corre, M., Danckwerts, D.K., Ringler, D., Bastien, M., Orlowski, S.,
Morey Rubio, C., Pinaud, D. and Micol, T. (2015). ‘Seabird recovery
and vegetation dynamics after Norway rat eradication at Tromelin
Island, western Indian Ocean’. Biological Conservation 185: 85–94.
Lock, J. (2006). ‘Eradication of brown rats Rattus norvegicus and black
rats Rattus rattus to restore breeding seabird populations on Lundy
Island, Devon, England’. Conservation Evidence 3: 111–113.
Long, J. (2003). Introduced Mammals of the World: their History,
Distribution and Influence. Wallingford, UK: CABI.
Lovegrove, R., Williams, I. and Williams, G. (2010). Birds in Wales.
London, UK: Bloomsbury Publishing.
Mathew, M. (1894). The Birds of Pembrokeshire and its Islands. London:
R.H. Porter.
Mayhew, P., Chisholm, K., Insley, H. and Ratcliffe, N. (2000). ‘A survey
of storm petrels on Priest Island in 1999’. Scottish Birds 21: 78–84.
Moors, P.J. and Atkinson, I.A.E. (1984). ‘Predation on Seabirds by
Introduced Animals and Factors Affecting its Severity’. In: J.P. Croxall,
P.G.H. Evans and R.W. Schreiber (eds.) Status and Conservation
of the World’s Seabirds, pp. 667–690. ICBP Technical Publication 2.
Cambridge, UK: International Council for Bird Preservation (ICBP).
Moors, P., Atkinson, I. and Sherley, G. (1992). ‘Reducing the rat threat to
island birds’. Bird Conservation International 2: 93–114.
544
Morgan, G. and Morgan, L. (2008). ‘A census of the Manx shearwater
Puffinus puffinus on Ramsey Island RSPB Nature Reserve,
Pembrokeshire in 2007’. RSPB Ramsey Island Annual Report 2007–
2008. St Davids, Pembrokeshire, UK: RSPB.
Morgan, G. (2012). ‘The bird populations of Ramsey and Grassholm’.
British Birds 105: 716–732.
Morgan, G. and Morgan, L. (2013). ‘A census of the Manx shearwater
Puffinus puffinus on Ramsey Island RSPB Nature Reserve,
Pembrokeshire in 2012’. RSPB Ramsey Island Annual Report 2012–
2013. St Davids, Pembrokeshire, UK: RSPB.
Morgan, G. and Morgan, L. (2017). ‘A census of the Manx shearwater
Puffinus puffinus on Ramsey Island RSPB Nature Reserve,
Pembrokeshire in 2016’. RSPB Ramsey Island Annual Report 2016–
2017. St Davids, Pembrokeshire, UK: RSPB.
Murray, S., Shewry, M., Mudge, G. and Spray, S. (2003). ‘A survey of
Manx shearwaters Puffinus puffinus on Rum, Inner Hebrides in 2001’.
Atlantic Seabirds 5: 89–100.
Newton, S., Thompson, K. and Mitchell, P. (2004). ‘Manx Shearwater
Puffinus puffinus’. In: P. Mitchell, S. Newton, N. Ratcliffe and T. Dunn,
(eds.) Seabird Populations of Britain and Ireland, pp. 63–80. London:
T & AD Poyser.
Perkins, A., Douse, A., Morgan, G., Cooper, A. and Bolton, M. (2017).
‘Using dual-sex call improves the playback census method for a
nocturnal burrow-nesting seabird, the Manx shearwater Puffinus
puffinus’. Bird Study 64: 146–158.
Perrins, C., Wood, M., Garroway, C., Boyle, D., Oakes, N., Revera, R.,
Collins, P. and Taylor, C. (2012). ‘A whole-island census of the Manx
shearwater Puffinus puffinus breeding on Skomer Island in 2011’.
Seabird 25: 1–13.
Ratcliffe, N., Vaughan, D., Whyte, C. and Shepherd, M. (1998).
‘Development of playback census methods for storm petrels Hydrobates
pelagicus’. Bird Study 45: 302–312.
Ratcliffe, N., Mitchell, I., Varnham, K., Verboven, N. and Higson, P.
(2009). ‘How to prioritize rat management for the benefit of petrels:
a case study of the UK, Channel Islands and Isle of Man’. Ibis 151:
699–708
Smith, S., Thompson, G. and Perrins, C. (2001). ‘A census of the Manx
shearwater (Puffinus puffinus) on Skomer, Skokholm and Middleholm,
west Wales’. Bird Study 48: 330–340.
Thomas, B. and Taylor, R. (2002). ‘A history of ground-based rodent
eradication techniques developed in New Zealand, 1959–1993’. In:
C.R. Veitch and M.N. Clout (eds.) Turning the tide: the eradication of
invasive species, pp. 301–310. Occasional Paper SSC no. 28. IUCN
SSC Invasive Species Specialist Group. IUCN, Gland, Switzerland and
Cambridge, UK.
Thomas, S., Brown, A., Bullock, D., Lock, L., Luxmoore, R., Roy, S.,
Stanbury, A. and Varnham, K. (2017). ‘Island restoration in the UK –
past present and future’. British Wildlife 28: 231–242.
Towns, D. and Broome, K. (2003). ‘From small Maria to massive
Campbell: Forty years of rat eradications from New Zealand islands’.
New Zealand Journal of Zoology 30: 377–398.
Towns, D., Atkinson, I. and Daugherty, C. (2006). ‘Have the harmful
effects of introduced rats on islands been exaggerated?’. Biological
Invasions 8: 863–891.
Towns, D., Byrd, G., Jones, H., Rauzon, M., Russell, J. and Wilcox, C.
(2011) ‘Impacts of Introduced Predators on Seabirds’. In: C. Mulder, W.
Anderson, D. Towns and P. Bellingham (eds.) Seabird Islands: Ecology,
Invasions and Restoration, pp. 56–90. New York: Oxford University
Press.
Walsh, P., Halley, D., Harris, M., del Nevo, A., Sim, I., and Tasker,
M. (1995). Seabird Monitoring Handbook for Britain and Ireland.
Peterborough: Joint Nature Conservation Committee, Royal Society
for the Protection of Birds, Institute of Terrestrial Ecology and Seabird
Group.
J.P. Bird, K. Varnham, J.D. Shaw and N.D. Holmes
Bird, J.P.; K. Varnham, J.D. Shaw and N.D. Holmes. Practical considerations for monitoring invasive mammal eradication outcomes
Practical considerations for monitoring invasive mammal
eradication outcomes
J.P. Bird1, K. Varnham2, J.D. Shaw1 and N.D. Holmes3
1
Centre for Biodiversity and Conservation Science, Level 5, Goddard Building, The University of Queensland, St Lucia,
QLD 4072 Australia. <jez.bird@uq.edu.au>. 2RSPB, The Lodge, Potton Road, Sandy, Bedfordshire SG19 2DL UK.
3
Island Conservation, 2100 Delaware Ave., Suite 1 Santa Cruz, CA 95060 USA.
Abstract Monitoring provides important evidence to evaluate if invasive mammal eradications on islands achieved
conservation goals and to inform future allocation of conservation funds. However, monitoring can be costly, and compete
with management actions for resources, so it is important that efforts are targeted. It can be difficult to obtain results of
monitoring from previous projects, which puts future projects at a disadvantage when projecting possible conservation
outcomes. In this short paper we discuss practical considerations for designing monitoring associated with an eradication
programme. We focus on the ecological outcomes of invasive species eradication, as opposed to specifically monitoring
if an eradication was successful or not. We identify major motivations for undertaking monitoring and present a decision
tree intended to improve the efficiency of monitoring by supporting project managers determining when and what to
monitor, and how to incorporate monitoring into project planning.
Keywords: decision tree, monitoring, post-eradication, project planning
INTRODUCTION
There is a substantial shortfall between the investment in
conservation worldwide and the amount required to tackle
the current biodiversity crisis (McCarthy, et al., 2012). It is
well established that this necessitates prioritised, efficient
allocation of resources, with evidence-based management
(Sutherland, et al., 2004; Wilson, et al., 2006; Kapos, et al.,
2008; Underwood, et al., 2008). Monitoring, defined as the
collection and analysis of repeated field-based empirical
measurements (Lindenmayer & Likens, 2010), provides
this evidence. Over the past fifty years the eradication of
invasive mammals from islands has developed into a reliable
and effective conservation tool, resulting in substantial
conservation gains (Veitch, et al., 2011; Jones, et al., 2016).
There has been noteworthy progress in determining why,
what, when and where to eradicate mammals from islands.
Prioritisation of management is more robust and evidencebased than ever. When and how much to invest, how to
balance different outcomes, and dealing with uncertainty
of outcomes have all been addressed in recent literature
(Donlan & Wilcox, 2007; Dawson, et al., 2015; Donlan,
et al., 2015; Helmstedt, et al., 2016). Yet why, what, when
and where to monitor following eradication is not always
apparent. Recent assessments have highlighted that data on
native species’ responses to eradication are rare, often not
quantitative, and not readily available through published
sources, suggesting that monitoring, or reporting on
monitoring, following eradication is uncommon (Jones,
et al., 2016; Brooke, et al., 2017; Towns, 2018). Both
eradication, and monitoring the outcomes that result, can
be costly (e.g. Helmstedt, et al., 2016; Springer, 2016).
Assuming both activities are being funded from the same
combined budget, there is a potential trade-off between
spending on eradication versus spending on monitoring
(Possingham, et al., 2012). This paper discusses what
to consider when designing monitoring of eradication
projects. We focus on monitoring the wider ecological
impacts of invasive species eradication, rather than shortterm post-eradication monitoring for signs of invasive
species that determines whether an eradication project has
succeeded or failed. The paper incorporates inputs from
the Island Invasives 2017 workshop: “Effective monitoring
of response to eradications” attended by 60 conference
participants. We aim to outline the main considerations for
practitioners assessing the monitoring needs for projects
they are involved in.
WHY, WHEN AND WHAT SHOULD WE
MONITOR?
Possingham, et al. (2012) identified five separate
benefits of long-term monitoring. Three of them—auditing
the outcomes of a project (Case study 1), detecting
unanticipated outcomes and researching mechanisms for
those outcomes—have ecological benefits. The other two
CASE STUDY 1: AUDITING THE OUTCOMES OF RAT ERADICATION AT ANACAPA, CALIFORNIA,
USA
Black rats (Rattus rattus) were successfully eradicated from the three islands of Anacapa in the Channel Islands,
California USA, in 2001–2002. The goal of the eradication project was to improve seabird nesting habitat, and aid
recovery of Scripps’s murrelet (Synthliboramphus scrippsi, formerly Xantus’s murrelet) and ashy storm-petrel
(Oceanodroma homochroa) (NPS, 2000). The project was funded via oil-spill restoration resources, and an additional
goal was to offset impacts that had occurred to these two species during the 1990 American Trader spill (ATTC, 2001).
Monitoring included tracking artificial eggs (mimicking Scripps’s murrelet eggs) before the eradication to quantify rat
predation on this life history stage, and after the eradication to confirm the expected outcome of removing that impact
(Jones, et al., 2005). Long-term monitoring of focal seabird species ensued for a decade including the hatching success,
distribution and abundance of Scripps’s murrelet on the island, which saw a three-fold increase in hatching success and
expansion of nesting (Whitworth, et al., 2013). The ashy storm-petrel was discovered breeding on the island 10 years
post-eradication, highlighting the contribution of the project towards stated goals (Whitworth, et al., 2013; Newton,
et al., 2016). The operation was also the first aerial broadcast of rodenticide in the USA, and short-term non-target
monitoring was undertaken to follow expected impacts (Howald, et al., 2010), and improve knowledge for further
planning of this activity in the USA. Surveillance monitoring of other taxa also occurred, including endemic deer mice,
herpetofauna and inter-tidal communities, to understand the broader impacts that occurred as a consequence of the
eradication.
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 545–551. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
545
Island invasives: scaling up to meet the challenge. Ch 3C Strategy: Outcomes
– informing stakeholders of outcomes and engaging the
public – have social benefits. Whether ecological or social,
several of these benefits involve measuring or reporting
against targets, so clearly defining the target outcomes
of eradications is often a prerequisite for designing
monitoring.
Monitoring may yield diminishing returns in terms
of advancing our ecological knowledge, when the same
outcome is monitored repeatedly. It can, therefore, detract
from investment in future management action. This is
an important consideration for repeated monitoring of
the same island or site (Possingham, et al., 2012), but
also for monitoring across projects where islands share
similar habitat types, and invasive mammal-native species
interactions. The target outcome of an eradication of a
particular invasive species e.g. population recovery of
a threatened species, may be confidently predicted if it
is driven by a simple mechanistic relationship or there
is sufficient evidence from previous eradications that
benefited the same or ecologically similar species. The
decision whether to monitor should, therefore, be informed
by the current state of evidence: what prior knowledge
exists and is it sufficient to confidently predict outcomes?
For invasive mammal eradications, the evidence-base
for predicting different outcomes is mixed. Individual
outcomes have been reported for several projects but not
consistently or comprehensively.
A key recommendation made during the Island
Invasives 2017 monitoring workshop was to compile
a synthesis of monitoring efforts to date, to identify
taxonomic or geographic gaps in coverage that will
help target future monitoring efforts. Although no
comprehensive synthesis exists currently, some studies
have collated and synthesised monitoring, either at a
regional level (e.g. Russell, et al., 2016; Towns, et al.,
2016), or globally for a taxonomic group. Schweizer,
et al., (2016) reviewed available evidence of vegetation
responses to goat (Capra hircus) and European rabbit
(Oryctolagus cuniculus) eradications. Although there was
evidence that vegetation responded following herbivore
removal, variation in monitoring methods, timeframe
and accounting for native versus non-native vegetation
response hindered the drawing of conclusions. Thus, the
authors recommended further monitoring to develop a
general model of expected vegetation responses. Brooke,
et al., (2017) collated seabird demographic responses
following invasive mammalian predator eradication,
highlighting that, in general, seabird populations increase
following invasive mammal eradications. However, not
all populations grew, insufficient data were available to
distinguish between threatened and non-threatened species,
and variation in response among major seabird taxa was
evident. Thus, while generally seabirds can be predicted to
respond positively following invasive mammal eradication,
we lack sufficient knowledge to predict how and why
this circumstance occurs, hence the recommendation for
systematic long-term monitoring to improve understanding
of the mechanisms of seabird population recovery (Brooke,
et al., 2018).
The social benefits that accrue from monitoring–
stakeholder feedback and public engagement–are more
linear because, while ecological knowledge grows
cumulatively from all projects that monitor, the social
returns are primarily project specific. Foreseeably,
the ecological need for monitoring may be low but, if
the operation had high public or stakeholder interest,
monitoring will be necessary.
Beyond a theoretical framework for monitoring,
The Nature Conservancy is one organisation looking at
their motivations for monitoring at an institutional level
(Montambault & Groves, 2009). They found monitoring
was a tool for managing risk and securing future investment–
the greater a project’s risk or higher the likelihood it could
lead to follow-on funding, the higher the investment that
should be made in monitoring (Case study 2). Eradication
operations with considerable ecological uncertainty, or
reputational risk, and those whose success could leverage
additional public, political or financial support for future
operations therefore all warrant a significant investment in
monitoring (Table 1).
Having identified the motivations for monitoring,
and decided on that basis whether monitoring is needed,
it becomes easier for a project team to decide what to
monitor and how. When the aim is to confirm that an
eradication achieved target outcomes, monitoring focusses
on those target beneficiaries. When the risk of unexpected
outcomes is high, broader surveillance monitoring is
appropriate. Both rely on assessing the state of target or
non-target species or habitats, whereas understanding
broader ecosystem responses is likely to require more
detailed research into ecological mechanisms.
The goal and audience for reporting ecological
outcomes of an eradication can influence the type of
monitoring undertaken. When there is a need to report
outcomes in a peer-reviewed publication to a technical or
scientific audience, a different approach such as a quantified
before-after comparison, may be required than for projects
reporting to non-technical audiences such as donors or
local communities (Case study 3), for which qualitative
approaches like photo-monitoring vegetation changes may
be sufficient. Further, the stakeholders using Traditional
Ecological Knowledge will require a different approach
CASE STUDY 2: LEVERAGING CONSERVATION GAINS THROUGH GOAT ERADICATIONS IN THE
GALÁPAGOS, ECUADOR
The ultimate goal of Project Isabela, initiated in 1997 and completed in 2006, was to facilitate the restoration of Pinta
(5,940 ha) and Santiago (58,465 ha) Islands and the larger, northern portion of Isabela Island (approximately 250,000
ha; the whole island encompasses 458,812 ha). The project began in response to the massive destruction by introduced
goats of both native vegetation and terrain (Galápagos Conservancy, 2017). Long-term vegetation monitoring was
established on six of the 12 islands in the Galápagos where goats had been introduced (Tye, 2006). Permanent plots and
transects showed that eradication or reduction of goat populations led to regeneration of native vegetation (Hamann,
1993; 1979), with a return to a near natural state in most cases after 20 years (Tye, 2006). The monitoring programme
successfully fulfilled a number of roles. It confirmed, overall, the success of goat eradication in facilitating recovery
of native vegetation and it provided lessons for subsequent eradication operations. In doing this, monitoring helped
to manage the risk associated with the operation. The programme highlighted cases where individual species did not
recover following goat eradication or exclusion so additional conservation management was required, including the tree
fern (Cyathea weatherbyana) on Alcedo, whose last two remnant populations were protected by fences in 1997 (Tye,
2006). Perhaps most importantly monitoring demonstrated to public, state and donor audiences the benefits of invasive
species management helping to leverage future investment. This led to the Charles Darwin Foundation (CDF) and the
Galápagos National Park Service (GNPS) convening a workshop in 2007, on the completion of Project Isabela, to
develop an action plan for managing rodents within the Galápagos (Galápagos Conservancy, 2017).
546
Bird, et al.: Monitoring invasive mammal eradication outcomes
Table 1 Motivations and conditions for monitoring biodiversity outcomes of invasive species eradications.
Why?
Confirm target
outcomes1
Detect non-target
outcomes1
Learn about whole
ecosystem responses1
When?
Outcomes are complex and difficult to predict, or
poorly studied
Large or complex systems where outcomes are
unpredictable
Ecosystem responses remain poorly studied.
Inform stakeholders1
If required, especially for larger operations
Engage the public1
Inhabited islands, regularly visited islands, large
operations, publicly funded operations. Projects
involving beneficiary species with a high public profile
Threatened species involved and outcomes uncertain
e.g. complex systems
Large operations funded by key donors, or receiving
political and public backing
Exemplars and trial operations in new geographies
paving the way for subsequent repetition/scaling-up
Ecological risk and
uncertainty2
Reputational risk2
Leverage2
What?
Target beneficiaries – quantitative
studies
Non-target surveillance –
quantitative studies
Ecological mechanisms of change
– quantitative question-driven
research. Community ecology
Target beneficiaries – qualitative
studies (see Case Study 1)
Target beneficiaries – qualitative
participatory monitoring
Target beneficiaries – empirical
studies
Target beneficiaries – qualitative
studies?
Target beneficiaries – empirical
studies
Sources: 1Possingham, et al., 2012 and 2Montambault & Groves, 2009.
than non-Western science frameworks. Thus, identifying
early the key audiences and their needs is recommended
as this will influence the cost and approach of monitoring.
Finally, there are a number of practical considerations
which may predispose projects to monitor, namely: when
existing baseline data are particularly good; when there
are existing established monitoring programmes e.g.
run by rangers, universities or participatory groups; and/
or when funding for monitoring does not compete with
management.
for eradication monitoring, they are not achievable for
all projects, nor may they be necessary to achieve project
goals. Here, we aim to provide general guidelines for
deciding what level of monitoring is required.
Fig. 1 presents a decision tree outlining the key
considerations which determine whether monitoring is
necessary, what needs to be monitored and the type of
monitoring needed. Although it is presented as a workflow,
several steps are inter-related and feed into one another.
Integrating monitoring into project planning
The most common motivation for monitoring is to
confirm the expected outcomes for native taxa after
removing a pest species from an island. It is therefore
essential that projects clearly define their objectives
(Prior, et al., 2018): why is eradication proposed?; what
is it expected to achieve? Outcomes should be explicitly
split into proximal outcomes, which will typically include
the removal of an invasive species and the undesirable
interactions with native species (e.g. predation), and
ultimate outcomes such as the recovery of a native species.
These ultimate outcomes are sometimes referred to as
impacts (Nam, et al., 2013). Conceptually, post-eradication
outcomes like improved survival and recruitment can
lead to impacts like population growth. Where possible,
outcomes should be specific, measurable, agreed-upon by
those involved in the project, realistic (i.e. ecologically
viable), and time-bound (Doran, 1981).
Fundamentally, monitoring should be considered in
the earliest stages of project planning. This allows for
additional baseline data to be collected if existing data are
insufficient for a robust before-after comparison, and for
monitoring to be costed and potentially included in the
project budget.
There is a wide spectrum of possible monitoring
investment for invasive mammal eradication projects,
ranging from not monitoring at all, through to comprehensive
whole ecosystem monitoring. The few whole ecosystem
studies that exist (e.g. Towns, et al., 2016; Griffiths, et al.,
2019) provide detailed learning into how systems respond
to the eradication of particular species and provide a model
for planning equivalent exercises elsewhere. Although
these excellent studies represent the optimum approach
1. Defining the desired outcomes of eradication
CASE STUDY 3: MONITORING ON ST AGNES AND GUGH, ISLES OF SCILLY, UK
Brown rats (Rattus norvegicus) were successfully eradicated from the islands of St Agnes and Gugh in 2013
(Thomas, et al., 2017). The islands have a combined area of 142 ha and a population of 82 people, making it the
largest community-led rat eradication project in the world to date. Engaging the community in all aspects of the project
including monitoring – and keeping them engaged throughout the life span of the project – was key to the project’s
success. Community members, especially schoolchildren, were involved in the work, with many people volunteering to
take part in monitoring of native shrews, invertebrates, plants and birdlife. The islands’ seabirds are of particular value
to the community, and islanders are involved in ongoing ‘chick check’ walks which monitor the breeding success of
Manx shearwaters (Puffinus puffinus) and European storm-petrels (Hydrobates pelagicus) two species which have bred
on the islands for the first time in living memory following the eradication of rats. The monitoring activities associated
with the eradication project have therefore fulfilled several roles – they have provided ongoing scientific data on the
wider ecological impacts of rat eradication and have provided powerful publicity and advocacy information regarding
the immediate benefits of eradication on species preyed upon by rodents, such as shearwaters and storm-petrels. The
monitoring has also galvanised and helped maintain ongoing community support for the project and ownership of its
long-term outcomes.
547
Island invasives: scaling up to meet the challenge. Ch 3C Strategy: Outcomes
Fig. 1 A decision tree to assist with planning biodiversity monitoring in relation to eradication programmes. After the target
outcomes of the eradication and the key audiences are defined, gathering existing evidence and answering a number
of questions will inform the scope that monitoring needs to encompass, as well as guide selection of monitoring targets
and the required approach to monitoring. Numbered points are discussed in the text.
2. Defining the key audiences
Just as what is monitored is driven by project goals
(e.g. seabird protection), to whom monitoring results
should be communicated should also be defined as a part
of project planning. For each project, target audiences
should be identified: a relevant group for whom the results
of ecological monitoring will be of interest, some of whom
may actively request the information, while others may
be informed more for advocacy and education purposes.
By defining these audiences, managers can prioritise and
determine what messages and products (e.g. peer-reviewed
publications, images, webpages, reports or public lectures)
monitoring needs to inform and, in a feedback loop, can
identify what to monitor. Key audiences may include the
local community, especially island residents or communities
close to the island; permit providers such as statutory
bodies and island managers wanting to understand the
wider ramifications of eradication; conservation scientists
and technical communities wishing to use monitoring data
to highlight ecological benefits of projects to advocate for
similar work; donors vetting project outcomes and return
on investment; and decision makers at local and regional
548
levels. At a higher level, the data may also be used to lobby
policy makers to enact or amend legislation relating to
invasive species and their management. Finally, project
managers may wish to engage the wider public with the
results of their monitoring work, seeking to develop more
broadly society’s understanding of the issues posed by
invasive species on islands.
3. Identifying existing resources
Determining the presence and suitability of existing
baseline data for the target island is an important activity.
Existing baseline information may satisfy pre-eradication
information needs and can inform future monitoring to
replicate the baseline methodology. This exercise may
identify stakeholders already engaged in monitoring on
the target island, or nearby control islands, whose ongoing
work may be tailored to inform eradication outcomes.
It is also valuable to assess the outcomes of other
eradication projects that benefited similar species or
ecosystems, for example ground-nesting seabirds like
terns (Sterna spp.) perform well after the removal of all
invasive mammalian predators (Brooke, et al., 2017). The
Bird, et al.: Monitoring invasive mammal eradication outcomes
consistency with which previous eradications delivered
particular outcomes will establish the level of confidence
in achieving desired management goals. Syntheses have
been undertaken for some taxa that outline broad-scale
responses (e.g. Jones, et al., 2016; Schweizer, et al., 2016;
Brooke, et al., 2017). Williams, et al., (2017) synthesised
information from 16 before-and-after studies documenting
seabird responses to predator removal and provide
practitioners with effectiveness and certainty ratings for
invasive mammal control as a conservation intervention.
If the results among projects vary considerably, or there
is a specific requirement for reporting to an audience on
localised information, then monitoring is warranted.
people involved in the project with particular expertise;
there are taxa present for which monitoring is likely to be
particularly cost-effective.
Sampling design should ideally occur before
eradication. In some instances a Before-After-ControlImpact (BACI) approach may be possible (Quinn &
Keough, 2002), whereby control islands (either those
with invasive species but where no eradication is carried
out, or those with no comparable invasive species at all)
can be compared to experimental islands (those with the
eradication e.g. Samaniego-Herrera, et al., 2017).
4. Adopting a whole-ecosystem approach
A major determinant of monitoring design is the
economic cost, relative to available budget. For monitoring
planned shortly after an eradication, an opportunity for
cost saving is to combine efforts with activities to confirm
the success or failure of the operation itself.
The amount invested should increase relative to risk
and leverage potential (Montambault & Groves, 2009), but
there are no clear guidelines on what proportion of a budget
to allocate for monitoring and evaluation. There has been
no review of proportional expenditures by conservation
projects on monitoring and evaluation, but within the
development sector and across major foundations typical
expenditure is 3–5% of programme costs (Austrian
Development Agency, 2009; Twersky & Arbreton,
2014), rising to an upper ceiling of 10% (Zondag, 2009).
Establishing a fixed limit for monitoring budgets helps to
guide monitoring design, and may result in iterative design
to keep monitoring within budget. Including monitoring
costs in the overall eradication budget is perhaps the most
straightforward way of funding monitoring, when it is a
relatively small component of the overall fund-raising
target. However, funds secured in this way are often
time-bound and not goal dependent—they often expire
before monitoring has been conducted for enough years
to demonstrate that a target outcome has been reached.
Addressing this issue by exploring financial mechanisms
such as endowment funds to separate and safeguard
monitoring budgets and ongoing biosecurity, or integrate
ecosystem monitoring with biosecurity monitoring, could
help future projects and improve upon the current approach
that relies on post-eradication fund-raising specifically for
monitoring.
When planning and designing monitoring, ecosystem
processes and community structure should be considered
(Zavaleta, et al., 2001; Prior, et al., 2018). By modelling
the trophic interactions in a system, flow-on effects can
be anticipated, reducing the likelihood of unexpected
outcomes (e.g. Baker, et al., 2017). There is a gradient
of approaches available to achieve this process, ranging
from simple food web diagrams through to models with
input from community/ecosystem ecologists. Generating
sophisticated models is challenging for many sites owing
to a lack of baseline information. However, even simple
exercises capturing current and projected interactions
within an ecosystem could aid planning. By considering
the trophic interactions on an island, those component taxa
of interest which are most likely to be impacted can be
identified and elevated to monitoring targets. This will also
clarify the complexity of the system which highlights the
potential need for wider surveillance monitoring beyond
anticipated outcomes.
5. Designing monitoring
There is a whole suite of taxa- and site-specific
monitoring methods that projects can utilise—it is not
our aim to discuss them here. Rather, we focus on three
key elements of monitoring design: i) choosing between
quantitative and qualitative monitoring methods; ii)
determining what to monitor; and iii) allowing for pre- and
post-eradication comparison.
The need for quantitative or qualitative monitoring is
influenced by the audience to whom monitoring results will
be communicated. As described above, a spectrum exists.
At one end, are projects for which quantitative monitoring is
required: for example, those with quantitative targets such
as percentage population changes or reductions in negative
trends; or those aiming to quantify outcomes to inform
other eradication operations (e.g. by providing evidence
for syntheses like Williams, et al., 2017). Further along
the spectrum are projects that may need only to provide
qualitative evidence of outcomes to laypersons’ audiences:
perhaps photo-plots illustrating the growth in vegetation
following an eradication; or “traffic-light” assessments of
ecological integrity (e.g. Tierney, et al., 2009) of an island
system following eradication.
To serve most purposes, monitoring can likely focus on
taxa or habitats identified when the target outcomes of the
eradication were defined. But, when potential secondary
outcomes have been identified, such as increases in invasive
invertebrates or prey-switching by meso-predators, taxa
or habitats predicted to be affected can also be selected
as monitoring targets. When outcomes are highly
uncertain, we recommend wider surveillance monitoring is
undertaken to detect hard-to-predict secondary outcomes.
In that case, taxa can be selected for monitoring based
first upon need (they are predicted to be affected, but
with unknown consequences), and then opportunity, e.g.
continued monitoring is worthwhile because baseline data
exist and monitoring can be continued easily; there are
6. Monitoring cost
Making the most of monitoring results
With so many eradication projects now being carried
out worldwide and many of them generating data through
associated monitoring, it is increasingly difficult for
scientists, managers and field officers to keep up to date
with new findings, and they can be hampered by language
barriers. Furthermore, the data generated are not always
disseminated widely. Understandably, positive changes to
target beneficiaries and to flagship species, are the most
widely reported. Changes in the abundance of other taxa,
especially plants and invertebrates, are less often reported,
or likely monitored (Jones, et al., 2016; Towns, 2018).
Understanding the outcome of previous eradication
projects’ pre- and post-eradication monitoring may
help new projects gain support for their work, may help
to identify and thus allow minimisation of negative
secondary impacts, and may help to optimise the allocation
of resources to conservation actions where monitoring
can be reduced. It is very important, therefore, that first
the results from any monitoring that has occurred are
disseminated, and second that the information is curated
in a readily accessible and searchable manner accessible
to technicians, land managers, scientists, conservation
bodies, educators and other interested parties. Ideally, they
would all be available via a single repository but nothing
549
Island invasives: scaling up to meet the challenge. Ch 3C Strategy: Outcomes
exists currently (although there is a searchable database
of island eradications; Holmes, et al., 2019). Although the
outcomes of individual eradications may not be considered
sufficiently novel for higher profile journals (Brooke, et
al., 2018), a number of journals specifically promote the
dissemination of evidence by promoting publication of the
outcomes of conservation interventions (Sutherland, et al.,
2017). There are opportunities for open access publishing
with no limit on the number of papers publishable or the
geographies covered, and a streamlined submission and
review process. There is a range of ways in which results
can then be disseminated more widely (Table 2).
Monitoring informs future conservation practice;
it enables us to increase our likelihood of success and
reduces uncertainty. We believe that there is a need to
broaden information availability and shared resources
through diverse platforms, in order to facilitate knowledge
exchange. To date, the findings of post-eradication
monitoring have not been consistently disseminated, so
a behavioural change must be supported and requires
incentivising. Including these costs in eradication budgets
and encouraging donors to support the collection of
evidence that confirms return on investment are first steps
in tackling the problem.
ACKNOWLEDGEMENTS
We thank all those involved in organising and
participating in the Island Invasives 2017 workshop:
“Effective monitoring of response to eradications”. We also
thank the Editorial team of the Conference Proceedings,
and two anonymous referees who provided valuable
feedback during the development of this paper.
Table 2 Summary of dissemination routes for pre- and post-eradication monitoring data, particularly to the invasive
species-specific technical community.
Media
Raw data
Analysed
results
Technical
reports
Layperson
reports
Dissemination route
Inter-agency scientific collaborations
Small organisations collaborating with
larger ones who can support with data
analysis and interpretation
Journals reporting outcomes of conservation
intervention in a searchable database
Aliens-l listserver
Held in central database (e.g. Database of
Island Invasive Species Eradications—
subject to copyright issues)
Eradication project/ organisation websites
Briefing documents, e.g. POST
(Parliamentary Office of Science and
Technology) notes
Aliens-l listserver
Regional websites, e.g. Pacific Invasives
Learning Network, PestSmart Connect
Annual compendium
Island Invasives Conference proceedings
Community forums (newsletters,
magazines, websites)
Through Aliens-l listserver
Held in central database?
Schools
Educational Universities (use examples in lectures on
materials
island restoration and species recovery)
Talks/presentations
Social
media
550
Web sites
Projects own Facebook pages, and links to
reports via twitter and instagram
Target audience
Invasive species practitioners and scientists
Any
Invasive species-specific scientific and technical
community
Invasive species-specific scientific and technical
community
Invasive species-specific scientific and technical
community
Invasive species-specific technical community
Local, regional and national government
Invasive species-specific technical community
Invasive species-specific technical community
Scientific and academic community, invasive speciesspecific scientific and technical community
Community in which eradication project was conducted,
communities in which similar projects are planned
Invasive species-specific technical community
Invasive species-specific technical community
Primary and secondary school children, and teachers
Students
Community in which eradication project was conducted,
communities in which similar projects are planned,
special interest groups (e.g. local bird or mammal
groups)
Scientific and academic community, invasive speciesspecific scientific and technical community. Community
in which eradication project was conducted, communities
in which similar projects are planned, special interest
groups (e.g. local bird or mammal groups)
Bird, et al.: Monitoring invasive mammal eradication outcomes
REFERENCES
ATTC (2001). Restoration Plan & Environmental Assessment for
Seabirds Injured by the American Trader Oil Spill. [Online.] Report of
the American Trader Natural Resource Trustee Council, U.S. Fish and
Wildlife Service, California Department of Fish and Game, and National
Oceanic and Atmospheric Administration. <https://casedocuments.
darrp.noaa.gov/southwest/amtrader/pdf/amtradefinal_restoreplanon.
pdf>. Accessed 21 October 2017.
Austrian Development Agency (2009). Guidelines for Project and
Programme Evaluations. Vienna, Austria: Austrian Development
Agency.
Baker, C.M., Gordon, A. and Bode, M. (2017). ‘Ensemble ecosystem
modeling for predicting ecosystem response to predator reintroduction’.
Conservation Biology 31: 376–384.
Brooke, M. de L., Bonnaud, E. Dilley, B.J., Flint, E.N., Holmes, N.D.,
Jones, H.P., Provost, P., Rocamora, G., Ryan, P.G. Surman, C. and others
(2017). ‘Seabird population changes following mammal eradications on
islands’. Animal Conservation 21: 3–12.
Brooke, M. de L., Bonnaud, E. Dilley, B.J., Flint, E.N., Holmes, N.D.,
Jones, H.P., Provost, P., Rocamora, G., Ryan, P.G. and Surman, C.
(2018). ‘Enhancing the value of future island eradications needs
improved understanding of past outcomes’. Animal Conservation 21:
19–20.
Dawson, J., Oppel, S. Cuthbert, R.J., Holmes, N., Bird, J.P., Butchart,
S.H., Spatz, D.R. and Tershy, B. (2015). ‘Prioritizing islands for the
eradication of invasive vertebrates in the United Kingdom overseas
territories’. Conservation Biology 29: 143–153.
Donlan, C.J. and Wilcox, C. (2007). ‘Complexities of costing
eradications’. Animal Conservation 10: 154–156.
Donlan, C.J., Luque, G.M. and Wilcox, C. (2015). ‘Maximizing return
on investment for island restoration and species conservation’.
Conservation Letters 8: 171–179.
Doran, G.T. (1981). ‘There’s a SMART way to write management’s goals
and objectives’. Management Review 70: 35–36.
Galapagos Conservancy (2017). ‘Project Isabela’. <https://www.
galapagos.org/conservation/conservation/project-areas/ecosystemrestoration/project-isabela/>. Accessed 21 October 2017.
Griffiths, R., Bell, E., Campbell, J., Cassey, P., Ewen, J.G., Green, C.,
Joyce, L., Rayner, M. Toy, R. Towns, D. Wade, L., Walle, R. and Veitch,
C.R. (2019). ‘Costs and benefits for biodiversity following rat and cat
eradication on Te Hauturu-o-Toi/Little Barrier Island’. In: C.R. Veitch,
M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) Island
invasives: scaling up to meet the challenge, pp. 558–567. Occasional
Paper SSC no. 62. Gland, Switzerland: IUCN.
Hamann, O. (1979). ‘Regeneration of vegetation on Santa Fé and
Pinta Islands, Galápagos, after the eradication of goats’. Biological
Conservation 15: 215–235.
Hamann, O. (1993). ‘On vegetation recovery, goats and giant tortoises
on Pinta Island, Galápagos, Ecuador’. Biodiversity and Conservation
2: 138–151.
Helmstedt, K.J., Shaw, J.D., Bode, M., Terauds, A., Springer, K.,
Robinson, S.A. and Possingham, H.P. (2016). ‘Prioritizing eradication
actions on islands: It’s not all or nothing’. Journal of Applied Ecology
53: 733–741.
Holmes, N.D., Keitt, B.S., Spatz, D.R., Will, D.J., Hein, S., Russell, J.C.,
Genovesi, P., Cowan, P.E. and Tershy, B.R. (2019). ‘Tracking invasive
species eradications on islands at a global scale’. In: C.R. Veitch, M.N.
Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) Island invasives:
scaling up to meet the challenge, pp. 628–632. Occasional Paper SSC
no. 62. Gland, Switzerland: IUCN.
Howald, G., Donlan, C.J., Faulkner, K.R., Ortega, S., Gellerman, H.,
Croll, D.A. and Tershy, B.R. (2010). ‘Eradication of black rats Rattus
rattus from Anacapa Island’. Oryx 44: 30–40.
Jones, H.P., Williamhenry, R., Howald, G.R., Tershy, B.R., and Croll, D.A.
(2005). ‘Predation of artificial Xantus’s murrelet Synthliboramphus
hypoleucus scrippsi nests before and after black rat Rattus rattus
eradication’. Environmental Conservation 32: 320–325.
Jones, H.P., Holmes, N.D., Butchart, S.H.M., Tershy, B.R., Kappes,
P.J., Corkery, I., Aguirre-Muñoz, A., Armstrong, D.P., Bonnaud, E.,
Burbidge, A.A., Campbell, K., Courchamp, F., Cowan, P.E., Cuthbert,
R.J., Ebbert, S., Genovesi, P., Howald, G.R., Keitt, B.S., Kress, S.W.,
Miskelly, C.M., Oppel, S., Poncet, S., Rauzon, M.J., Rocamora, G.,
Russell, J.C., Samaniego-Herrera, A., Seddon, P.J., Spatz, D.R., Towns,
D.R., and Croll, D.A. (2016). ‘Invasive mammal eradication on islands
results in substantial conservation gains’. Proceedings of the National
Academy of Sciences 113 :4033–4038.
Kapos, V., Balmford, A., Aveling, R., Bubb, P., Carey, P., Entwistle, A.,
Hopkins, J., Mulliken, T., Safford, R., Stattersfield, A., Walpole, M. and
Manica, A. (2008). ‘Calibrating conservation: New tools for measuring
success’. Conservation Letters 1: 155–164.
Lindenmayer, D.B. and Likens, G.E. (2010). ‘The science and application
of ecological monitoring’. Biological conservation 143: 1317–1328.
McCarthy, D.P., Donald, P.F., Scharlemann, J.P., Buchanan, G.M.,
Balmford, A., Green, J.M., Bennun, L.A., Burgess, N.D., Fishpool,
L.D., Garnett, S.T. and others (2012). ‘Financial costs of meeting global
biodiversity conservation targets: Current spending and unmet needs’.
Science 338: 946–949.
Montambault, J.R. and Groves, C.R. (2009). Improving Conservation
Practice by Investing in Monitoring Strategy Effectiveness.
Conservation Measures Working Paper 2. Arlington, Virginia: The
Nature Conservancy.
Nam, S., Ampofo, F.O., Shields, L., Alatamirano, E.S., and Toure, M.S.
(2013). Enhancing Learning in Monitoring and Evaluation: Six Cases
from Philanthropic Organizations Working in the Health Sector. Seattle,
Washington: Bill and Melinda Gates Foundation.
Newton, K.M., McKown, M., Wolf, C., Gellerman, H., Coonan, T.,
Richards, D., Harvey, A.L., Holmes, N., Howald, G., and Faulkner, K.
(2016). ‘Response of native species 10 years after rat eradication on
Anacapa Island, California’. Journal of Fish and Wildlife Management
7: 72–85.
NPS (2000). Anacapa Island Restoration Project: Final Environmental
Impact Statement. Ventura County, California: National Park Service.
<http://www.nps.gov/chis/learn/nature/upload/
AIRP%20Final%20
EIS%20Full.pdf>. Accessed 21 October 2017.
Possingham, H.P., Wintle, B.A., Fuller, R.A., and Joseph, L.N. (2012).
‘The Conservation Return on Investment from Ecological Monitoring’.
In: D. Lindenmayer and P. Gibbons (eds.) Biodiversity Monitoring in
Australia, pp. 49–58. Clayton, Australia: CSIRO Publishing.
Prior, K.M., Adams, D.C., Klepzig, K.D., and Hulcr, J. (2018). ‘When
does invasive species removal lead to ecological recovery? Implications
for management success’. Biological Invasions 20: 267–283.
Quinn, G.P. and Keough, M.J. (2002). Experimental Design and Data
Analysis for Biologists. Cambridge: Cambridge University Press.
Russell, J.C., Cole, N.C., Zuël, N., and Rocamora, G. (2016). ‘Introduced
mammals on Western Indian Ocean islands’. Global Ecology and
Conservation 6: 132–144
Samaniego-Herrera, A., Clout, M.N., Aguirre-Muñoz, A., and Russell,
J.C. (2017). ‘Rodent eradications as ecosystem experiments: A case
study from the Mexican tropics’. Biological Invasions 19: 1761–1779.
Schweizer, D., Jones, H.P., and Holmes, N.D. (2016). ‘Literature review
and meta-analysis of vegetation responses to goat and European rabbit
eradications on islands’. Pacific Science 70: 55–71.
Springer, K. (2016). ‘Methodology and challenges of a complex multispecies eradication in the sub-Antarctic and immediate effects of
invasive species removal’. New Zealand Journal of Ecology 40: 273–
278.
Sutherland, W.J., Pullin, A.S., Dolman, P.M., and Knight, T.M. (2004).
‘The need for evidence-based conservation’. Trends in Ecology and
Evolution 19: 305–308.
Sutherland, W.J., Dicks, L.V., Ockendon, N., and Smith, R.K. (2017).
What Works in Conservation. Cambridge, UK: Open Book Publishers.
Thomas, S., Brown, A., Bullock, D., Lock, L., Luxmoore, R., Roy, S.,
Stanbury, A. and Varnham, K. (2017) ‘Island Restoration in the UK –
past, present and future’. British Wildlife 28(4): 231–242.
Tierney, G.L., Faber-Langendoen, D., Mitchell, B.R., Shriver, W.G., and
Gibbs, J.P. (2009). ‘Monitoring and evaluating the ecological integrity
of forest ecosystems’. Frontiers in Ecology and Environment 7: 308–
316.
Towns, D., Borrelle, S., Thoresen, J., Buxton, R., and Evans, A. (2016).
‘Mercury Islands and their role in understanding seabird island
restoration’. New Zealand Journal of Ecology 40: 235–249.
Towns, D.R. (2018). ‘Understanding seabird responses to invasive
mammal eradications from islands needs systematic monitoring’.
Animal Conservation 21: 15–16.
Twersky, F. and Arbreton, A. (2014). Benchmarks for Spending on
Evaluation. Menlo Park, California: William and Flora Hewlett
Foundation.
Tye, A. (2006). ‘Restoration of the vegetation of the Dry Zone in
Galapagos’. Lyonia 9: 29–50.
Underwood, E.C., Shaw, M.R., Wilson, K.A., Kareiva, P., Klausmeyer,
K.R., McBride, M.F., Bode, M., Morrison, S.A., Hoekstra, J.M., and
Possingham, H.P. (2008). ‘Protecting biodiversity when money matters:
Maximizing return on investment. PLOS One 3: e1515.
Veitch, C.R., Clout, M.N., and Towns, D.R. (eds.) (2011). Island
invasives: eradication and management. Occassional Paper SSC no. 42.
Gland, Switzerland: IUCN and Auckland, New Zealand: CBB.
Whitworth, D.L., Carter, H.R., and Gress, F. (2013). ‘Recovery of a
threatened seabird after eradication of an introduced predator: Eight
years of progress for Scripps’s murrelet at Anacapa Island, California’.
Biological Conservation 162: 52–59.
Williams, D.R., Child, M.F., Dicks, L.V., Ockendon, N., Pople, R.G.,
Showler, D., Walsh, J.C., zu Ermgassen, E.K.H.J., and Sutherland,
W.J. (2017). ‘Bird Conservation’. In: W.J. Sutherland, L.V. Dicks, N.
Ockendon, and R.K. Smith (eds.) What Works in Conservation, pp.
95–244. Cambridge, UK: Open Book Publishers.
Wilson, K.A., McBride, M.F., Bode, M., and Possingham, H.P. (2006).
‘Prioritizing global conservation efforts’. Nature 440: 337–340.
Zavaleta, E.S., Hobbs, R.J., and Mooney, H.A. (2001). ‘Viewing invasive
species removal in a whole-ecosystem context’. Trends in Ecology and
Evolution 16: 454–459.
Zondag, M.-J. (2009). ‘Practical Handbook for Ongoing Evaluation’.
INTERACT Program Secretariat. <http://wiki.interact-eu.net/pages/
viewpage.action?pageId=23756932>. Accessed 21 October 2017.
551
R.N. Fisher, J. Niukula, P. Harlow, S. Rasalato, R. Chand, B. Thaman, E. Seniloli, J. Vadada, S. Cranwell, J. Brown, K. Lovich and N.
Thomas-Moko
Fisher, R.N.; J. Niukula, P. Harlow, S. Rasalato, R. Chand, B. Thaman, E. Seniloli, J. Vadada, S. Cranwell, J. Brown, K. Lovich and N.
Thomas-Moko. Community-based conservation and recovery of native species on Monuriki Island, Fiji
Community-based conservation and recovery of native species on
Monuriki Island, Fiji
R.N. Fisher1, J. Niukula2, P. Harlow3, S. Rasalato4, R. Chand5, B. Thaman2, E. Seniloli6, J. Vadada2, S. Cranwell7,
J. Brown8, K. Lovich8 and N. Thomas-Moko6
U.S. Geological Survey, Western Ecological Research Center, San Diego CA, 92101, USA. <rfisher@usgs.gov>.
National Trust of Fiji Islands, Suva, Fiji. 3Taronga Conservations Society Australia, Mosman, NSW, Australia. 4Ahura
Resorts, Fiji. 5Kula Wild Adventure Park, Sigatoka, Fiji. 6NatureFiji-MareqetiViti, Suva, Fiji. 7Birdlife Pacific, Suva,
Fiji. 8San Diego Zoo Global, San Diego, CA, USA.
1
2
Abstract The small uninhabited island of Monuriki (40.4 ha) in western Fiji is of national and international conservation
concern for its several protected species. Exotic invasive species and a Category 5 cyclone have exacerbated conservation
challenges. The cooperation of local, national, and international stakeholders continues to be crucial in restoration of the
island’s native flora and fauna. This summary presents a timeline of restoration efforts and current status of the recovery
programme for Monuriki. The critically endangered dry forest habitat of Fiji is only found in a few isolated patches on
disparate islands. The Fijian crested iguana (Brachylophus cf. vitiensis) is a critically endangered species restricted to a
few small islands in this dry forest zone of western Fiji. The population of crested iguanas on Monuriki Island is the third
largest remaining population. Even before iguanas were first documented on the island in the early 1980s, individuals had
been removed by local resorts for display purposes, a practice that was previously unregulated. In the late 1990s, the first
efforts to conserve and restore Monuriki Island were discussed, but conservation initiatives were not implemented until
the development of the Crested Iguana Recovery Plan in 2008. In 2011, domestic goats (Capra hircus) and non-native
rats (Rattus exulans) were removed from the island, and 10 pairs of iguanas were brought into captive breeding facilities
within Fiji. In 2015, the first 32 captive-bred crested iguanas were released back on Monuriki Island. More than half of
these iguanas (N=26) were radio-tracked for 56 days post-release in order to assess survivorship and help provide insights
into their short-term movement patterns. Of the 26 iguanas that were tracked, nearly 70% (N=18) were found after 56
days indicating excellent short-term survival. In February 2016, Tropical Cyclone Winston, a Category 5 storm passed
through Fiji and devastated some of the tropical dry forest habitat on Monuriki. With sustained winds of up to 230 km/
hr nearly all of the canopy leaves from trees on Monuriki Island were removed and large amounts of debris covered the
forest floor. Following the cyclone, a brief wildlife survey revealed Monuriki’s iguana and bird populations were still
present. In 2017, the crested iguana captive breeding programme was brought to an end when 16 of the original 20 iguana
founders, and an additional 32 captive bred offspring, were reintroduced onto Monuriki. This was accomplished, in part,
due to successful breeding and reestablishment of the remaining wild iguanas on the island. Despite a major storm event,
reestablishment likely resulted from reduced egg and hatchling predation by the rats, and excellent habitat recovery after
goat removal. Overall these invasive species eradications have proven highly successful for the recovery of the iguanas,
wedge-tailed shearwaters (Puffinus pacificus), and several other non-target species including the banded rail (Gallirallus
philippensis) and endangered Fijian peregrine falcon (Falco peregrinus). Furthermore, eradication of non-native species
has also helped the recovery of the highly threatened tropical dry forest ecosystem in which these species exist.
Keywords: Brachylophus cf. vitiensis, eradication, Fijian crested iguanas, goats, island restoration, Pacific rats, Puffinus
pacificus, wedge-tailed shearwater
INTRODUCTION
Tropical dry forest habitats are globally rare and often
contain highly endemic faunas. These forests are typically
impacted by anthropogenic fires to convert them into
lands for agriculture on mainland regions and degraded
by multiple invasive alien species such as grazing and
predatory mammals, and various invertebrates on islands.
In Fiji, most of the dry forest on the two large islands of Viti
Levu and Vanua Levu has been transitioned into sugar cane,
cattle grazing, or invasive grasslands (Olson, et al., 2010).
Dry forest persists only on some of the smaller islands, or
in very limited patches on larger islands. Of the smaller
islands Monuriki and Monu have been identified as Key
Biodiversity Areas (KBA’s) by Conservation International
because of their significance as critical refugia for the Fijian
crested iguana (Brachylophus cf. vitiensis) and tropical
dry forest (Conservation International, 2005; Olson, et
al., 2010). These islands, and particularly Monuriki, also
support the largest wedge-tailed shearwater (Puffinus
pacificus) population in Fiji. This paper outlines progress
made with the restoration of Monuriki Island by working
with the traditional land owners through an innovative and
inclusive conservation partnership.
Location
The uninhabited Monuriki Island (12.610ᵒ S, 177.034ᵒ
E) lies within the Mamanuca group in the province of
Nadroga, western Fiji (Fig. 1). This 40.4 ha volcanic
island reaches its peak at 177 m above sea level (Fig. 2).
Monuriki is owned by the Mataqali Vunaivi, the traditional
Fijian clan living on the nearby island of Yanuya. Monuriki
is listed under the National Biodiversity Strategic Action
Plan as a site of national significance due to its tropical
dry forest and two particular species of international
or national conservation concern, the Fijian crested
iguana (Brachylophus cf. vitiensis) and the wedge-tailed
shearwater (Puffinus pacificus) (Coulston, et al., 2010;
Olson, et al., 2010). Monuriki is the location of the third
largest population of the endemic Fijian crested iguana
(IUCN, 2014). This iguana is listed on CITES Appendix
I, as Critically Endangered by the IUCN Red List (IUCN,
2014), and Endangered by the US Fish and Wildlife
Service; it is protected in Fiji under the Endangered and
Protected Species Act (2002). Monuriki Island crested
iguanas are genetically distinct from all other crested
iguana populations (Keogh, et al., 2008), and the 2008
Iguana Species Recovery Plan (Harlow, et al., 2008)
prioritised Monuriki as the single most important site
In:
552C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 552–557. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Fisher, et al.: Recovery of native species on Monuriki Island
for immediate conservation action for this taxon. These
iguanas were discovered only in 1980. At that time there
was “a high density of iguanas” on Monuriki (Gibbons,
1984); however, less than 20 years later a survey revealed
fewer than 100 iguanas remained (Harlow & Biciloa,
2001). A more recent survey indicated the population had
dropped precipitously further with more extensive surveys
reporting only eight individuals found in 2003 (Harlow, et
al., 2007). The island also hosts several nesting colonies
of the wedge-tailed shearwater (Puffinus pacificus), a
species known from seven islands in Fiji and among which
Monuriki supports the most significant population. These
sea birds excavate burrows, often in the fragile coastal
strand substrates, to rear their chicks. It was estimated that
more than a thousand pairs of wedge-tailed shearwaters
annually breed on this island (Rasalato, et al., 2012).
Exotic faunal and floral species have invaded many of the
islands in Fiji and pose a serious environmental threat to
Monuriki’s native biodiversity.
Threats to native species
Fire is an anthropogenic threat to Monuriki, due to the
island’s small size and lack of natural ignition sources.
Exacerbating this is the threat of exotic non-native species.
The history of exotics on the island of Monuriki may date
back more than 3000 YBP. Pacific rats (Rattus exulans)
were the first exotic species introduced to this island, most
likely as stowaways with the early human arrivals (Roberts,
1991). This adaptable species has been able to sustain
itself on most islands left unchecked until eradication is
implemented. Rats are known to prey on eggs and chicks of
nesting birds, as well as lizards, juvenile tuatara (Sphenodon
punctatus), and seeds (Towns, et al., 2006). Domesticated
goats (Capra hircus) were established on Monuriki during
the 1970s. Originally brought as livestock, they provided an
income for the Yanuya owners of the island. As voracious
grazers, goats denuded the island of its undergrowth and
ate the seedlings of forest trees, and leaf litter causing
serious habitat degradation and severe erosion. The dry
forest habitat may recover from infrequent dry season
burning in the next rainy season provided seedlings are left
intact. However, when goats are present the seedlings are
grazed, preventing the regeneration of these native plants
and trees. This causes the endangered dry forest habitat
to convert to a mostly non-native composition while any
remaining mature native trees senesce and eventually
die. Following this cycle with fire and goat grazing on
Monuriki, the lack of native food plants posed a threat to
the diminished population of iguanas. Most of the surviving
vegetation was unpalatable to both goats and iguanas. The
open ground left by the continual grazing created space for
opportunistic invasive exotic and unpalatable plants to take
Fig. 1 Map of Fiji with arrow showing the location of
Monuriki Island.
hold, including the native but invasive vaō (Neisoperma
oppositifolium). Normally found in low abundance, vaō
became overabundant with disturbance and an increase in
light through the canopy increasing its representation in
the forest composition. Goats also threatened the groundnesting shearwaters by trampling nests, and causing the
collapse of fragile burrows containing eggs, chicks, and
nesting adults. The loss of insulating vegetation (leaf litter
and understorey structure), moderating water runoff and
erosion during heavy rain events, also potentially reduced
shearwater breeding success due to inundation of burrows
and nests. Most documented extinctions and current
causes of declining numbers of Pacific island birds result
from the effects of invasive alien species, and particularly
vertebrates such as rats and goats (McCreless, et al., 2016).
Tourism and poaching are additional disturbances to
this island which greatly impact iguanas, and increased
foot traffic during the breeding season might impact
wedge-tailed shearwaters, and other natives such as the
banded rail (Gallirallus phillippensis), and peregrine
falcon (Falco peregrinus). Monuriki is the site where the
award-winning 1999 Dreamworks movie “Castaway” was
filmed, and subsequently the island has become popular as
a tourist stop with the remaining movie set maintained as a
primary attraction. Ecotourism is the major contributor to
Fiji’s economy in this region. An estimated 70–100 tourists
visit the island daily although the number could be much
larger some days. Local resorts are still removing iguanas
for display on other nearby islands as an “ecotourism”
prop to draw in customers, and for tourist activities such
as staged photos.
Restoration plans and community conservation
In the late 1990s, the first efforts to conserve and
restore Monuriki were discussed. In 1998, 2000, and 2003,
surveys of Monuriki detected a rapid decline in the iguana
population as a result of continued major habitat degradation
by goats, with only adult iguanas being detected and no
evidence of recent recruitment (Harlow, et al., 2007). The
landowners were approached on at least three occasions to
remove goats from Monuriki, but declined to participate
each time. In 2004 the IUCN Iguana Specialist Group (ISG)
met in Fiji to discuss the current impacts on Monuriki, as
part of identifying potential conservation actions possible
for the species, including captive breeding. Through the
development of the IUCN Crested Iguana Recovery Plan
in 2008, conservation action steps for Monuriki were fullydeveloped and finally implemented. In 2009, BirdLife
International undertook surveys that documented rats and
goats to be major threats to nesting seabirds on Monuriki
and nearby Monu and Kodomo islands (NTF, 2012). It was
concluded that, if left unchecked, the persistence of rats and
goats would lead to the loss of the dry forest and nesting
seabirds on Monuriki. With the endorsement of the Yanuya
Fig. 2 View of Monuriki Island from 2017, looking to the
south-west.
553
Island invasives: scaling up to meet the challenge. Ch 3C Strategy: Outcomes
Village Chief and chief landowner, Sitiveni Drigi, an
eradication programme was therefore implemented during
2010–2011, specifically targeting invasive rats and feral
goats remaining on Monuriki (NTF, 2012). The goal of this
programme was to restore an ecosystem suitable for the
wedge-tailed shearwater, birds native to Fiji’s tropical dry
forest, and the critically endangered crested iguana. Regular
post-eradication monitoring, along with environmental
education campaigns specifically targeting biosecurity
among island users including tourists and fishermen,
were critical components to sustaining the eradication
results and restoration outcomes (Donlan & Keitt, 1999).
The eradications were designed in collaboration with
eradication experts (i.e., Pacific Invasives Initiative), local
land owners, the Provincial and National government, and
consultants. Two years of monitoring on Monuriki (and
Kodomo) did not detect rats and goats and by the end of
2013 their eradication was confirmed.
METHODS
Consultation
Many stakeholders were included in discussions
regarding the invasive species eradication plan. The plan
was launched as a community-based effort to successfully
restore and protect the natural state of Monuriki. The
involvement of the landowning unit in the village of
Yanuya (the Mataqali Vunaivi) and their chief, the late
Taukei Yanuya (Ratu Sitiveni Drigi) played a pivotal role
in ensuring the restoration and protection of Monuriki was
supported by the Yanuya community.
Other stakeholders since 2009 include (but are not
limited to):
● BirdLife International
● Commissioner Western’s office of the Fiji
Government
● Community members and village groups
● Critical Ecosystem Partnership Fund
● David & Lucile Packard Foundation
● Department of Heritage and Arts
● Disney Conservation Fund
● School of Geography at the University of the South
Pacific
● Iguana Specialist Group (IUCN)
● International Iguana Foundation
● Kula Wild Adventure Park
● Mamanuca Environment Society
● Ministry of Agriculture (Fiji)
● Ministry of Local Government, Housing &
Environment (Fiji)
● Nadroga/Navosa Provincial Office
● National Trust of Fiji Islands
● NatureFiji-MareqetiViti
● New Zealand Department of Conservation
● Pacific Invasives Initiative
● San Diego Zoo Global
● South Sea Cruises
● Survivor Entertainment Group
● Taronga Conservation Society Australia
● US Embassy Suva, Fiji, Regional Environmental
Affairs Office
● US Geological Survey, Western Ecological Research
Center
● Yanuya Rugby Team.
554
Consultations with landowners were (and will continue
to be) conducted before, during and after each activity or
site visit. Transparency of all information and intentions of
any actions are disclosed to stakeholders (BirdLife, 2011b).
From 2009–2010, discussions that addressed goat
grazing on the island were held with landowners. Although
the BirdLife surveys in 2009 confirmed goats to be a
significant factor in the decline of the native species, goats
were also a contributor to the village of Yanuya’s financial
income through market sale or occasional harvest for
meat. Therefore, compensation for the village of Yanuya to
halt goat grazing was agreed upon by the community and
supported by various stakeholders. In 2010 a Memorandum
of Understanding (MOU) was signed with the Mataqali
Vunaivi, National Trust, and Kula Eco Park (now Kula
Wild Adventure Park) for rat and goat removal, and iguana
harvest for the captive breeding programme. The National
Trust of Fiji and BirdLife International jointly carried out
operations to eradicate rats in August of 2011 and goats
between June 2010 and November 2011. At the same time,
it was decided the best way to conserve the crested Fijian
iguana was to harvest 10 sexually mature adult pairs (N =
20) from the remaining iguanas on the island for captive
breeding and subsequent reintroduction. From April 2010
to February 2012 Monuriki iguanas were collected and
brought to captive breeding facilities located at Kula Eco
Park as part of the MOU.
Goat removal and eradication
The local Yanuya Rugby Team was employed to
muster and catch goats on the island utilising mustering
routes and techniques established by the local communities
from decades of catching goats on Monuriki. From June
to November 2010, 151 goats were mustered from the
island over 12 days, and as of January 2011 an estimated
20 goats remained (BirdLife, 2011a). Captured goats were
taken to the Viti Levu mainland for sale. Two professional
hunters from New Zealand using trained dogs eliminated
more than 50 additional goats over an 11-day period in
September 2011. A final four day follow up hunting effort
in November 2011 detected no additional goats (BirdLife,
2011a). To compensate for expected revenue loss of these
animals, the owners received FJ$100 per goat. Posteradication monitoring of the forest vegetation using fixed
photo points and of the wedge-tailed shearwater population
was conducted to assess the response to the goat eradication
(Rasalato, et al., 2012).
Rat eradication
Eradication of Pacific rats (Rattus exulans) was
carried out by delivering specially formulated rodenticide
(brodifacoum at 20 ppm) baits (PestOff 20R) from a
helicopter using standard procedures and equipment
including a specifically designed spreader bucket
calibrated to the required application rate (20 kg/ha) and
GPS (Seniloli, et al., 2011). To determine the success of
the rat eradication, a series of transects with rat-trap (Victor
Professional) and rat-detection stations were created in
two to three main locations across the island for each
monitoring event in 2012, 2013, 2015, 2016, and 2017. The
transects were set up in areas of the eastern, northern, and
south-western beaches with between 10 and 20 stations for
each transect (Rasalato, et al., 2012; Fig. 3). Each station
comprised at least one snap trap, but the first assessments
also included a peanut butter wax tag, a tracking tunnel
and a second snap trap. The peanut butter wax tags were
nailed to trees at random heights so as to reduce hermit
crab access. Ink pads were placed in tracking tunnels with
roasted coconut placed on the pads to act as baits. Snap
traps were also baited with roasted coconuts and positioned
to minimise non-target interference (e.g. hermit crabs).
Fisher, et al.: Recovery of native species on Monuriki Island
These monitoring stations were set-up for three trap nights
and were maintained and checked daily for any signs of rat
activity (Rasalato, et al., 2012).
Biosecurity control
There is on-going training and outreach to the local tour
companies about the conservation activities on Monuriki
Island. Furthermore, biosecurity protocols were established
to help reduce the potential negative impacts of the tours
and other visitations to the island. A biosecurity plan
prepared in 2013 (Thaman & Niukula, 2013) is reviewed
and updated every two years (Seniloli, et al., 2015).
The plan includes three main biosecurity procedures:
preventing the entry of invasive alien species, systematic
checking for such species, and rapid response procedures
if any are found. Measures include the establishment of a
community-based ranger programme to train local rangers
on invasive species surveys, response methods and the
prevention of wildlife poaching.
Kula Eco Park captive breeding programme
Concurrently with the rat and goat eradication efforts
(2010‒2011), 20 Monuriki iguanas were harvested and
brought to Kula Eco Park, on the main Fijian island of Viti
Levu, to develop a captive breeding colony. Pairs were
successfully bred in managed care with the intention of
re-introducing the offspring to their home island once the
forest vegetation had recovered from grazing.
Species recovery
Monitoring for native species recovery has taken
place so comparisons to pre-eradication surveys can be
conducted. This includes using standard protocols for the
iguanas and shearwaters, and recording other incidental
species recoveries (Harlow, et al., 2007; Rasalato, et al.,
2012). Vegetation surveys were conducted in fixed plots
prior to the mammal eradications, and these plots were
resurveyed in 2016, after the eradications, following the
same survey protocols (Harlow, et al., 2007).
RESULTS
Goat removal and eradication
Since the goat removal and eradication was undertaken,
there have been no detections of goats on the island during
the last five years (through to 2017), confirming this action
to be a success.
Rat eradication
In March 2012, five months after the helicopter spread
of rodenticide, no rats were trapped nor were there any
obvious signs of rat presence such as droppings, gnawed
fruits or sightings (Rasalato, et al., 2012). Similarly, no
indications of rat presence were found during subsequent
assessments to date (2013, 2015, 2016, 2017), confirming
eradication of Pacific rats from Monuriki Island and an
ongoing rat free status generally.
Biosecurity control
Ships/vessels, commercial and private, are required
to abide by the biosecurity measures detailed in the most
recent version of the biosecurity plan (Seniloli, et al.,
2015). These measures include setting up and maintaining
bait and trap stations near wharfs and landing sites, regular
checks of vessels for alien species stowaways, proper
storage of food and regular decontamination of equipment
including footwear; detection of any alien species on the
island requires immediate reporting to a designated regional
support centre. Although outlined, implementation of all
of these actions has been slow, due to lack of resources.
Camera stations to detect and assess risk of reinvasion,
especially by rodents, were set up to monitor tourist
visitations at designated areas of Monuriki. Some cameras
are obvious and others are hidden; this programme helped
identify what items are being brought ashore. Community
awareness and involvement in implementing and enforcing
biosecurity measures has been important in preventing
additional invasive species from establishing on the island.
This includes the cooperation of tour boats and yachts that
must now follow biosecurity guidelines by having to report
to local landowners or designated personnel and crew for
a biosecurity briefing before anchoring near the island,
although this is still to be fully enforced (NTF, 2012).
Kula Eco Park captive breeding programme
After four wet seasons following goat removal, the
vegetation of Monuriki showed significant recovery, so
we initiated the reintroduction of iguanas from Kula Eco
Park. In mid-May 2015, 32 captive-bred crested iguanas,
all implanted with unique PIT tags, were released into
four different areas on Monuriki Island (Chand, et al.,
2016) after a major community ceremony highlighting
this milestone in the programme. Community members
participated in the release of the iguanas and the event has
been recorded in a video documenting the story (<https://
vimeo.com/163325268>). In February 2017, 32 additional
captive bred juveniles along with 16 of the original wild
caught adults were released; 10 of each group were tracked
for five months to measure post-release survivorship.
Because this event signalled the end of the captive
breeding programme and to thank the community for their
permission and participation, a second major community
event involving many levels of the Government was
planned around the release. Only a few young iguanas
remain in captivity for release in 2018 or 2019.
Released iguanas were monitored in 2017 and any wild
captured individuals or recaptures from the 2015 release
were measured and weighed to document post-release
growth and general health. Transmitters used to help track
released iguanas were removed before final re-release
after the five-month period. Currently, the crested iguana
population on Monuriki is recovering with the release of
captive breeding programme animals, and naturally with
existing wild animals (see below). Due to this success the
captive breeding programme was ended on 24 February
2017, after the final release of the remaining 16 wild
founder iguanas. Overall, between 2015 and 2017, a total
of 80 iguanas, including the founders, were released into
the wild.
Species recovery
Fig. 3 Monuriki Island showing the transects with the rat
sampling stations on the west and east beaches.
Between February and June 2017, 35 wild iguanas (not
passing through captivity) were caught and marked. Many
555
Island invasives: scaling up to meet the challenge. Ch 3C Strategy: Outcomes
of these were young animals that would have hatched after
the rat eradication was completed. This sample of iguanas
revealed that, within five years, the small remaining
population of iguanas is reproducing and recruiting back
into the recovering habitat.
A monitoring survey was conducted for wedge-tailed
shearwater nests post-eradication. Out of the 159 burrows
searched, 110 (69%) were occupied by chicks and one
had an egg. Pre-eradication wedge-tailed shearwater
nest site occupancy was 41% whereas post-eradication
this had risen to 69% occupancy (Rasalato, et al., 2012).
The size of the wedge-tailed shearwater population on
the south-western beach colony during the 2011‒2012
seasons was estimated to be 1383 breeding pairs (Rasalato,
et al., 2012). As an additional positive outcome, banded
rails (Gallirallus philippensis) and peregrine falcons
(Falco peregrinus; critically endangered in Fiji) are more
frequently encountered on Monuriki following rat and goat
eradications.
Habitat monitoring to document terrestrial dry forest
habitat recovery following goat removal was conducted
in 10 previously established 100 m2 plots in the lowlands
behind Rogua Beach. Vaō trees dominated the forests
on Monuriki before the eradications. This species is
unpalatable to both goats and iguanas. However due to
goats’ removal of the edible understory and lower branches
of the taller edible trees, the goats were now starting to
chew on the stems of the vaō. During the presence of goats
and rats, vaō comprised 91.3% of the seedlings per plot,
but only 58.7% of the seedlings per plot after goats and
rats were removed. Individual vaō trees (> 2 m in height)
had decreased from 155 to 78 individuals across the plots
after mammal removal. Although this plant is native, it
spreads like an invasive with disturbance. There is no plan
to control it, but the recovery of the other native forest
diversity will reduce its cover over time. For example,
only three other tree species seedlings were documented
when goats were present; now there are about 11 species of
seedlings present per plot including Hibiscus, Diospyros,
and Pongomia. Ground cover of vine (some invasive but
edible such as Passiflora foetida) increased from 0% with
goats to 30.2% after goats were removed. Habitat recovery
was determined to be successful through these habitat
surveys, although no similar repeated surveys have been
done on the higher slopes where a greater diversity of dry
forest trees are present.
DISCUSSION
Overall, the Monuriki Island restoration programme
has been a great success and a model for Fiji and other
nations in the region. Table 1 is a timeline that reviews
the overall impacts to the island and the major milestones
relevant to the plan. Monitoring will continue over the
next decade as the tree canopy expands and invasive plants
continue to be removed from the island. On the two main
beaches, invasive plants are being removed manually
and dry-forest trees are being planted within the coconut
groves that persist. Over time these areas will recover to
dry forest also. Most of the obvious habitat damage from
cyclone Winston was on these coconut groves, breaking
them in half, and the native trees on the ridges that were
damaged seem to be recovering well from the event.
Community engagement
This programme has continued with renewed investment
in the local community through the development of a
regional Ranger Programme for Fiji iguana conservation.
Various stakeholders have supported the development
and training of local residents to act as regional iguana
experts and habitat managers. This programme includes
556
Table 1 Timeline of impacts and recovery actions on
Monuriki Island.
ca.
3,000
years ago
Goats introduced to Monuriki
1970s
Iguanas discovered on Monuriki
1980
Iguanas captured for resort displays
1980s–2000s
1999
Castaway was filmed
IUCN Iguana Specialist Group Suva, 2004
Fiji
Crested Iguana Species Recovery Plan 2008
Discussions with landowners over goats 2009–2010
Goat eradication operation
2010–2011
20 Monuriki iguanas harvested for 2010–2012
captive colony
Rat eradication operation
2011
Goat and rat eradication confirmed
2013
32 captive-bred iguanas released
2015
Cyclone Winston (Category 5)
2016
Reality TV Show now being filmed on 2016–2017
Monuriki
32 captive-bred iguanas released
2017
16 remaining original founder iguanas 2017
released
Rattus exulans introduced to Fiji
capacity building by training local rangers in field survey
techniques, habitat management methods, reforestation
efforts, guest experience training for tourist interactions,
and anti-poaching. Support for the Rangers and other
local level science educators to attend conferences and
workshops (such as the IUCN Iguana Specialists Group)
geared toward engagement in conservation initiatives
for threatened species has continued to provide valuable
training and resources for long-term capacity building
within Fiji. Education and outreach materials have been
developed with the goal of reaching the regional children
through programmes for the classrooms. Visits to the
local communities to conduct outreach programmes have
provided additional opportunities to reach the local
communities and encourage their continued support in the
conservation of the native threatened species.
To protect the regenerating forest, the community
Rangers also established tourist hiking paths with the
intent of educating visitors about tropical dry forest habitat
while keeping their impacts on the island to a minimum.
Interpretive kiosks are being developed and will be
installed at the tourist beach.
Additional threats
In February of 2016, Tropical Cyclone Winston
passed through Fiji with a peak intensity of ten-minute
sustained winds of 230 km/hr that removed a significant
proportion of the canopy leaves from trees on Monuriki
Island (Fig. 4). Terrestrial surveys conducted in March
2016, indicated that the iguanas and birds were still present
and increasing, but that long-term studies after this storm
event would be critical for helping to understand species
resiliency and recovery in the wake of massive tropical
storms. BirdLife continues to monitor the wedge-tailed
shearwater population and, while no assessment was made
following cyclone Winston (which occurred during the
Fisher, et al.: Recovery of native species on Monuriki Island
chick feeding phase), population measures assessed will
inform recruitment trends. The final 16 original founding
iguanas, along with their 32 offspring produced as part
of the captive breeding programme, were released after
this extreme weather event (Table 1). Robust population
monitoring for the iguanas will take place during 2018 to
assess the longer-term survivorship of the released iguanas
and track recovery of the remaining wild individuals.
2016‒2017 Reality TV show, Survivor, now being
filmed on Monuriki
Regional conservation efforts for Fijian iguanas have
continued and grown to include additional partners. One
such collaborator, Survivor Entertainment Group, has been
strongly supportive of the programme since their arrival to
the region. Although this partner may not be conventional
when considering species conservation and habitat
restoration, they have embraced and supported our efforts
to save the Fijian iguanas from extinction and have assisted
in efforts to continue local level engagement. There might
be impacts of the filming activities on the shearwaters,
but these are hard to measure, and are being minimised
by marking active nest sites and putting in avoidance
trails to move film crews and contestants around these
sensitive sites. By providing significant local employment
opportunities and incorporating and investing in biosecurity
training methods and native species conservation as part of
their local strategic plan they have become an advocate for
these restoration and recovery efforts in the region.
ACKNOWLEDGEMENTS
The following people helped in the field to accomplish
this work, these include: B. Nagle, G. Coulston, P. Biciloa,
A. Naikatini, T. Tuamotu, C. Rochester, S. Pasachnik, S.
Hathaway, E. Matatia, N. Felstead, S. Anstey, M. Glaser,
K. Waga, M. Segaidina, and B. Rashni. Additional help
was received from the National Trust of Fiji Islands,
Nadroga Provincial Office, Mataqali Vunaivi, and
Mamanuca Environment Society. Funding was provided
by the Polynesia/Micronesia CEPF, International Iguana
Foundation, Dutch Iguana Society, San Diego Zoo
Global, Taronga Conservation Society Australia, Survivor
Entertainment Group, and the USGS Ecosystems Mission
Area. BirdLife International would like to thank the David
and Lucile Packard Foundation for supporting this work
and for the goat eradication, Glen Coulston, and, in its
implementation, Ian Warfe and Luke Robertson from
the Department of Conservation, New Zealand. The
Fig. 4 Rogua Beach on Monuriki in March of 2016 after
Cyclone Winston. Broken and denuded trees are seen
along the beach and up the hillslope.
professional services of John Oakes, Otago, New Zealand
for supporting technically the aerial baiting operation and
the team of Island Hoppers, Nadi, Fiji for the provision of
helicopter services.
REFERENCES
BirdLife International. (2011a). Monuriki and Kadomo Goat Eradication
Operation Report. Unpublished report. Suva, Fiji: BirdLife
International, Pacific Secretariat.
BirdLife International, National Trust of Fiji, Pacific Invasives Initiative.
(2011b). Operational Plan for the Eradication of Feral Goats in the
Mamanuca Islands, Republic of Fiji. Unpublished report. Suva, Fiji:
BirdLife International, Pacific Secretariat.
Chand, R., Niukula, J., Vadada, J., Fisher, R., Lovich, K., Pasachnik,
S., Rasalato, S., Thaman, B., Seniloli, E., Tuamoto, T., Yanuya, T.
and Harlow, P. (2016). ‘Captive Breeding and Re-introduction of the
Monoriki Island Crested Iguana in Fiji’. In: P.S. Soorae (ed.) Global Reintroduction Perspectives: 2016. Case-studies from Around the Globe.
Gland, Switzerland: IUCN/SSC Reintroduction Specialist Group and
Abu Dhabi, UAE: Environment Agency-Abu Dhabi.
Conservation International. (2005). Conservation Outcomes: Polynesia
and Micronesia, Fiji Inset, Draft Map, 1:3,200,000. Washington,
DC, USA: Conservation International and The Critical Ecosystem
Partnership Fund.
Coulston, G., Ratu, M. and Nagle, B. (2010). Monuriki Island Goat
Removal; Report of a Feasibility Study Visit. Unpublished report.
Auckland, New Zealand: Pacific Invasives Initiative.
Donlan, C.J. and Keitt, B.S. (1999). ‘Using research and education to
prevent extinction’. California Coast and Ocean 15: 20–23.
Gibbons, J. (1984). ‘Iguanas of the South Pacific’. Oryx 18 (2): 82–91.
Harlow, P.S. and Biciloa, P.N. (2001). ‘Abundance of the Fijian
crested iguana (Brachylophus vitiensis) on two islands’. Biological
Conservation 98: 223–231.
Harlow, P.S., Fisher, M., Tuiwawa, M., Biciloa, P.N., Palmeirim, J.M.,
Mersai, C., Naidu, S., Naikatini, A., Thaman, B., Niukula, J. and
Strand, E. (2007). ‘The decline of the endemic Fijian crested iguana
Brachylophus vitiensis in the Yasawa and Mamanuca archipelagos,
western Fiji’. Oryx 41: 44–50.
Harlow, P.S., Hudson, R. and Alberts, A. (2008). ‘Fijian Crested Iguana
Brachylophus vitiensis Species Recovery Plan 2008–2012’. IUCN
Species Survival Commission, Iguana Specialist Group. < http://www.
iucn-isg.org/publications/actionsplans/>.
Keogh, J.S., Edwards, D.L., Fisher, R.N. and Harlow, P.S. (2008).
‘Molecular and morphological analysis of the critically endangered
Fijian iguanas reveals cryptic diversity and a complex biogeographic
history’. Philosophical Transactions of the Royal Society, Series B. Vol.
363 (1508): 3413–3426.
McCreless, E.E., Huff, D.D., Croll, D.A., Tershy, B.R., Spatz, D.R.,
Holmes, N.D., Butchart, S.H.M. and Wilcox, C. (2016.) ‘Past and
estimated future impact of invasive alien mammals on insular threatened
vertebrate populations’. Nature Communications 7: 12488.
NTF. (2012). Restoring the Native Vegetation of Monuriki Island, Fiji.
Final Report Submitted to Critical Ecosystems Partnership Fund. Suva,
Fiji: National Trust of Fiji Islands.
Olson, D., Farley, L., Patrick, A., Watling, D., Tuiwawa, M., Masibalavu,
V., Lenoa, L., Bogiva, A., Qauqau, I., Atherton, J., Caginitoba, A.,
Tokota’a, M., Prasad, S., Naisilisili, W., Raikabula, A., Mailautoka, K.,
Morley, C. and Allnutt, T. (2010). ‘Priority forests for conservation in
Fiji: landscapes, hotspots and ecological processes’. Oryx 44: 57–70.
Rasalato, S., Niukula, J., Thaman, B., Mateiwai, M. and Vadada, J. (2012).
Post-eradication Monitoring Survey on Monuriki Island, Mamanuca
group. Unpublished field report. Suva, Fiji: BirdLife International,
Pacific Secretariat.
Roberts, M. (1991). ‘Origin, dispersal routes, and geographic distribution
of Rattus exulans, with special reference to New Zealand’. Pacific
Science 45: 123–130.
Seniloli, E. Rasalato, S. Tuamoto, T. and Cranwell, S. (2011). Feasibility
of Rat and Goat Eradication in the Mamanuca group Fiji. Unpublished
report. Suva, Fiji: BirdLife International, Pacific Secretariat.
Seniloli, E., Rasalato, S. and Tuamoto T. (2015). Draft Biosecurity Action
Plan for Monuriki and Kadomo Islands, Mamanuca Group, Fiji. Suva,
Fiji: Nadroga/Navosa Provincial Council, BirdLife International,
National Trust of Fiji Islands, NatureFiji-MarqetiViti.
Thaman, B. and Niukula J. (2013). Draft Biosecurity Action Plan for
Monoriki and Kadomo Islands, Mamanuca Group, Fiji. Suva, Fiji:
Nadroga/Navosa Provincial Council, BirdLife International, National
Trust of Fiji Islands, NatureFiji-MarqetiViti,
Towns, D.R., Atkinson, I.A.E. and Daugherty, C.H. (2006). ‘Have
the harmful effects of introduced rats on islands been exaggerated?’
Biological Invasions 8: 863–891.
557
R. Griffiths, E. Bell, J. Campbell, P. Cassey, J.G. Ewen, C. Green, L. Joyce, M. Rayner, R. Toy, D. Towns, L. Wade, R. Walle and C.R. Veitch
Griffiths, R.; E. Bell, J. Campbell, P. Cassey, J.G. Ewen, C. Green, L. Joyce, M. Rayner, R. Toy, D. Towns, L. Wade, R. Walle and C.R. Veitch.
Costs and benefits for biodiversity following rat and cat eradication on Te Hauturu-o-Toi/Little Barrier Island
Costs and benefits for biodiversity following rat and cat eradication on
Te Hauturu-o-Toi/Little Barrier Island
1
R. Griffiths , E. Bell2, J. Campbell3, P. Cassey4, J.G. Ewen5, C. Green6, L. Joyce7, M. Rayner8, R. Toy9, D. Towns10, L.
Wade11, R. Walle7 and C.R. Veitch12
1
Island Conservation, 561 Woodcocks Rd, RD 1, Warkworth 0981, New Zealand. <rgriffiths@islandconservation.org>.
Wildlife Management International Ltd, PO Box 607, Blenheim 7240, New Zealand. 3Ecological Research Associates,
2 Harewood Grove, Upper Hutt 5019, New Zealand. 4Centre for Conservation Science and Technology, and School
of Biological Sciences, University of Adelaide, Adelaide, South Australia, Australia. 5Institute of Zoology, Zoological
Society of London, Regents Park NW1 4RY, London, UK. 6Department of Conservation, PO Box 68-908, Parnell,
7
Auckland, 1141, Auckland. Department of Conservation, 12/30 Hudson Rd, Warkworth 0984, New Zealand. 8Auckland
Museum, Private Bag 92018, Auckland, 1141, New Zealand. 978c Little Sydney Road, Motueka, New Zealand. 10Institute
for Applied Ecology, Auckland University of Technology, Auckland 1142, New Zealand. 11Hamilton Road, RD 2,
Warkworth 0982, New Zealand. 1248 Manse Road, Papakura, 2113, New Zealand.
2
Abstract Considerable benefits can be achieved for indigenous biodiversity when invasive vertebrates are removed from
islands. In New Zealand, two logistically challenging eradications were undertaken, one to remove cats (Felis catus)
and the other Pacific rats (Rattus exulans) from Te Hauturu-o-Toi/Little Barrier Island (Hauturu). Here we document the
short- and long-term impacts of these interventions on the biodiversity of Hauturu. We also assess the extent to which
predicted outcomes were reflected in the measured responses for a wide range of species. Short-term impacts of the
eradication program encompassed individual mortality for some native species but no measurable impact to populations.
In contrast, at least 11 native vertebrates and one native invertebrate species increased in abundance after rat and cat
removal. Fifteen of 34 plant species monitored had significantly more seedlings on Hauturu after rat eradication compared
with control islands, indicating future changes in forest composition. Several native species previously not recorded on
the island were discovered, including the New Zealand storm petrel (Fregetta maoriana) (formerly considered extinct),
the forest ringlet butterfly (Dodonidia helmsi) and eight species of aquatic invertebrate. The chevron skink (Oligosoma
homalonotum) has been found in increasing numbers and tuatara (Sphenodon punctatus), raised in captivity on the island,
are now re-established and breeding in the wild. These results illustrate an island gradually recovering after a long period
of modification. We conclude that more success stories such as Hauturu must be told if we are to allay the public’s
concerns about such eradication campaigns. And more public support is required if the conservation community is to
tackle invasive species at a scale commensurate with the threats they pose.
Keywords: conservation management, ecosystem, restoration, species recovery
INTRODUCTION
Worldwide, more than 1,000 invasive vertebrate
eradications have been successfully completed to prevent
biodiversity loss (DIISE, 2017) and many benefits to
species and ecosystems have been documented (Jones,
et al., 2016). However, eradication projects continue to
attract controversy (e.g. Howald, et al., 2010; Griffiths,
et al., 2012; Capizzi, et al., 2019) suggesting that, despite
transparent consultation processes, sectors of the public
remain unconvinced of the relative cost benefits of this
conservation strategy.
To illustrate the value of invasive vertebrate eradication,
we present the short- and long-term impacts on biodiversity
following the removal of cats (Felis catus) and Pacific rats
(Rattus exulans) from Te Hauturu-o-Toi/Little Barrier
Island (hereafter referred to as Hauturu). Specifically, we
ask whether the claimed benefits of cat and Pacific rat
eradication were met.
The eradication of cats from Hauturu raised little
public concern and, under New Zealand environmental
law, did not require consent. In contrast, the proposed rat
eradication raised cultural and environmental concerns
and, because rodent bait was broadcast by helicopter,
required local government consent (Resource Management
Act 1991). Some members of the public were opposed to
the aerial application of rodent bait and some Māori iwi
(tribes) contested the removal of rats because of their
cultural significance. Consequently, public hearings were
held and an Assessment of Environmental Effects (AEE)
(Griffiths, 2002) was presented to a panel of independent
commissioners. The AEE identified the legal mandate for
the removal of rats and the risk to native species if rats
remained. The application was approved as the potential
benefits to native biodiversity were judged to outweigh the
short-term environmental costs.
Cats were removed from Hauturu in an operation that
spanned four years from 1977 to 1980. To support this
work, a 67 km long track network was established across
the island and three huts built at strategic locations (Veitch,
2001). Leg-hold traps and baits containing the toxin 1080
were the principal methods employed to remove cats,
although cage traps, the introduction of pathogens and
dogs were also used (Veitch, 2001). Mitigation of potential
impacts on non-target species was undertaken through
careful placement of traps.
Rats were eradicated in 2004 by the New Zealand
Department of Conservation (DOC) in an operation
utilising the aerial application of Pestoff 20R™ rodent
bait containing brodifacoum at 20 ppm (Griffiths, 2004).
Rodent bait was applied twice by three helicopters in two
successive operations during winter, the first on 8 and 9
June and the second on 12 July. At the same time, baits
were placed in bait stations within all buildings and huts on
the island. The operation used a total of 55 tonnes of rodent
bait with rates for the first and second bait applications
averaging 11.7 kg/ha (ca. 1 bait per 1.7 m2) and 6.16 kg/
ha (ca. 1 bait per 3.2 m2), respectively, across the island.
The success of the eradication operation was confirmed in
January 2006 after extensive monitoring both on and off
the track network across the island with tracking tunnels,
spotlight searches and indicator dogs (Griffiths, 2006).
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
558
up to meet the challenge, pp. 558–567. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Griffiths, et al.: Costs and benefits for biodiversity on Little Barrier Island
Our findings are based on research and monitoring
completed on the island from 1962 to the present. We
review published and unpublished studies, but also present
new data that illustrate changes following rat and cat
removal.
seabirds. It is likely that other species such as the black
petrel (Procellaria parkinsoni) would have suffered a
similar fate had cats not been removed.
STUDY AREA
A literature search of published and unpublished
monitoring, undertaken to measure the environmental
impacts of cat and rat eradication, was conducted but
also included general research and monitoring completed
on Hauturu for other reasons. We assessed impacts as
measured costs, measured benefits, and unknown costs
or benefits to biodiversity. Unknown costs and benefits
were largely a function of an absence of monitoring prior
to and following eradication, and/or due to environmental
changes unrelated to the eradications. For example,
Hauturu supports breeding populations of the New Zealand
lesser short-tailed bat (Mystacina tuberculata) and longtailed bat (Chalinolobus tuberculatus), both endemic to
New Zealand. Bat populations were not monitored before,
during or after cat and rat eradication.
Hauturu (3,083 ha), 36°11′56.76″S, 175°4′53.04″E is
almost midway between Great Barrier Island/Aotea and
the mainland (Fig. 1). Rugged and steeply dissected, the
island arose from the partly eroded core of a composite
volcanic cone that formed 1.5–3 Ma (Lindsay & Moore,
1995)(Fig. 2). Hauturu was first settled around the 14th C by
the descendants of the Maori ancestor and voyager Toi te
Huatahi and was occupied continuously until the arrival of
Europeans in the 1800s. Over this and the ensuing period
of European settlement, approximately one third of the
island (the south-west) was cleared, burnt and subjected to
grazing by sheep (Ovis aries) and cattle (Bos taurus) (most
of these areas have since reverted to native vegetation and
are now secondary successional forest or older). Because
of its unique forest and threats to the diverse birdlife, the
island was gazetted as New Zealand’s first Nature Reserve
in 1896 (Young, 2004).
Except for feral cats and rats, the island escaped
many of the invasive vertebrate introductions to the main
islands of New Zealand and, consequently, its fauna is
still largely representative of northern New Zealand prior
to European colonisation. By area, the island supports a
greater diversity of native birds and reptiles than any other
part of of New Zealand. Nonetheless, the introduction of
rats and cats to Hauturu had a huge impact. Pacific rats,
considered to have arrived early during the period of
Maori settlement (Campbell, 2011), likely extirpated most
of the small seabird species still seen on nearby rat-free
islands. Rat predation may also explain the absence of milk
tree (Streblus banksii), coastal maire (Nestegis apetala)
(Campbell, 2011), large land snails and slugs on the island
(Campbell, 2011). The introduction of cats sometime
around 1867 resulted in the extinction of the last population
of North Island snipe (Coenocorypha barrierensis) and
extirpation of the tieke (Philesturnus rufusater), greyfaced petrel (Pterodroma macroptera), and probably other
Fig. 1 Map showing the location of Hauturu.
METHODS
We first summarise the predicted outcomes (Tables
1 and 2), then collate previously published monitoring
results obtained from a Web of Science and Google Scholar
search completed on 24 January 2017 using the key words
‘Little Barrier Island’ and ‘Hauturu’. We then summarise
unpublished research and monitoring reports and other
unpublished data including the methods used (Table 3) and
analyse data on terrestrial birds (see mist-netting below).
Changes in island species composition after rat and cat
eradication were determined by comparing recent literature
with historical reports. For simplicity, results are grouped
by taxa: marine birds, terrestrial birds, reptiles, freshwater
fish, terrestrial invertebrates, aquatic invertebrates and
terrestrial plants.
Fig. 2 The rugged nature of Hauturu and locations of mist
nets (black dots) used to sample forest bird abundance
before and after rat eradication.
559
Island invasives: scaling up to meet the challenge. Ch 3C Strategy: Outcomes
Table 1 Predicted benefits of cat eradication from Hauturu with species marked* identified in Recovery Plans for threatened
species.
Native biodiversity Evidence
Black petrel
Direct evidence of cat predation and declining
breeding distribution
Cook’s petrel/titi
Direct evidence of cat predation and declining
breeding distribution
Tieke
Potential for reintroduction
Kakapo*
Kokako*
Grey-faced petrel
Potential for ex-situ management of Stewart
Island population heavily impacted by feral cats
Potential for reintroduction
Recolonisation expected
References
(Girardet, et al., 2001)
(Imber, et al., 2003a)
(Girardet, et al., 2001; Hoosen & Jamieson,
2003)
(Lloyd & Powlesland, 1994; Anon., 1996;
Elliott, et al., 2001)
(Innes & Flux, 1999)
(Girardet, et al., 2001)
Table 2 Predicted responses to rat removal from Hauturu as identified in the assessment of environmental effects (Griffiths,
2002), with conventions as in Table 1.
Native biodiversity Evidence
References
Species that faced local extinction in the absence of intervention
Cook’s petrel
Fledgling recruitment on Whenua Hou (Codfish Island) increased to (Imber, et al., 2003a)
90% following rat eradication
Tuatara*
Documented impacts on juvenile recruitment. Local extinction
(Gaze, 2001; Towns, et al.,
predicted by population models completed for Marotere and Taranga 2007; Cree, 2014)
Islands
Wetapunga*
No off-site data but examples of localised extinctions in other
(Towns, et al., 1990;
flightless crickets e.g. tusked weta (Motuweta isolata; Mercury
Sherley, 1998; Towns, et
Islands)
al., 2006)
No off-site data on impacts by Pacific rats on this species but video (Eckroyd, 1995)
Dactylanthus
evidence of inflorescence destruction by rats on Hauturu
taylorii*
Giant-flowered
Examples of localised extinctions from other archipelagos (e.g.
(Towns, et al., 2003)
broom
Marotere Islands)
Species predicted to benefit from intervention
Grey-faced petrel
Increases in abundance documented elsewhere after Pacific rat
(Towns & Atkinson, 2004;
removal (e.g. Korapuki Island, Stanley Island)
G Taylor pers. comm.)
Diving petrel
Localised extinctions reported in other archipelagos and increased
(Towns & Atkinson, 2004;
abundance following Pacific rat removal (e.g. Mercury Islands)
G Taylor, pers. comm.)
Tieke
Evidence of increased abundance after Pacific rat removal (e.g. Red (Robertson, et al., 1993)
Mercury Island)
Towns’ skink
Examples of localised extinctions in other archipelagos (e.g.
(Towns, et al., 2003)
Mokohinau, Marotere Islands) or confinement to refugia (Hauturu)
Duvaucel’s gecko Confined to refugia in presence of Pacific rat (Hauturu). More
(Towns, 1996; Hoare, et
terrestrial activity and increased abundance after Pacific rat removed al., 2007)
(e.g. Mercury, Marotere and Ohinau Islands)
Pisonia brunoniana Examples of recovery after Pacific rat removal (e.g. Mercury Islands) (Campbell & Atkinson,
2002)
and 10 other plant
species
Species that will possibly benefit from intervention
Short-tailed bats
No off-site models for this species. Potential release from competition
for invertebrates and Dactylanthus taylorii inflorescences
Long-tailed bats
No off-site models for this species. Potential release from competition
for invertebrates
Smaller native
No off-site models for these species. Potential release from
passerine birds
competition for invertebrates and nectar sources
Day-active skinks Increased abundance of selected species when Pacific rats removed (Towns, 1991)
(e.g. Mercury Islands) but no models for chevron and striped skinks
present on Hauturu
560
Griffiths, et al.: Costs and benefits for biodiversity on Little Barrier Island
Table 3 Methods used to evaluate short- and long-term changes to biodiversity subsequent to cat and rat eradication on
Hauturu.
Species Group
Bats
Indicator
Methods
Sources
(Daniel & Williams, 1984)
Species composition
Literature review
Species abundance
Anecdotal reports
Short term mortality over
the course of eradication
operations
Cat trapping data and carcass
searches before and after rat
eradication.
(Veitch, 2001; Griffiths, 2004)
Species composition
Literature review
(Hutton, 1868; Turbott, 1947; Girardet,
et al., 2001; Stephenson, et al., 2008;
Rayner, et al., 2009; Gaskin & Rayner,
2013; Rayner, et al., 2015)
Cooks petrel breeding success
and distribution
Monitoring of marked burrows 1971
to present
(Imber, et al., 2003a; Rayner, et al.,
2007b)
Black petrel breeding success
Monitoring of marked burrows 1971
to present
(E. Bell, unpubl. data; Imber, 1987;
Imber, et al., 2003b)
New Zealand storm petrel
Monitoring of marked burrows
(M.J. Rayner, unpubl. data; Ismar, et al.,
2015)
Short term mortality over
the course of eradication
operations
Cat trapping data and carcass
searches before and after rat
eradication
(Veitch, 2001; Griffiths, 2004)
Literature review and bird counts
(Hutton, 1868; Turbott, 1947; Girardet,
et al., 2001; Veitch, et al., 2019)
General species abundance
Mist-netting capture rates and bird
counts
(Girardet, et al., 2001; Veitch, et al.,
2019). Data analysis of unpublished
mist-netting data described below.
Tieke, hihi and tui abundance.
Distance sampling and bird counts
(Toy, et al., in press; Veitch, et al., 2019)
Kiwi
Call counts
(Wade 2009; Wade 2014a)
Short term mortality over
the course of eradication
operations
Cat trapping data and carcass
searches pre and post rat eradication
(Veitch, 2001; Griffiths, 2004)
Species abundance and
distribution.
Pitfall trapping (10 L plastic buckets
baited 24 h with tinned pear) and
search effort
(Brown, 2013)
Short term mortality over
the course of eradication
operations
Cat trapping data and carcass
searches before and after rat
eradication
(Veitch, 2001; Griffiths, 2004)
Freshwater fish
Species composition
Trapping and spotlight surveys
(Winterbourn, 1964; Wade, 2014b)
Terrestrial
invertebrates
Species composition
Anecdotal observations
(S.Wheatley, pers. comm.; R. Walle,
pers. obs.)
Wetapunga abundance
Detections per unit of search effort
(Green, et al., 2011)
Aquatic
invertebrates
Species composition
Benthic sampling and light trapping
(Winterbourn, 1964; Wade, 2014b)
Threatened
native and
invasive alien
plants
See production, seedling
recruitment; abundance and
distribution
(D. Havell, unpubl. data; Campbell,
Monitoring of seed set for D.
2011)
taylorii; search effort for other
threatened species and priority weeds
Canopy trees,
palms and
lianes
Juvenile recruitment on
rat-inhabited versus rat-free
islands, post-eradication
response, seedling response in
exclosures
Seedling numbers of 34 species
counted on marked linear plots,
twice before rat eradication and two
years after on Hauturu and on two
control islands with rats
Marine birds
Terrestrial birds Species composition
Reptiles
Mist-netting
Data collected from mist-netting completed before
and after rat eradication were used to assess the impact
of the application of rodent bait on the abundance and
composition of forest birds. As these data are not published
we summarise it here. Four trips were completed, one prior
(January 2004) and three subsequent (August/September
(Campbell, 2011)
2004, February 2005, August 2005) to the eradication
comprising 413 mist-netting events at 69 sites across five
valleys in an area of approximately 350 ha on the southwest side of the island (Fig. 2). Each trip lasted for between
five and seven whole days of mist-netting. Each mistnetting event had an average duration (±sd) of 399 minutes
(6 hours 39 minutes) ± 170 minutes and ranged from 06:11
561
Island invasives: scaling up to meet the challenge. Ch 3C Strategy: Outcomes
(opening time) to 19:47 (closing time). On average, the
median time a net remained open was 11:44 hrs.
In the analyses we assessed whether temporal factors
(e.g. year, season, day, time of day) were directly associated
with variability in the number of individuals and species
caught in mist-nets. All analyses were compiled in SAS
V.9.0. Generalised linear mixed models (with Poisson
distributed errors) were used to assess the variability
between both the total number of individuals (bird
abundance) and the number of species (species richness)
caught per mist-netting event with respect to temporal
factors.
With the data assumed to follow a Poisson distribution
due to its non-negative, count nature, we used Basic
Generalised Linear Models followed by a more complicated
Generalised Linear Mixed Effects Model to tease out
changes between years for individual species. Richness
and Shannon diversity were the variables being predicted,
with Year, Season, Total Number of Birds, and Corrected
Net Length as explanatory variables. Site and Net were
included in the model as random effects, separately and
together. The ‘best’ model was then used to fit the six
species most commonly caught as the predictor variable.
composition) analysed for forest bird species. The
only change measured was a significant increase in the
number of both bellbird Anthornis melanura and parakeet
(Cyanoramphus novaezelandiae and/or C. auriceps)
captured. In total, 1,570 birds (twenty-three species)
were caught in mist-nets. The total number of birds (bird
abundance) caught varied between seasons (F1,374 = 5.53, P
= 0.02) but did not differ between periods of mist-netting
completed before and after the bait applications targeting
rats (F1,374 = 0.34, P = 0.56). Number of species (species
richness) (F1,374 = 0.01, P = 0.93) and relative similarity in
the composition of forest birds caught in mist nets (F1,4 =
0.53, P = 0.51) did not differ significantly over the course
of the rat eradication. These data correspond with the
findings of a non-toxic bait trial completed ahead of the
rat eradication, that determined the risk to terrestrial bird
populations to be low (Greene & Dilks, 2004).
Reptiles
No individual mortality as a result of the cat and rat
eradications was documented. However, anecdotal evidence
and pitfall trapping (Fig. 3) suggest an unexplained decline
in skink numbers following cat eradication, prior to rats
being removed.
RESULTS
Freshwater fish
Potential and actual costs to native biodiversity
A freshwater fish survey conducted in 2000
detected redfin bullies (Gobiomorphus huttoni), banded
kokopu (Galaxias fasciatus) and longfin eel (Anguilla
dieffenbachii) (McGlynn, et al., 2000). No mortality
following bait application for rats was observed, but a less
extensive survey completed in 2009 detected only banded
kokopu and longfin eels (Wade, 2014b).
Seabirds
Thirteen Cook’s petrel (Pterodroma cookii) were
trapped in cat leg-hold traps and euthanised (Veitch, 2001).
The breeding success of Cook’s petrel decreased following
the removal of cats then increased after rat eradication. This
was hypothesised to be a function of mesopredator release
resulting in higher numbers of rats at higher elevations
after cat eradication leading to greater impacts on Cook’s
petrel breeding success (Rayner, et al., 2007b). No other
short term negative impacts on seabirds as a result of cat
and rat eradication were observed.
Terrestrial birds
Thirty-two kiwi (Apteryx mantelli) were caught in traps
during the cat eradication. Two were euthanised, the rest
released unharmed (Veitch, 2001). Three kiwi were found
dead after application of rodent bait (Fisher, et al., 2011)
and are presumed, based on the necropsy of one individual,
to have died from secondary poisoning. Despite the loss
of these individuals, no change in calling frequency was
observed in kiwi call count surveys completed after the rat
eradication (Wade, 2009).
Individual mortality following bait application to target
rats was documented for eight other terrestrial bird species
including blackbird (Turdus merula), robin (Petroica
australis), pukeko (Porphyrio porphyrio), kakariki
(Cynoramphus novaezelandiae and C. auriceps), harrier
(Circus approximans), kaka (Nestor meridionalis) and
morepork (Ninox novaeseelandiae) (Veitch, 2001; Fisher,
et al., 2011). Numbers of each species found after the rat
eradication are presented in Fisher et al. (2011). However,
no significant short-term population impacts were detected
in an analysis of bird count data collected over the course
of the cat eradication (Girardet, et al., 2001). Bird counts
from 2012 to 2017, after rat eradication, using the same
methods as Girardet, et al. (2001), showed no significant
change in overall abundance but significant changes in the
abundance of some species (C.R. Veitch unpubl. data).
Data collected from mist-netting conducted before
and after rat eradication showed no significant change
(either increase or decline) in any of three components
of catchability (bird abundance, species richness, species
562
Invertebrates
As predicted, no negative impacts on invertebrates
were observed.
Plants
Aside from small-scale clearance of vegetation to form
the trail network to complete cat eradication, no negative
impacts on plants were observed but a greater impact on
some plant species may have resulted from the release of
rats from cat predation.
Potential and measured benefits to native biodiversity
Seabirds
Cooks petrel breeding success in high altitude
habitats (with 90% of the population), averaged 5%
prior to rat eradication but increased to approximately
Fig. 3 Catch per unit effort for all skink species combined on
Hauturu between 2002 and 2013 (Source: Department of
Conservation, Warkworth, New Zealand).
Griffiths, et al.: Costs and benefits for biodiversity on Little Barrier Island
60% the following breeding season as a result of reduced
predation pressure (Rayner, et al., 2007a; Rayner, et al.,
2007b). Improved Cook’s petrel breeding success was
also circumstantially reflected in a tenfold increase in the
number of recently fledged chicks presented at bird rescue
centres on the adjoining mainland the season following
eradication and ongoing (M.J. Rayner, unpubl. data). An
obvious massive increase in the extent and volume of
nocturnal vocalisation by Cook’s petrels has been observed
over the last 12 years suggesting ongoing population
recovery facilitated by increased breeding success and
recruitment (M.J. Rayner, pers. obs.).
Up to 600 pairs of black petrel are now thought to
breed on Hauturu (Bell, et al., 2016) up from the 50–100
pairs estimated by Imber (1987) prior to cat eradication.
Similarly, breeding success increased from 1977 (50%),
1978 (60%) and 1996 (71.8%) (Imber, unpubl. data)
to 2015/16 (85%). Before 1980 up to 67% of fledglings
emerging from burrows were killed by cats and fewer than
5% of chicks were expected to have fledged (Imber, 1987).
Between 1 and 28% of adult black petrels were also killed
by cats at the colony between 1972 and 1976 (Imber, 1987).
Comparisons of breeding activity within the same burrows
during 1996/97 and 2015/16 showed a stable occupation
rate of ca. 57% over the 19-year period and ca. 3% decline
in breeding activity (Bell, et al., 2016).
New Zealand storm petrels (Fregetta maoriana),
thought to be extinct for 110 years, were rediscovered at
sea in 2003 and, after much effort, were located breeding
on Hauturu in 2013 (Stephenson, et al., 2008; Rayner, et
al., 2015). Mark recapture data collected between 2015
and 2017 suggest a minimum population size of 1,000
individuals (M.J. Rayner unpubl. data) and, based on at sea
sightings, the population is steadily increasing.
Grey-faced petrels were discovered breeding after
an apparent 60-year absence in 2009 and anecdotal
observations of old colony sites suggest a gradual increase
in these populations (M.J. Rayner, unpubl. data; Rayner,
et al., 2009). The calls of other seabird species, such as
common diving petrels (Pelecanoides urinatrix) and
fluttering shearwaters (Puffinus gavia), have been also
documented subsequent to rat eradication (M.J. Rayner
unpubl. data) and may reflect recolonisation of the island’s
coastline by these predator-sensitive species.
Terrestrial birds
Three bird species have been introduced or reintroduced
to Hauturu since the removal of cats: kakapo (Strigops
habroptilus) during 1982, kokako (Callaeus wilsoni)
during 1980–1988 and tieke during 1984–1988. Following
their reintroduction, both tieke and kokako populations
expanded rapidly and are now abundant across the island
(K. Parker & I. Flux, pers. comm.). Kakapo were removed
from the island in 1998 due to ongoing nest predation
by rats but were re-established in 2012. Breeding by
some individuals has subsequently been documented but
whether the population will ever become self-supporting is
at present unknown (L. Joyce, unpubl. data).
Annual distance sampling completed between 2005
and 2013 in the south-west of the island initially charted
a decrease in numbers of hihi (Notiomystis cincta) and
tui (Prosthemadera novaeseelandiae) (Toy, et al., 2018).
Hihi numbers appeared to stabilise from 2009 onwards but
the density of tui continued to vary. The recorded density
of tieke changed little over the survey period. Forest bird
counts undertaken between 2013 and 2017 within the same
area recorded significantly higher numbers of bellbird,
tomtit (Petroica macrocephala), parakeets, robin, kokako
and tieke and a decline in numbers of whitehead (Mohoua
albicilla), tui, hihi, rifleman (Acanthisitta chloris),
grey warbler (Gerygone igata), blackbird and silvereye
(Zosterops lateralis) when compared to counts undertaken
before and during the cat eradication and prior to the rat
eradication (C.R. Veitch, unpubl. data). No significant
change was detected in the overall number of forest birds
(C.R. Veitch, unpubl. data).
No significant change in calling frequency was detected
in kiwi call count surveys over the period 1993 to 2014
although frequencies recorded were consistently higher
than sites monitored on the North Island of New Zealand
(Wade, 2014a). Despite the return of the brown teal (Anas
chlorotis) that were removed during the rat eradication and
the introduction of additional individuals, the brown teal
population has not expanded. The brown teal population is
considered permanent, but numbers present may be more a
reflection of the species’ breeding success on nearby Great
Barrier Island (Aotea).
Banded rail (Gallirallus philippensis), last seen on
the island in 1946 (Sibson, 1947), have returned to the
island and reared young (C.R. Veitch, unpubl. data) and
spotless crake (Porzana tabuensis), never previously
recorded on the island, are now present and breeding (C.R.
Veitch, unpubl. data). Another short-term impact worthy
of note is the appearance and establishment of bellbirds
at Tawharanui Regional Park subsequent to rat eradication
(Brunton, et al., 2008). Invasive vertebrates were removed
from Tawharanui at the same time as the rat eradication on
Hauturu and this coupled with an increase in the number
of bellbirds (as indicated by mist-netting data) may have
created conditions suitable for dispersal and subsequent
population establishment.
Reptiles
Following rat removal, numbers of reptiles caught in
pitfall traps steadily increased (Fig. 3). Towns (Oligosoma
townsi), moko (O. moko) and shore skink (O. smithi)
showed the biggest increase (see Fig. 4), contributing to
an 18-fold increase in the total number of skinks caught
per 100 trap nights since the rat eradication (Brown,
2013). Although numbers are too low to quantify changes
to the island’s chevron skink (Oligosoma homalonotum)
population, the number of additional skinks found after
rat eradication may indicate population recovery. Prior to
the rat eradication only one chevron skink had ever been
found on the island. Four have been found since rats were
removed.
Limited monitoring of the island’s gecko populations
was undertaken, but spotlight surveys completed in
2009 and 2013 suggest populations are recovering
Fig. 4 Catch per unit effort for five species of skinks on
Hauturu between 2002 and 2013 (Source: Department
of Conservation, Warkworth, New Zealand).
563
Island invasives: scaling up to meet the challenge. Ch 3C Strategy: Outcomes
from pre-eradication declines. Sighting rates of Pacific
(Dactylocnemis pacificus), forest (Mokopirirakau
granulatus) and common (Woodworthia maculatus) gecko
in 2013 more than doubled relative to 2009 (Brown, 2013)
and Duvaucel’s gecko (Hoplodactylus duvaucelli) have
been sighted more frequently (Hoare, 2009).
Tuatara (Sphenodon punctatus) were thought to be
extinct on Hauturu until the species was rediscovered
in 1991–1992 (Whitaker & Daugherty, 1991). Nine
adults were taken into captivity on the island to ensure
the relict population did not go extinct before rats were
removed (Moore, et al., 2008). Since the rat eradication
was confirmed successful in 2006 more than 196 young
tuatara, raised in captivity, have been released at three sites
and all captive adult tuatara have been returned to the wild.
Additional adult survivors have been detected, breeding
in the wild population has been noted and the population
appears to be expanding (S. Keall, pers. comm.).
Aquatic invertebrates
A survey of aquatic invertebrates completed in
1963 was repeated in 2014 to identify changes in faunal
composition. In total, 33 macroinvertebrate taxa from 12
orders were recorded from benthic samples. Six species
of mayfly (Mauiulus luma, Isothraulus abditus, Zephlebia
spectabilis, Arachnocolus phillipsi, Ichthybotus hudsoni
and Neozephlebia scita), and two species of caddisfly
(Trichoptera) (Oxyethira albiceps and a Chathamiidae
sp.) not recorded in the 1963 survey, were found in 2014
(Wade, 2014b).
Terrestrial invertebrates
An annual monitoring programme to assess the
recovery of wetapunga (Deinacrida heteracantha), New
Zealand’s largest giant weta, was instigated in 2005, a year
after rats were removed. Numbers of wetapunga found in
each survey had more than doubled by 2009 (Green, et
al., 2011). Results indicate that the numbers increased by
50% every second year. Subsequent captive breeding, for
translocation to other islands, showed that wetapunga have
a two to three-year life cycle (P. Barrett, pers. comm.),
potentially explaining the stepped rate of increase on
Hauturu. During the monitoring programme, occupancy
of daytime refuge sites remained low, suggesting the
population may increase further over time. A repeat of
the programme would be required to verify the level of
increase and thus give a longer-term measure of the benefit
of the rat eradication.
In 2017, surveys throughout New Zealand for the
endemic forest ringlet butterfly (Dodonidia helmsii)
revealed the species’ presence on Hauturu. This species
was widespread throughout much of the country but is now
rare or absent from many areas of its previous distribution
(S. Wheatley, pers. comm.). Despite the presence of
suitable habitat, the forest ringlet had not previously been
recorded on the island. Multiple individuals were found,
indicating a resident population (L. Wade, unpubl. data; J.
Knight, pers. comm.). Gibbs (1980) highlights the potential
for introduced social wasps, the German wasp (Vespula
germanica) and European common wasp (V. vulgaris),
as a cause for the decline in forest ringlet populations.
Interestingly, the European common wasp was noted
by previous island rangers as a significant nuisance on
Hauturu (C. Smuts-Kennedy, pers. comm.) but subsequent
to the rat eradication social wasps have not been reported.
The relationship between rat eradication and social wasp
populations is currently the subject of a PhD study at the
University of Auckland (J. Schmack, pers. comm.).
564
Plants
Nineteen of 34 plant species monitored on fixed plots
had more than 20 seedlings and were analysed further
(Campbell, 2011). Significantly more seedlings were
found for 14 species following rat eradication, Pisonia
brunoniana, Coprosma macrocarpa, Ixerba brexioides,
Knightia excelsa, Rhopalostylis sapida, Phyllocladus
trichomanoides, Nestegis lanceolata, Dacrycarpus
dacrydioides, Ripogonum scandens, Hedycarya arborea,
Dysoxylum
spectabile,
Pittosporum
umbellatum,
Macropiper excelsum and Corynocarpus laevigatus (Fig.
5). Seedlings of 11 others were searched for in 2008 and
2009. In 2009 Coprosma arborea seedlings were very
abundant. Fewer seedlings were counted of Agathis
australis, Beilschmiedia tarairi, B. tawa, Prumnopitys
ferruginea and Vitex lucens. A few species that Pacific
rats severely affect (e.g. Coprosma repens, Elaeocarpus
dentatus, Melicytus novae-zelandiae, Pouteria costata),
showed little early response because of their initial rarity
(Campbell & Atkinson 2002; Campbell, 2011).
Prior to rat eradication, seedlings of N. lanceolata, R.
sapida and R. scandens were rare, but in 2008, N. lanceolata
was found on most plots, and R. sapida and R. scandens
seedlings were common in moister sites. The number
of seedlings of other tree species had also significantly
increased. Seedlings of B. tarairi, C. laevigatus and
P. trichomanoides were twice as numerous in Kunzea
ericoides stands after rat eradication, D. spectabile was
five times more common and R. scandens 41 times.
Threatened plants also showed a positive response to
the removal of rats. Improved seed set by the endangered
Carmichaelia williamsii was noted and the endangered
Euphorbia glauca colonised new areas. Seed production
Fig. 5 Log ratios of seedling numbers on Hauturu (after
eradication) and control islands with rats, with 95%
confidence intervals. Log ratios of < 0, 0, > 0 indicate
seedling counts that decreased, remained the same or
increased, respectively.
Griffiths, et al.: Costs and benefits for biodiversity on Little Barrier Island
in the endangered Dactylanthus taylorii increased and
individuals have been discovered at new locations on the
island including a site where seeds were hand sown (D.
Havell, pers. comm.). No increase in seedling recruitment
was noted for the invasive plant species Asparagus
scandens, Cortaderia jubata, C. selloana and Araujia
hortorum (managed as part of an ongoing eradication/
control programme) following rat and cat eradication.
DISCUSSION
The legislation that defines the management of
Nature Reserves such as Hauturu, the Reserves Act 1977,
mandates the removal of exotic species to protect native
ecosystems. Cat and rat eradication on Hauturu became a
matter of urgency because of their impacts on individual
species. For cats, these impacts included extirpation of
tieke and grey faced petrels as well as threats to other
seabird species. The effects of rats were much wider and
included impacts on seabirds, tuatara, lizards, invertebrates
and an array of plants (Griffiths, 2002). Cats, rats or the
two in combination were probably also responsible for
the extinction of the last population of North Island snipe
(Tennyson & Martinson, 2006).
Short term negative impacts from the cat eradication
operation on Hauturu were minor and limited to the mortality
of some non-target bird species caught in traps, apparent
declines in skink populations and the removal of relatively
small amounts of vegetation as a consequence of track
and hut construction. An unanticipated outcome of the cat
eradication, was a reduction in breeding success of Cook’s
petrel nesting at higher elevations. This was attributed
to mesopredator release leading to increased predation
pressure by a rat population no longer suppressed by cats
(Rayner, et al., 2007b). This mechanism may also explain
why pitfall trapping charted a decline in skink capture rates
between cat and rat removal. Increased pressure on other
rat foods (invertebrates, seeds and seedlings) may also
have been sustained, but was not monitored.
As predicted in the AEE for rat eradication (Griffiths,
2002), negative impacts of the application of rodent bait
were short-lived and minor and included no more than
the loss of some individuals of at least eight bird species.
Monitoring could not detect changes in abundance for
these species indicating that they were not affected at the
population level. The only native species not detected
subsequent to the rat eradication was the red-finned bully.
The absence of this species in the survey completed 10 years
after the rat eradication could have been a consequence of
the application of rodent bait for rats but equally the species
could have been extirpated by a storm event or simply
that insufficient search effort has been undertaken. This
species is diadromous and likely to recolonise or could be
reintroduced provided that suitable habitat on Hauturu is
still available.
In contrast, the benefits of cat and rat eradication
have been significant, and all species deemed vulnerable
to extinction have since recovered. Tieke and kokako
were successfully established following cat removal. Rat
eradication resulted in recolonisation of the island by
grey faced petrels and immediate recovery of the island’s
Cook’s petrel population. As predicted by Griffiths (2002),
there were increases in the abundance of skinks, geckos
and invertebrates such as the wetapunga and in seedling
recruitment by numerous tree species. All but one of the
species identified as likely to benefit from rat eradication
have shown evidence of recovery. The exception is the
reintroduced tieke which could have reached carrying
capacity ahead of the rat eradication.
Some species previously not recorded from Hauturu
have made remarkable appearances. Examples include the
New Zealand storm petrel, forest ringlet butterfly and eight
new aquatic invertebrates. All such species were likely
present in refugia but undetectable until rats were removed.
It is unlikely that these will be the last discoveries to be
made on the island. Several seabird species are expected
to recolonise and highly cryptic species are still likely
waiting discovery or rediscovery on Hauturu in the future.
For example, only one record of striped skink has been
made on the island, but this cryptic species will likely be
found again in the future. As noteworthy as unexpected
appearances is the disappearance of German and common
wasps following rat eradication. The disappearance of
these two highly invasive species may have enabled the
recovery of the forest ringlet.
Other ecological changes post rat and cat eradication
have been unclear. Predicted increases in the abundance of
forest birds after cat and rat removal have yet to eventuate.
Some species undoubtedly benefited following the removal
of cats, but monitoring methods were insufficiently
sensitive to detect population changes; bird counts were
too variable to discern significant trends (Girardet, et al.,
2001). It is also likely that some of the potential benefits
of cat removal were confounded by the presence of
rats. Attempts to manage kakapo on the island post cat
eradication, for example, were thwarted by the continued
presence of rats that preyed upon eggs and chicks.
After the removal of rats, initial positive trends for
species such as hihi and tui were followed by declines
to pre-eradication levels. Whether, this was a result of
insufficient sampling effort, inadequacy of the methods
used, changes in inter-specific competition among birds,
or simply that the removal of cats and rats had little effect
is not well understood. Forest birds were also monitored
in the most accessible part of the island, the island’s SW
corner, which had been subject to the greatest impacts
of logging and grazing. Consequently, ongoing forest
successional changes may confound monitoring results.
The most recent set of bird counts completed in 2017
suggest that some forest bird species have increased in
abundance whereas others have declined (Veitch unpubl.
data). Further monitoring is needed to confirm these trends.
The effects of cat and rat removal on black petrel is also
less straightforward. Although petrel numbers appear
stable and breeding success has improved, the influence
of other factors such as birds fledged from Hauturu being
lured away to the much larger and noisier colony on Great
Barrier Island (E. Bell, unpubl. data) may be affecting
population recruitment.
As evidenced by altered patterns of seedling
recruitment following rat eradication, changes in forest
composition will occur. Future changes within the island’s
plant, invertebrate and reptile communities are likely to
be strongly influenced by the recovery of Cook’s petrel
and the return of other seabirds. The enormous influence
of seabirds on ecological communities has been well
described (e.g. Jones, 2010; Smith, et al., 2011). However,
given the large size of Hauturu and the extent of its forest
communities, the full impact of these ‘ecosystem engineers’
is at present unknown.
On a global scale, there are no comparative invasive
mammal eradications that have been completed in such
a complex environment. The value of these eradications
thus derives not only from the responses of resident species
and recolonisation of those lost, but also in increasing
our understanding of the ways invasive species influence
island community structure. The changes reported here
have been documented over only 13 years, but the removal
565
Island invasives: scaling up to meet the challenge. Ch 3C Strategy: Outcomes
of cats and rats has instigated a process of recovery that for
many species on Hauturu will not be realised for decades.
For those species reduced to relict populations and with
low reproductive output, the post-eradication response
will be slow. Other species such as snipe and some large
species of lizards have been lost and will not contribute to
Hauturu’s ecosystems without intervention to re-establish
them. The timescales involved for the recovery process
are daunting. For example, the release of the near extinct
tuatara population has begun a process of recovery for this
species that may require centuries to play out.
From a social perspective, the islands of Auckland’s
Hauraki Gulf have been a source of inspiration for many
members of the public. The removal of rats and subsequent
reforestation of Tiritiri Matangi Island inspired volunteers
to invest tens of thousands of hours to plant trees (Galbraith
& Cooper, 2013). Hauturu has been no exception with its
own Community Trust formed in 1997. The Little Barrier
Island/Hauturu Supporters Trust provides on average
$100,000 NZD annually to conservation programmes on
the island. Visitors to Hauturu often return awe struck (R.
Griffiths, pers. obs.) from their first view of ‘primeval’
New Zealand.
It is now important that the story of Hauturu, the impact
that invasive species had on the island, and the recovery
witnessed subsequent to cat and rat removal is shared on
the world stage. It is only through the telling of such stories
that the public’s imagination will be captured along with
the attention of government and private agencies. This
and other projects have served to cement a sense of pride
in New Zealand’s biodiversity and have culminated in a
pledge by New Zealand’s Government for the country to be
free of introduced predators by 2050 (Parkes, et al., 2017).
ACKNOWLEDGEMENTS
We would like to dedicate this paper to the many
organisations and individuals that contributed to the
protection of Hauturu. The legacy that stands on the
horizon today serves as a fitting tribute to your efforts.
Thank you also to the many individuals that contributed to
the development of this paper, in particular David Wilson,
David Havell and Alicia Warren.
REFERENCES
Anon. (1996). ‘Kakapo recovery plan 1996–2005’. Threatened Species
Recovery Plan No. 21. Wellington, NZ: Department of Conservation.
Bell, E.A., Mischler, C.P., MacArthur, N. and Sim, J.L. (2016). Black
Petrel (Procellaria parkinsoni) Population Study on Hauturu-o-Toi/
Little Barrier Island, 2015/16. Report to the Conservation Services
Programme. Wellington, NZ: Department of Conservation.
Brown, N. (2013). Little Barrier Island Annual Reptile Monitoring
Programme Jan/Feb 2013 Report. Warkworth, NZ: Department of
Conservation.
Brunton, D.H., Evans, B.A. and Ji, W. (2008). ‘Assessing natural dispersal
of New Zealand bellbirds using song type and song playbacks’. New
Zealand Journal of Ecology: 147–154.
Campbell, D. and Atkinson, I. (2002). ‘Depression of tree recruitment
by the Pacific rat (Rattus exulans Peale) on New Zealand's northern
offshore islands’. Biological Conservation 107: 19–35.
Campbell, J. (2011). ‘Seedling recovery on Hauturu/Little Barrier Island,
after eradication of Pacific rats Rattus exulans’. DOC Research &
Development Series 325.
Capizzi, D., Sposimo, P., Sozio, G., Gotti, C., Baccetti, N. and Patrassi,
F. (2019). ‘Black rat eradication on Italian Islands: planning by looking
backward’. In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and
C.J. West (eds.) Island invasives: scaling up to meet the challenge, pp.
15–20. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Cree, A. (2014). Tuatara Biology and Conservation of a Venerable
Survivor. Christchurch, NZ: Canterbury University Press.
566
Daniel, M. and Williams, G. (1984). ‘A survey of the distribution,
seasonal activity and roost sites of New Zealand bats’. New Zealand
Journal of Ecology 7: 9–25.
DIISE (2017). ‘The Database of Island Invasive Species Eradications,
developed by Island Conservation, Coastal Conservation Action
Laboratory UCSC, IUCN SSC Invasive Species Specialist Group’.
University of Auckland and Landcare Research New Zealand. <http://
diise.islandconservation.org>. Accessed 30 September 2017.
Eckroyd, C.E. (1995). ‘Dactylanthus taylorii recovery plan’. Threatened
Species Recovery Plan 16. Wellington, NZ: Department of Conservation.
Elliott, G.P., Merton, D.V. and Jansen, P.W. (2001). ‘Intensive management
of a critically endangered species: the kakapo’. Biological Conservation
99(1): 121–133.
Fisher, P., Griffiths, R., Speedy, C. and Broome, K. (2011). ‘Environmental
monitoring for brodifacoum residues after aerial application of baits
for rodent eradication. In: C.R. Veitch, M.N. Clout and D.R. Towns
(eds.) Island invasives: eradication and management, pp. 300–304.
Occassional Paper SSC no. 42. Gland, Switzerland: IUCN and
Auckland, New Zealand: CBB.
Galbraith, M. and Cooper, H. (2013). ‘Tiritiri Matangi – an overview of
25 years of ecological restoration’. New Zealand Journal of Ecology
37(3): 258−260.
Gaskin, C. and Rayner, M.J. (2013). Seabirds of the Hauraki Gulf: Natural
History, Research and Conservation. Auckland, NZ: Auckland Council.
Gaze, P.D. (2001). ‘Tuatara Recovery Plan 2001−2011’. Threatened
Species Recovery Plan 47. Wellington, NZ: Department of Conservation.
Gibbs, G.W. (1980). New Zealand Butterflies: Identification and Natural
History. Auckland, NZ: Collins.
Girardet, S., Veitch, C. and Craig, J. (2001). ‘Bird and rat numbers on
Little Barrier Island, New Zealand, over the period of cat eradication
1976–80’. New Zealand Journal of Zoology 28(1): 13–29.
Green, C.J., Gibbs, G.W., Barrett, P.A. and City, M. (2011). ‘Wetapunga
(Deinacrida heteracantha) population changes following Pacific rat
(Rattus exulans) eradication on Little Barrier Island’. In: C.R. Veitch,
M.N Clout and D.R. Towns, (eds.). Island invasives: eradication and
management, pp. 305–308. Occassional Paper SSC no. 42. Gland,
Switzerland: IUCN and Auckland, New Zealand: CBB.
Greene, T.C. and Dilks, P.J. (2004). Effects of a Non-toxic Bait Application
on Birds: Assessing the Impacts of a Proposed Kiore Eradication
Programme on Little Barrier Island. Wellington, NZ: Department of
Conservation.
Griffiths, R. (2002). Assessment of Environmental Effects for the
Eradication of Kiore (Rattus exulans) from Little Barrier Island
(Hauturu). Warkworth, NZ: Department of Conservation.
Griffiths, R. (2004). Little Barrier Island Kiore Eradication – Post
Operational Report. Warkworth, NZ: Department of Conservation.
Griffiths, R. (2006). Hauturu Kiore Eradication – Post Eradication
Target Species Monitoring and Assessment of Operational Success.
Warkworth, NZ: Department of Conservation.
Griffiths, R., Buchanan, F., Broome, K. and Butland, B. (2012). ‘Rangitoto
and Motutapu – a Starting Point for Future Vertebrate Pest Eradications
on Inhabited Islands’. In: Timms R.M. (ed) Proceedings of the 25th
Vertebrate Pest Conference, pp. 22–27. Monterey, CA, USA.
Hoare, J.M. (2009). Hauturu (Little Barrier Island) Reptile Survey
– 23 February to 3 March 2009. Warkworth, NZ: Department of
Conservation.
Hoare, J.M., Pledger, S., Nelson, N.J. and Daugherty, C.H. (2007).
‘Avoiding aliens: behavioural plasticity in habitat use enables
large, nocturnal geckos to survive Pacific rat invasions’. Biological
Conservation 136(4): 510–519.
Hoosen, S. and Jamieson, I.G. (2003). ‘The distribution and current
status of New Zealand saddleback Philesturnus carunculatus’. Bird
Conservation International 13: 79–95.
Howald, G., Donlan, C.J., Faulkner, K.R., Ortega, S., Gellerman, H.,
Croll, D. and Tershy, B. (2010). ‘Eradication of black rats Rattus rattus
from Anacapa Island’. Oryx 44(1): 30–40.
Hutton, W. (1868). ‘Notes on the birds of Little Barrier Island’.
Transactions of the Royal New Zealand Institute 1: 160–162.
Imber, M.J. (1987). ‘Breeding ecology and conservation of the black
petrel (Procellaria parkinsoni)’. Notornis 34(1): 19–39.
Imber, M., West, J. and Cooper, W. (2003a). ‘Breeding biology of Cook’s
petrel (Pterodroma cookii) on Hauturu (Little Barrier Island) and
Whenua Hou (Codfish Island)’. Notornis 50: 221–230.
Griffiths, et al.: Costs and benefits for biodiversity on Little Barrier Island
Imber, M.J., McFadden, I., Bell, E.A. and Scofield, R.P. (2003b). ‘Postfledging migration, age of first return and recruitment, and results of
inter-colony translocation of black petrels (Procellaria parkinsoni)’.
Notornis 50(4): 183–190.
Towns, D.R. (1991). 'Response of lizard assemblages in the Mercury
Islands, New Zealand, to removal of an introduced rodent: the kiore
(Rattus exulans)’. Journal of the Royal Society of New Zealand 21:
119–136.
Innes, J. and Flux, I. (1999). ‘North Island kokako recovery plan
1999–2009’. Threatened Species Recovery Plan 30. Wellington, NZ:
Department of Conservation.
Towns, D.R. (1996). ‘Changes in habitat use by lizards on a New Zealand
island following removal of the introduced Pacific rat Rattus exulans'.
Pacific Conservation Biology (2): 286–292.
Ismar, S.M.H., Gaskin, C.P., Fitzgerald, N.B., Taylor, G.A., Tennyson,
A.J.D. and Rayner, M.J. (2015). ‘Evaluating on-land capture methods
for monitoring a recently rediscovered seabird, the New Zealand stormpetrel Fregetta maoriana’. Marine Ornithology 43: 255–258.
Towns, D.R., Parrish, R. and Board, N.W.T. (2003). Restoration of
the Principal Marotere Islands. Wellington, NZ: Department of
Conservation.
Jones, H.P. (2010). ‘Seabird islands take mere decades to recover
following rat eradication’. Ecological Applications 20(8): 2075–2080.
Jones, H.P., Holmes, N.D., Butchart, S.H., Tershy, B.R., Kappes, P.J.,
Corkery, I., Aguirre-Muñoz, A., Armstrong, D.P., Bonnaud, E. and
Burbidge, A.A. (2016). ‘Invasive mammal eradication on islands
results in substantial conservation gains’. Proceedings of the National
Academy of Sciences 113(15): 4033–4038.
Lindsay, J. and Moore, P. (1995). ‘Geological features of Little Barrier
Island, Hauraki Gulf’. Tane 35: 25–38.
Lloyd, B.D. and Powlesland, R.G. (1994). ‘The decline of kakapo
Strigops habroptilus and attempts at conservation by translocation’.
Biological Conservation 69: 75–85.
McGlynn, M., McDowall, B., Roberts, C., Petrov, I. and Marsh, S. (2000).
Vertebrate Stream Survey of Hauturu. Unpublished data. Department of
Conservation & National Institute for Water and Atmospheric Research.
Moore, J.A., Nelson, N.J., Keall, S.N. and Daugherty, C.H. (2008).
‘Implications of social dominance and multiple paternity for the genetic
diversity of a captive-bred reptile population (tuatara)’. Conservation
Genetics 9(5): 1243–1251.
Parkes, J.P., Byrom, A.E. and Edge, K.-A. (2017). ‘Eradicating mammals
on New Zealand island reserves: what is left to do?’. New Zealand
Journal of Ecology 41(2): 263–270.
Rayner, M.J., Clout, M.N., Stamp, R.K., Imber, M.J., Brunton, D.H.
and Hauber, M.E. (2007a). ‘Predictive habitat modelling improves
the population census accuracy of a burrowing seabird: a study of the
endangered Cook’s petrel’. Biological Conservation 138: 235–247.
Rayner, M.J., Hauber, M.E., Imber, M.J., Stamp, R.K. and Clout, M.N.
(2007b). ‘Spatial heterogeneity of mesopredator release within an
oceanic island system’. Proceedings of the National Academy of
Sciences of the United States of America 104(52): 20862–20865.
Rayner, M.J., Dunphy, B.J. and Landers, T.J. (2009). ‘Grey-faced petrel
(Pterodroma macroptera gouldi) breeding on Little Barrier Island, New
Zealand’. Notornis 56: 222–223.
Rayner, M.J., Gaskin, C.P., Fitzgerald, N.B., Baird, K.A., Berg, M.M.,
Boyle, D., Joyce, L., Landers, T.J., Loh, G.G. and Maturin, S. (2015).
‘Using miniaturized radiotelemetry to discover the breeding grounds
of the endangered New Zealand storm petrel Fregetta maoriana’. Ibis
157(4): 754–766.
Towns, D.R. and Atkinson, I.A.E. (2004). Restoration Plan for Korapuki
Island (Mercury Islands) 2004–2024. Wellington, NZ: Department of
Conservation.
Towns, D.R., Atkinson, I.A.E. and Daugherty, C.H. (2006). ‘Have
the harmful effects of introduced rats on islands been exaggerated?’.
Biological Invasions 8: 863–891.
Towns, D., Parrish, G., Tyrell, C., Ussher, G., Cree, A., Newman,
D., Whitaker, A. and Westbrooke, I. (2007). ‘Responses of tuatara
(Sphenodon punctatus) to removal of introduced Pacific rats from
islands’. Conservation Biology 21(4): 1021–1031.
Toy, R., Greene, T.C., Greene, B.S., Warren, A. and Griffiths, R. (2018).
‘Changes in density of hihi (Notiomystis cincta), tieke (Philesturnus
carunculatus) and tui (Prosthemadera novaeseelandiae) on Little
Barrier Island (Te Hauturu-o-Toi), Hauraki Gulf, Auckland’. New
Zealand Journal of Ecology 42(2): 149-157
Turbott, E. (1947). ‘Birds of Little Barrier Island’. New Zealand Bird
Notes 2(92): 93–108.
Veitch, C.R. (2001). ‘The eradication of feral cats (Felis catus) from Little
Barrier Island, New Zealand’. New Zealand Journal of Zoology 28(1):
1–12.
Wade, L. (2009). Hauturu (Little Barrier Island) Kiwi Survey July 2009.
Warkworth, NZ: Department of Conservation.
Wade, L. (2014a). Hauturu (Little Barrier Island) Comparative
Kiwi Survey May and July 2014. Warkworth, NZ: Department of
Conservation.
Wade, L. (2014b). A Survey of Aquatic Invertebrates and Fish in a
Selection of Intermittent Streams on Hauturu/Little Barrier Island.
Unpublished report. Whangarei, NZ: NorthTec, Applied and
Environmental Sciences.
Whitaker, A.H. and Daugherty, C.H. (1991). Research on the Tuatara
(Sphenodon punctatus) of Little Barrier Island, 5–12 February 1991.
Unpublished report. Auckland, NZ: Auckland Conservancy, New
Zealand Department of Conservation.
Winterbourn, M.J. (1964). ‘A survey of the stream fauna of Little Barrier
Island’. Tane 10: 58–69.
Young, D. (2004). Our Islands, Ourselves. Dunedin, NZ: University of
Otago Press.
Robertson, H.A., Colbourne, R.M. and Nieuland, F. (1993). ‘Survival
of little spotted kiwi and other forest birds exposed to brodifacoum rat
poison on Red Mercury Island’. Notornis (40): 253−262.
Sherley, G. (1998). ‘Threatened weta recovery plan’. Threatened Species
Recovery Plan No. 25. Wellington, NZ: Department of Conservation.
Sibson, R. (1947). 'A visit to Little Barrier Island'. New Zealand Birds
Notes 2: 134–144.
Smith, J., Mulder, C. and Ellis, J. (2011). ‘Seabirds as Ecosystem
Engineers: Nutrient Inputs and Physical Disturbance’. In: C.P.H.
Mulder, W.B. Anderson, D.R. Towns, and P.J. Bellingham (eds.)
Seabird Islands. Ecology, Invasion and Restoration, pp. 27–55. Oxford.
New York: Oxford University Press.
Stephenson, B.M., Gaskin, C.P., Griffiths, R., Jamieson, H., Baird, K.A.,
Palma, R.L. and Imber, M.J. (2008). ‘The New Zealand storm-petrel
(Pealeornis maoriana Mathews, 1932): First live capture and species
assessment of an enigmatic seabird’. Notornis 55(4): 191–206.
Tennyson, A.J.D. and Martinson, P. (2006). Extinct Birds of New Zealand.
Wellington, NZ: Te Papa Press.
Towns, D.R., Atkinson, I.A.E. and Daugherty, C.H. (1990). ‘The Potential
for Ecological Restoration in the Mercury Islands’. In: D.R. Towns,
C.H. Daugherty and I.A.E. Atkinson (eds.) Ecological restoration of
New Zealand Islands. Conservation Sciences Publication No. 2: 91–
108.
567
C.N. Kaiser-Bunbury
Kaiser-Bunbury, C.N. Restoring plant-pollinator communities: using a network approach to monitor pollination function
Restoring plant-pollinator communities: using a network approach to
monitor pollination function
C.N. Kaiser-Bunbury1,2
1
Centre for Ecology and Conservation, Biosciences, University of Exeter, Penryn Campus, Penryn, Cornwall,
TR10 9FE, UK. 2Ecological Networks, Department of Biology, TU Darmstadt, 64287 Darmstadt, Germany.
<C.Kaiser-Bunbury@exeter.ac.uk>.
Abstract Ecological restoration is a common tool to mitigate the loss of species and habitats, ultimately aiming to
restore ecosystem functioning. Large-scale experimental evidence is lacking, however, on whether standard management
techniques, e.g. the removal of invasive alien plants, indeed restore ecosystem functions at the community level. One
key ecosystem function is animal mediated pollination. Based on findings from an experimental network study on rocky
outcrops (inselbergs) on the island of Mahé in the Seychelles, I present recommendations for conservation practitioners
about how to incorporate a network approach into an evaluation of management effectiveness. Responses to restoration
actions by plant-pollinator communities and pollination functions lead to several conclusions regarding the resilience of
native fauna and flora and ecosystem functioning. Pollination network structure appears to be directly related to the quality
and resilience of pollination services, which suggests that network analysis can be used to monitor management efficacy.
I provide recommendations and advice to encourage the uptake of a network approach by conservation practitioners
seeking to restore ecosystem functions.
Keywords: biomonitoring, management effectiveness, pollination networks, Seychelles, vegetation restoration
INTRODUCTION
Despite recent efforts to slow biodiversity decline
worldwide, habitat degradation continues to degrade
and simplify ecosystems, especially in the species-rich
tropics (Butchart, et al., 2010). To mitigate the effects of
habitat modification on ecosystems and assist species and
ecosystem functions to adapt to changing environmental
conditions, conservation practitioners employ a diverse
set of management tools, including ecological restoration
(Sodhi & Ehrlich, 2010). Such management tools often
rely on a few well-studied target species to assess their
outcomes, primarily because of limited time and resources.
However, too little is known about the efficacy of restoration
for achieving self-sustaining species communities and
functioning ecosystems. Habitat restoration usually
modifies ecosystems with the purpose of providing suitable
habitat for target native species (Miller & Hobbs, 2007).
Non-target species, however, can serve essential functional
roles in the restored habitat and failure to recognise these
species and the ecosystem-level interactions and processes
that they are involved in may compromise restoration
efforts and assessment (Ehrenfeld, 2000). Pollination is
one such key ecosystem function; most tropical plants and
crops heavily rely on pollination services for reproduction
(Klein, et al., 2007; Ollerton, et al., 2011). Pollinators
are rarely targets of habitat restoration (Williams, 2011),
although this is slowly changing in agricultural areas
where the benefits of wild bees in crop pollination have
been considered (Kremen & M’Gonigle, 2015). Given that
ecosystems are characterised by networks of interactions
between organisms (McCann, 2007), the effect of habitat
restoration on pollination interactions is often best studied
with a network approach (Jordano, 1987; Proulx, et al.,
2005). Thus, to assess the impact of habitat restoration
on integrity of pollination services, an understanding
of the implications of structural changes of pollination
networks on functional performance is critical. Recent
work proposed close links between network structure
and ecosystem functioning (Coux, et al., 2016; Gómez, et
al., 2011; Schleuning, et al., 2015), but field experiments
at the community level are required to shed light on the
relationships between habitat restoration, pollination
network structure, the resilience of plant-pollinator
communities, and the quality of pollination services.
Restoration practitioners worldwide place vegetation
rehabilitation at the centre of habitat restoration, which
often involves removal of exotic plants and assisted
recovery of native plant communities (Clewell &
Aronson, 2013). Assistance takes the form of fencing
off native habitat against large herbivores or exotic
seed predators (see e.g. Florens & Baider, 2013), or the
reintroduction of large herbivores to replace now extinct
seed dispersers (Hansen, 2015). These interventions
enable native vegetation and their mutualists to establish
and adapt to subtle changes in native and novel processes,
which increase resilience against future disturbance. One
important prerequisite for a self-sustaining restored plant
community is a large and diverse native fruit crop, which
is dependent to some degree on the quality and quantity
of pollination services. To provide optimal functional
performance, plant-pollinator communities mutually rely
on diverse and reliable resources (pollen and nectar) and
services (pollination). Weighted network metrics, which
take into account the quantitative importance of species
for their mutualistic partners, have been developed to
assess the consequences of vegetation rehabilitation on
pollination services by teasing apart changes in abundance,
species diversity and the topology of species interactions,
e.g. species generalisations (Banašek-Richter, et al., 2004;
Blüthgen, et al., 2006; Tylianakis, et al., 2007).
RESTORING PLANT-POLLINATOR
COMMUNITIES
In a recent study, Kaiser-Bunbury, et al. (2017) showed
for the first time, with a large-scale field experiment, that
not only were species communities fundamentally changed
by restoration (the removal of invasive alien shrubs), but
also plant-pollinator interactions became more resilient
as a result of restoration. Restoration altered pollinator
behaviour and increased pollinator species richness
(Kaiser-Bunbury, et al., 2017). In this instance, the removal
of invasive plants modified pollinator foraging patterns,
which increased pollinator efficiency (i.e. more pollen
delivered per visit) and frequency (i.e. higher visitation rate
per flower) of native plants in the restored community (see
Fig. 3 in Kaiser-Bunbury, et al., 2017). Simultaneously,
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
568
up to meet the challenge, pp. 568–570. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Kaiser-Bunbury: Restoring plant-pollinator communities
pollinator species became more generalised in restored
communities, creating greater functional redundancy and
lower mutual dependencies. These results appeared at
first contradictory, as specialised pollinators tend to be
more effective pollinators than generalists, due to lower
interspecific pollen transfer (see Morales & Traveset, 2008
and references within). However, the data also suggested
that while pollinator species became more generalised as a
result of restoration, individual pollinators had increased
floral constancy, providing high quality pollination services
even at relatively low visitation frequencies (KaiserBunbury, et al., 2017). Several plant species at the restored
sites (nine species at restored vs. two species at unrestored
sites) further benefitted from attracting more pollinator
species – on average an increase in pollinator species
richness by approximately 114% compared to the same
plant species at the unrestored sites, thereby lowering their
dependency on a few pollinator species for reproduction.
The effects of restoration on the plant-pollinator
community and pollination services were reflected
by changes in pollination network structure (KaiserBunbury, et al., 2017). The findings on the connection
between network structure and ecological processes
are important for two reasons. Firstly, they corroborate
previous theoretical and empirical, non-experimental
work that suggested a direct relationship between
network properties and ecosystem functioning (Gómez,
et al., 2011; Schleuning, et al., 2015; Coux, et al., 2016).
Secondly, network metrics, which are commonly used to
characterise network properties, can now be employed to
inform scientists and practitioners about the ecological
and conservation status of communities and ecosystem
functions when, for example, compared to baseline data.
With future shifts in conservation approaches towards the
protection of ecosystem services and functions (Harvey, et
al., 2017), suitable tools and methods need to be developed
that allow conservation biologists and practitioners to
monitor and evaluate such processes. The Kaiser-Bunbury,
et al. (2017) study provided an important cornerstone for
interpreting processes in ecological communities by using
a network approach.
Network ecologists have advocated for some time
the potential of a network approach in applied ecology,
based on advances in understanding the processes that
shape community level interactions (e.g. Memmott, 2009;
Kaiser-Bunbury, et al., 2010; Tylianakis, et al., 2010). More
recently, a selection of network indicators, i.e. aggregate
network metrics describing community properties, was
proposed, which characterise the diversity and distribution
of interactions at the species, guild (e.g. plants, pollinators)
and network level (Kaiser-Bunbury & Blüthgen, 2015).
These network indicators were selected because of
ecological characteristics, sound empirical and theoretical
support, conceptual similarities to well-established
diversity indicators, and computational ease with which
they can be generated (Kaiser-Bunbury & Blüthgen, 2015).
The authors presented a conceptual framework on how to
use network indicators to guide conservation decisions
by evaluating management effectiveness, and proposed
island ecosystems as suitable model system. Island biotas
are not only in urgent need of extensive conservation
action, but the simplicity of island ecosystems also
facilitates comprehensive studies on interaction networks
(Kaiser-Bunbury, et al., 2010). Thus, how can the insights
gained from studies on network structure and ecosystem
functioning (e.g., Kaiser-Bunbury, et al., 2017) be applied
to biomonitoring and assessments of management
effectiveness by island conservation practitioners?
IMPLICATIONS FOR ECOLOGICAL
RESTORATION
Biotic interactions (here I refer to mutualistic interactions
such as pollination and seed dispersal, but antagonistic i.e.,
trophic, interactions may equally be used) can be shortlived and highly variable across seasons, years or even
longer time spans (Medan, et al., 2006; Olesen, et al., 2010;
CaraDonna, et al., 2017). Network indicators that describe
the ecological processes determining network structure
may be most suitable to monitor ecologically meaningful
changes in biotic interactions that reflect community-wide
adaptations to specific restoration actions, for example,
the removal of invasive species, reforestation with native
plants, or landscape modifications. Methodological and
ecological advances, however, are rarely used to their
full potential for evaluating and monitoring conservation
progress (Gardener, et al., 2010). To benefit from such
advances, network indicators could be used to inform
managers on whether conservation interventions actually
restore or maintain ecosystem integrity (Noss, 2004). In the
Seychelles, the positive effects of restoration on pollinator
communities and native plant reproduction were reflected
in corresponding changes in network indicators (KaiserBunbury & Blüthgen, 2015). These included the total
number of visits and interactions, interaction diversity and
evenness, and the degree of network- (H2ʹ) and specieslevel (dʹ) specialisation (Kaiser-Bunbury, et al., 2017).
Thus, recording community-wide biotic interactions
and calculating network indicators for observed biotic
interactions can provide restoration practitioners with a
measure of effectiveness for achieving the overall goal of
restoring ecosystem functioning.
A network approach may appear challenging, overly
complicated and costly to most conservation practitioners.
Instead of providing comprehensive instructions on how to
apply a network approach in restoration, I aim to illustrate
that using biotic interactions and network analyses
are viable and effective tools to monitor conservation
progress and adapt management approaches based on the
outcome of the performance assessment. Below I outline
four recommendations for consideration by practitioners
who are interested in embracing a network approach in
biodiversity conservation.
1) Clearly define conservation goals that can be
validated with network indicators. Network indicators can
only illustrate the properties of one specific ecosystem
function at a time, for example, pollination, seed dispersal,
or predation. It is therefore important to identify the
ecosystem function to be targeted by the conservation
intervention (Kaiser-Bunbury & Blüthgen, 2015). Decisionmaking tools that take into account multiple ecosystem
functions may be required to prioritise conservation action
(McCarthy & Possingham, 2007). Clear conservation
objectives and outcomes will then provide the basis for
selecting network indicators and setting threshold values of
conservation targets (Kaiser-Bunbury & Blüthgen, 2015).
2) Actively engage with applied network ecologists
who can assist with establishing data recording protocols
and conducting network analysis, possibly via electronic
data collection in the field and automated analysis
(Kaiser-Bunbury & Blüthgen, 2015). At first, the network
approach may appear dauntingly complex. However, the
involvement of network ecologists in the planning phase
of any conservation action will ensure that a suitable
sampling protocol is developed, facilitating data analysis
and interpretation to evaluate management effectiveness.
Network ecologists are also more likely to follow advances
in the field and can update protocols, sampling techniques
and analyses based on the most up-to-date research. In
return for the time invested, ecologists will have access to
569
Island invasives: scaling up to meet the challenge. Ch 3C Strategy: Outcomes
empirical data for publications and contribute actively to
maximising the impact of their research.
3) Be realistic in sampling design. Collecting data on
biotic interactions involving all species in the community
is often considered extremely time and labour intensive,
and therefore costly. It is not necessary, however, to
record ‘every single interaction’. Interaction networks are
inherently under-sampled (Vázquez, et al., 2009) but still
provide meaningful insights into ecosystem complexity
and functioning. It is more important to identify the
most time and cost efficient sampling method (see e.g.
Hegland, et al., 2010) and assess sampling completeness
with appropriate extrapolation techniques (Colwell &
Coddington, 1994). Depending on the conservation goals,
sampling of subsets or at a lower frequency/density may
suffice to reveal changes in network structure as a result of
the restoration intervention.
4) Select the most suitable sampling approach for your
habitat, available resources, and the accessibility of the
management site. For example, pollination interactions
can be observed using standardised transects, which is a
time-efficient sampling method most suited to meadows,
heathlands and other low-growing plant communities.
Alternatively, by observing target plants for a set amount
of time, pollination interactions can be recorded in a forest
or shrubland habitat with a 3-dimensional structure and
a patchy distribution of flowers (for a comparison of the
methods see Gibson, et al., 2011).
Why should conservation practitioners and ecologists
invest extra time and resources into monitoring processes?
In short, moving conservation actions towards an
ecosystem functions oriented approach (sensu Harvey, et
al., 2017) will require tools that can monitor and evaluate
the multi-facetted dimensions of biodiversity. The network
approach can generate detailed insights into the functioning
of ecological communities, is developing rapidly, and
presents a promising and exciting method for improving
biodiversity conservation in the 21st century.
ACKNOWLEDGEMENTS
I thank two reviewers for helpful comments on an
earlier version of this manuscript. I acknowledge funding
from the German Research Foundation (DFG; Heisenberg
Fellowship KA 3349/3-1).
REFERENCES
Banašek-Richter, C., Cattin, M.-F. and Bersier, L.-F. (2004). ‘Sampling
effects and the robustness of quantitative and qualitative food-web
descriptors’. Journal of Theoretical Biology 226(1): 23–32.
Blüthgen, N., Menzel, F. and Blüthgen, N. (2006). ‘Measuring
specialization in species interaction networks’. BMC Ecology 6: 9.
Butchart, S.H.M., Walpole, M., Collen, B., van Strien, A., Scharlemann,
J.P.W., Almond, R.E.A., Baillie, J.E.M., Bomhard, B., Brown, C.,
Bruno, J., Carpenter, K.E., Carr, G.M., Chanson, J., Chenery, A.M.,
Csirke, J., Davidson, N.C., Dentener, F., Foster, M., Galli, A., Galloway,
J.N., Genovesi, P., Gregory, R.D., Hockings, M., Kapos, V., Lamarque,
J.F., Leverington, F., Loh, J., McGeoch, M.A., McRae, L., Minasyan,
A., Morcillo, M.H., Oldfield, T.E.E., Pauly, D., Quader, S., Revenga,
C., Sauer, J.R., Skolnik, B., Spear, D., Stanwell-Smith, D., Stuart, S.N.,
Symes, A., Tierney, M., Tyrrell, T.D., Vie, J.C. and Watson, R. (2010).
‘Global biodiversity: Indicators of recent declines’. Science 328(5982):
1164–1168.
CaraDonna, P.J., Petry, W.K., Brennan, R.M., Cunningham, J.L.,
Bronstein, J.L., Waser, N.M. and Sanders, N.J. (2017). ‘Interaction
rewiring and the rapid turnover of plant–pollinator networks’. Ecology
Letters 20(3): 385–394.
Clewell, A.F. and Aronson, J. (2013). Ecological Restoration: Principles,
Values, and Structure of an Emerging Profession. Washington, D.C.,
USA: Island Press.
Colwell, R.K. and Coddington, J.A. (1994). ‘Estimating terrestrial
biodiversity through extrapolation’. Philosophical Transactions of
the Royal Society of London Series B-Biological Sciences 345(1311):
101–118.
Coux, C., Rader, R., Bartomeus, I., Tylianakis, J.M. and Mouillot, D.
(2016). ‘Linking species functional roles to their network roles’.
Ecology Letters 19(7): 762–770.
570
Ehrenfeld, J.G. (2000). ‘Defining the limits of restoration: The need for
realistic goals’. Restoration Ecology 8(1): 2–9.
Florens, F.B.V. and Baider, C. (2013). ‘Ecological restoration in a
developing island nation: How useful is the science?’. Restoration
Ecology 21(1): 1–5.
Gardener, M.R., Atkinson, R. and Rentería, J.L. (2010). ‘Eradications
and people: Lessons from the plant eradication program in Galapagos’.
Restoration Ecology 18(1): 20–29.
Gibson, R.H., Knott, B., Eberlein, T. and Memmott, J. (2011). ‘Sampling
method influences the structure of plant–pollinator networks’. Oikos
120(6): 822–831.
Gómez, J.M., Perfectti, F. and Jordano, P. (2011). ‘The functional
consequences of mutualistic network architecture’. PLOS ONE 6(1):
e16143.
Hansen, D.M. (2015). ‘Non-native megaherbivores: The case for novel
function to manage plant invasions on islands’. AoB Plants 7: plv085.
Harvey, E., Gounand, I., Ward, C.L. and Altermatt, F. (2017). ‘Bridging
ecology and conservation: From ecological networks to ecosystem
function’. Journal of Applied Ecology 54(2): 371–379.
Hegland, S.J., Dunne, J., Nielsen, A. and Memmott, J. (2010). ‘How to
monitor ecological communities cost-efficiently: The example of plantpollinator networks’. Biological Conservation 143(9): 2092–2101.
Jordano, P. (1987). ‘Patterns of mutualistic interactions on pollination
and seed dispersal: connectance, dependence, asymmetries and
coevolution’. The American Naturalist 129(5): 657–677.
Kaiser-Bunbury, C.N., Traveset, A. and Hansen, D.M. (2010).
‘Conservation and restoration of plant-animal mutualisms on oceanic
islands’. Perspectives in Plant Ecology Evolution and Systematics
12(2): 131–143.
Kaiser-Bunbury, C.N. and Blüthgen, N. (2015). ‘Integrating network
ecology with applied conservation: A synthesis and guide to
implementation’. AoB Plants 7: plv076.
Kaiser-Bunbury, C.N., Mougal, J., Whittington, A.E., Valentin, T.,
Gabriel, R., Olesen, J.M. and Blüthgen, N. (2017). ‘Ecosystem
restoration strengthens pollination network resilience and function’.
Nature 542(7640): 223–227.
Klein, A.-M., Vaissière, B.E., Cane, J.H., Steffan-Dewenter, I.,
Cunningham, S.A., Kremen, C. and Tscharntke, T. (2007). ‘Importance
of pollinators in changing landscapes for world crops’. Proceedings of
the Royal Society B: Biological Sciences 274(1608): 303–313.
Kremen, C. and M’Gonigle, L.K. (2015). ‘Small-scale restoration in
intensive agricultural landscapes supports more specialized and less
mobile pollinator species’. Journal of Applied Ecology 52(3): 602–610.
McCann, K. (2007). ‘Protecting biostructure’. Nature 446(7131): 29.
McCarthy, M.A. and Possingham, H.P. (2007). ‘Active adaptive
management for conservation’. Conservation Biology 21(4): 956–963.
Medan, D., Basilio, A.M., Devoto, M., Bartoloni, N.J., Torretta, J.P. and
Petanidou, T. (2006). ‘Measuring Generalization and Connectance
in Temperate, Year-long Active Systems’. In: N.M. Waser and J.
Ollerton (eds.) Plant-pollinator Interactions: From Specialization
to Generalization, pp. 245–259. Chicago: The University of Chicago
Press.
Memmott, J. (2009). ‘Food webs: A ladder for picking strawberries or a
practical tool for practical problems?’. Philosophical Transactions of
the Royal Society B: Biological Sciences 364(1524): 1693–1699.
Miller, J.R. and Hobbs, R.J. (2007). ‘Habitat restoration—do we know
what we’re doing?’. Restoration Ecology 15(3): 382–390.
Morales, C. and Traveset, A. (2008). ‘Interspecific pollen transfer:
Magnitude, prevalence and consequences for plant fitness’. Critical
Reviews in Plant Sciences 27: 221–238.
Noss, R.F. (2004). ‘Some suggestions for keeping national wildlife
refuges healthy and whole’. Natural Resources Journal 44: 1093–1111.
Olesen, J.M., Dupont, Y.L., O’Gorman, E.J., Ings, T.C., Layer, K., Melian,
C.J., Trojelsgaard, K., Pichler, D.E., Rasmussen, C. and Woodward, G.
(2010). ‘From Broadstone to Zackenberg: Space, time and hierarchies
in ecological networks’. Advances in Ecological Research: Ecological
Networks 42: 1–69.
Ollerton, J., Winfree, R. and Tarrant, S. (2011). ‘How many flowering
plants are pollinated by animals?’ Oikos 120(3): 321–326.
Proulx, S.R., Promislow, D.E.L. and Phillips, P.C. (2005). ‘Network
thinking in ecology and evolution’. Trends in Ecology and Evolution
20(6): 345–353.
Schleuning, M., Fründ, J. and García, D. (2015). ‘Predicting ecosystem
functions from biodiversity and mutualistic networks: An extension of
trait-based concepts to plant–animal interactions’. Ecography 38(4):
380–392.
Sodhi, N.S. and Ehrlich, P.R. (2010). Conservation Biology for All.
Oxford, New York: Oxford University Press.
Tylianakis, J.M., Tscharntke, T. and Lewis, O.T. (2007). ‘Habitat
modification alters the structure of tropical host-parasitoid food webs’.
Nature 445(7124): 202–205.
Tylianakis, J.M., Laliberté, E., Nielsen, A. and Bascompte, J. (2010).
‘Conservation of species interaction networks’. Biological Conservation
143(10): 2270–2279.
Vázquez, D.P., Blüthgen, N., Cagnolo, L. and Chacoff, N.P. (2009).
‘Uniting pattern and process in plant–animal mutualistic networks: A
review’. Annals of Botany 103(9): 1445–1457.
Williams, N.M. (2011). ‘Restoration of nontarget species: Bee
communities and pollination function in riparian forests’. Restoration
Ecology 19(4): 450–459.
L. Luna-Mendoza, A. Aguirre-Muñoz, J.C. Hernández-Montoya, M. Torres-Aguilar, J.S. García-Carreón, O. Puebla-Hernández, S. Luvianos-Colín, A.
Cárdenas-Tapia and F. Méndez-Sánchez. Ten years after feral goat eradication: the active restoration of plant communities on Guadalupe Island, Mexico
L. Luna-Mendoza, A. Aguirre-Muñoz, J.C. Hernández-Montoya, M. Torres-Aguilar, J.S. García-Carreón, O. Puebla-Hernández, S. Luvianos-Colín, A.
Cárdenas-Tapia and F. Méndez-Sánchez
Ten years after feral goat eradication: the active restoration of plant
communities on Guadalupe Island, Mexico
L. Luna-Mendoza1, A. Aguirre-Muñoz1, J.C. Hernández-Montoya1, M. Torres-Aguilar2, J.S. García-Carreón3,
O. Puebla-Hernández1, S. Luvianos-Colín1, A. Cárdenas-Tapia1 and F. Méndez-Sánchez1
Grupo de Ecología y Conservación de Islas, A.C. Moctezuma 836, Centro, Ensenada, Baja California, México 22800.
<luciana.luna@islas.org.mx>. 2Comisión Nacional de Áreas Naturales Protegidas. Av. del Puerto 375, Local 30,
Fracc. Playa Ensenada, Ensenada, Baja California, México 22880. 3Comisión Nacional Forestal. Periférico Poniente
5360, San Juan de Ocotán, Zapopan, Jalisco, México 45019.
1
Abstract As the first step towards the ecological restoration of its islands, Mexico has completed 60 eradications of
invasive mammals thanks to a strong partnership between Grupo de Ecología y Conservación de Islas, A.C. (GECI),
the federal government, local fishing communities, academia, and private donors. The removal of invasive mammals
has led to the dramatic recovery of the islands’ ecosystems. On Guadalupe Island, after completing the goat eradication
in 2007, the native vegetation started to recover. Plants considered extinct or extirpated have been rediscovered, and
plant species new to the island have been recorded. However, in order to achieve the island’s full recovery, the active
restoration of degraded soils and vegetation are needed. To date, GECI, in collaboration with the National Forestry
Commission (CONAFOR) and the National Commission for Natural Protected Areas (CONANP), is implementing a 700
ha project to accelerate the restoration of the native vegetation communities. The project involves reforestation, erosion
control, and fire prevention actions on different plant communities: forests and sage scrub. An on-site nursery has been
established, seedlings—mostly from endemic trees—are being grown, and on-site reforestation planting has started. Up
to June 2018, we have planted almost 40,000 trees, and will produce 160,000 seedlings during this year. Mechanical
methods to control and prevent erosion have been used as we have installed more than 2,400 m of contour barriers, 57 m3
of dams, and rehabilitated firebreaks. The actions will continue: the long-term goal being the comprehensive restoration
of the vegetation communities devastated by feral goats. The Guadalupe Island experience will be useful to inform the
restoration of other Mexican islands.
Keywords: erosion control, Guadalupe cypress, Guadalupe pine, reforestation, vegetation recovery
INTRODUCTION
Islands support a disproportionate amount of
biodiversity in relation to their area (Myers, et al., 2000),
but also are vulnerable ecosystems, highly susceptible to
any alteration to their fragile equilibria (Holdgate, 1967;
Simberloff, 1995; Whittaker & Fernández-Palacios, 2007).
The trophic webs of oceanic islands have been described
as “very simplified, with little ecological or taxonomical
redundancy” (Courchamp, et al., 2003). This is one of the
reasons why invasive alien species are considered one
of the biggest threats to insular ecosystems and the main
cause of insular biodiversity loss, as well as the alteration
of ecosystem functions (Reaser, et al., 2007; Veitch, et al.,
2011). Native insular species normally lack evolutionary
defences, having traits that evolved in the absence of
regular immigrants, and in consequence they fail to
adapt to new threats posed by invasive species (Brook,
et al., 2008; Berglund, et al., 2009). In contrast, invasive
species have attributes that facilitate their establishment
in novel environments due to their broad ecological niche
(generalists) and high degree of behavioural flexibility.
Consequently, they normally thrive when introduced to
new environments (Mack, et al., 2000; Courchamp, et al.,
2003; Sol, 2007). Among invasive mammals, feral goats
(Capra hircus) are one of the most destructive species.
Their effects include overgrazing, soil compaction, and
tree and shrub damage through browsing (Coblentz,
1978; Parkes, et al., 1996; Campbell & Donlan, 2005;
Chynoweth, et al., 2013).
Guadalupe Island is a priority site in terms of
biodiversity conservation. It is a Biosphere Reserve, as well
as an Important Bird Area (IBA; Vidal, et al., 2009) and an
Alliance for Zero Extinction site (AZE, 2010). In addition,
it is categorised as a Marine Priority Conservation Area by
the Commission for Environmental Cooperation of North
America (Morgan, et al., 2005), and it is included in the
Southern Californian Pacific Marine Ecoregion (Wilkinson,
et al., 2009). Unfortunately, invasive mammals – including
feral goats – were introduced in the 19th century with
devastating consequences for the island’s flora and fauna.
Goats depleted entire vegetation communities. Moran
(1996) stated that “…it is most important before plants
are lost, to remove all goats from the island, reversing the
process of degradation and encouraging in every way the
renewal of the natural vegetation. Even at best, some rare
plants may die out unless propagated and replanted...”.
Other authors agreed, claiming that conservation actions
for the island must begin by removing the feral goats and
be followed with a plan of active restoration (León de la
Luz, et al., 2003, Aguirre-Muñoz, et al., 2005). Also, for
the Mexican Government there was an understanding that
urgent restoration actions were needed. As a result, in
terms of ecological restoration, much has been done during
the past decade to tackle the threats posed to Guadalupe’s
biodiversity, particularly those from introduced species
(Aguirre-Muñoz, et al., 2011; Luna-Mendoza, et al.,
2007). Therefore, a long-term restoration and conservation
programme has been developed for the island, aimed at
removing invasive mammals to protect Guadalupe’s native
flora and fauna – especially seabirds – and preventing
more extinctions. The successful eradication of goats, in a
collaboration between GECI, federal government agencies
and private donors, was the beginning of the island’s
recovery. The next phase is to do active restoration, mostly
through reforestation of several vegetation communities.
SITE DESCRIPTION
Guadalupe Island is a 242 km2 remote oceanic island
located in the Pacific Ocean, 260 km off the Baja California
peninsula, Mexico (29o N, 118o 20'W). It represents
Mexico’s last frontier on its western and northern margins;
a unique territory in many ways, particularly in terms of
biodiversity, a “naturalists’ paradise” in the words of Dr
Edward Palmer after his 1875 visit to Guadalupe (Huey,
1925).
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 571–575. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
571
Island invasives: scaling up to meet the challenge. Ch 3C Strategy: Outcomes
Guadalupe is a 5.8 km high seamount that emerges
from a depth of 4.5 km, with a maximum elevation of
1.3 km above sea level (Delgado-Argote, et al., 1993). It
comprises a main island, three islets and several offshore
rocks. Guadalupe was discovered in 1602 by Spanish
explorer Juan Sebastían Vizcaíno (León Portilla, 1989). Yet,
it remained pristine and uninhabited until the beginning of
the 19th century when Russian, English and American fur
hunters visited the island in search of fur seals, sea otters
and elephant seals (Hanna, 1925; Huey, 1925).
Guadalupe is a protected area decreed as a Biosphere
Reserve by the Mexican government in 2005. The Reserve
is managed by the National Commission of Natural
Protected Areas (CONANP), and is safeguarded by the
Ministry of the Navy, which has been watching over this
important territory since the early 1900s. Besides the Navy
base on the southern tip of the island, there are two more
settlements: a settlement of the Abuloneros y Langosteros
fishing cooperative on the west coast, and a biological
field station of the Mexican NGO Grupo de Ecología
y Conservación de Islas, A.C. (GECI, for its Spanish
acronym) at about 1,200 m above sea level on the northwest portion of the island. Also, CONANP personnel are
present permanently on site. A total of 100 people inhabit
Guadalupe (CONANP, 2013). The only economic activity
on the island is commercial fishing, carried out solely by
the fishing cooperative to sustainably harvest valuable
marine resources such as abalone (Haliotis spp.) and
lobster (Panulirus interruptus) (Searcy-Bernal, et al.,
2010; Méndez-Sánchez, 2012).
Climate
Guadalupe has a Mediterranean climate, characterised
by hot, dry summers and cool, wet winters (Camps &
Ramos, 2012; Granda, et al., 2014). Temperature is
relatively stable throughout the year, with a mean of 17.2
± 2 ºC. Relative humidity oscillates between 69 ± 8% to
82 ± 5% without a well-defined seasonal pattern (Castro,
et al., 2005) and average annual cumulative precipitation
is 193 ± 119 mm (CONAGUA-SARH Delgadillo in
Moran, 1996; SARH-Colegio de Postgraduados, 2010).
However, given the islands’ complex topography, some
microclimates are also recognised. The south end of
Guadalupe is drier compared to the rest of the island, and
the humidity increases northwards with elevation, mostly
due to the fog influence. There are some records of ice
and snow in winter, restricted to the cypress forest at the
higher elevations (Moran, 1996; N. Silva-Estudillo, pers.
comm.). It is also likely that rainfall is relatively more
abundant at higher elevations (Moran, 1996). Guadalupe
Island’s local climate can also be influenced by regional
climatic conditions. Occasional tropical storms from the
south bring heavy rainfall to the island between summer
and autumn (August to October) (Moran, 1996). In
addition, the normal precipitation pattern can be disrupted
by irregular El Niño or La Niña events, associated with
supra- and subnormal precipitation, respectively, between
December and March (winter and spring) (Minnich, et
al., 2009). This oceanic island is heavily influenced by
the California Current, which generates a peculiar pattern
of wind, fog, and rainfall (León de la Luz, et al., 2003;
Garcillán, et al., 2012 ). Winds prevail from the northwest, while the island’s climate is influenced by a nearpermanent fog system which allows the presence of forests
on the island despite the low precipitation.
Flora
In a classification of Mexican Biogeographic Provinces,
Guadalupe Island is considered as a separate province (i.e.
Guadalupe Island province) within the Baja California
Province, which is part of the Nearctic Region (Morrone,
et al., 2002). This classification is based on distributional
572
patterns of plants, invertebrates and birds (Morrone, et
al., 1999). Floristically, it is very similar to the Channel
Islands, USA (Raven, 1965). Originally, the island was
home to a rich flora that included several insular and
Guadalupe endemics. In total, 225 vascular plant species
have been recorded on the island, 7% insular endemic
(shared with other islands of the region) and 12% endemic
to Guadalupe (Junak, et al., 2005; Rebman, et al., 2005;
GECI, unpublished data). In addition, 36 non-vascular
plants have been recorded for the site (Crum & Miller,
1956; Crum, 1972 in Moran, 1996) as well as 104 lichen
species (Weber, 1994 in Moran, 1996).
Several original vegetation communities have been
described, based on historic records (Moran, 1996;León de
la Luz, et al., 2005; Oberbauer, 2005). The communities
include forests, woodlands, chaparral (shrubs), native
grassland and communities dominated by low shrubs.
Some of the representative species of these vegetation
communities are the endemic Guadalupe cypress
(Cupressus guadalupensis), Monterey pine (Pinus radiata
var. binata) and Guadalupe palm (Brahea edulis). Several
native species such as the juniper (Juniperus californica now restricted to <10 individuals), and the shrubs island
redberry (Rhamnus pirifolia), and laurel sumac (Malosma
laurina) were also characteristic of some communities,
along with endemic succulents such as cistanthe (Cistanthe
guadalupensis) and liveforever (Dudleya guadalupensis),
the endemic Guadalupe senecio (Senecio pameri); and three
endemic species of shrubby tarweeds (Deinandra spp.).
Some insular endemics were also representative of these
environments, such as the island hazardia (Hazardia cana;
only present on San Clemente Island, USA and Guadalupe)
and the insular oak (Quercus tomentella; only present on
five of the Channel Islands, USA and Guadalupe).
INTRODUCTION OF INVASIVE MAMMALS
Guadalupe Island remained pristine and uninhabited
until the beginning of the 19th century when fur hunters
arrived (Hanna, 1925; Huey, 1925). It is likely that
house mice (Mus musculus) and feral cats (Felis catus)
were introduced following these first human settlements
(Hanna, 1925; Huey, 1925; Moran, 1996). Both species
were introduced around 1880, rapidly establishing feral
populations (Moran, 1996). In addition, whalers, in order
to have a source of fresh meat during their voyages,
introduced goats (Moran, 1996). Together, goats and cats
have been responsible for the extinction and extirpation of
many native and endemic species (e.g. Jehl Jr & Everett,
1985; León de la Luz, et al., 2003), and the impacts of
house mice remain to be evaluated. In addition, feral dogs
(Canis familiaris) also established a population on the
island (Moran, 1996). Other mammals, such as cows (Bos
taurus), were also introduced but never established feral
populations (Aguirre-Muñoz, et al., 2011; J. Rico-Cerda
pers. comm.).
Effects of feral goats on native flora
After goats were introduced, only one of the original
vegetation communities, comprising low shrubs present on
islets where goats or mice never were introduced, remained
pristine. The other plant communities either disappeared,
became restricted to a very patchy distribution, or were
represented only by isolated individual plants. At least
26 plant taxa became extinct or were extirpated due to
feral goats (Moran, 1996; León de la Luz, et al., 2003;
Oberbauer, 2005; GECI, unpublished data). Not only
were entire vegetation communities depleted, and many
endemic species lost, but also many non-native species
have been introduced. Since 1875, at least 69 plant species
have been introduced to the island, mostly European
grasses and forbs (Junak, et al., 2005; Rebman, et al., 2005;
GECI, unpublished data). The heavy modification of the
Luna-Mendoza, et al.: Guadalupe Island ten years after goat eradication
ecosystem caused by the feral goats, in combination with
the arrival of invasive plants, resulted in a vast extension
of bare ground and vegetation dominated by European
grasses (grassland community), such as slender wild oat
(Avena barbata) and red brome (Bromus rubens).
CURRENT RESTORATION AND CONSERVATION
ACTIONS
Ecosystem resilience – passive restoration
On Guadalupe Island, after completing the goat
eradication in 2007, native vegetation started to naturally
recover. Plants considered extinct or extirpated have been
rediscovered and there have been new plant records for
the island, including at least one undescribed species. A
survey conducted in 2001 on the endemic variety of the
Monterey pine—there are five endemic varieties: three in
the USA and two in Mexico, all the original seed source for
plantations around the world—estimated that there were
only 220 adult pines left (Rogers, et al., 2006). Since the
goats were eradicated, the number of new seedlings has
increased to several thousands (Fig. 1). Not only have the
trees have recovered, but shrubs are also returning with full
strength, competing well with invasive grasses. Ceanothus
arboreus, a shrub able to reach 6 or 7 m and a new record
for the island (Junak, et al., 2005), is now very common
around the cypress and pine-oak forests. Also, the maritime
desert scrub in the area most impacted by the goats’
presence has changed from almost 0% native vegetation
coverage (areas dominated by European grasses) to 52%
(Ceceña-Sánchez, 2014).
Active restoration of plant communities
In a review of passive vs active restoration effects on
forest recovery, Meli, et al. (2017) suggest observing the
system for a few years after intervention to inform better
decisions regarding active restoration actions. In the case
of Guadalupe Island, feral goat eradication was completed
in 2007. The active restoration project started in 2015.
Over a period of almost 10 years we documented and
measured the recovery of species. Not all recovered at the
same speed. Some trees, shrubs, and forbs are recovering
at a fast pace. However, there are many species that still
remain very fragile, given their low numbers (Juniperus
californica < 10 known individuals; insular oak, Quercus
tomentella < 50 adult trees; Cistanthe guadalupensis,
almost absent from the main island and surviving only
on islets), and there are others whose distribution has
decreased historically from forests to small isolated patches
(e.g. Cupressus guadalupensis).
In order to achieve the island’s full recovery, the
active restoration of vegetation and eroded and degraded
soils was the next conservation step. The negative effects
Fig. 1 Recovery of Guadalupe pine (Pinus radiata var.
binata) from 220 individuals (adult trees) to several
thousands in ten years. Photo credits: GECI Archive/J.A.
Soriano.
of overgrazing and soil compaction are exacerbated
on a volcanic island where soil (even some of the most
productive soil (Ugolini & Dahlgren, 2002)) is limited and
very susceptible to loss due to erosion. A study focused
on the cypress forest on Guadalupe Island concluded than
the erosion rates were exceptionally high, with a minimum
recorded loss of 44 ton/ha/year and the maximum 142 ton/
ha/year (Ramos Franco, 2007). Although it was estimated
only for the cypress forest, the erosion problem is evident
across the whole island, especially at higher elevations.
To date, GECI in collaboration with the National Forestry
Commission (CONAFOR) and CONANP, and other
partners, such as the Mexican Navy (SEMAR) and the
local fishermen’s cooperative Abuloneros y Langosteros, is
implementing a 700 ha project to accelerate the recovery of
native vegetation communities.
The project involves reforestation, erosion control, and
fire prevention actions for different plant communities.
Reforestation is being implemented over 583 ha: 33 ha of
palm forest; 120 ha of pine-oak forest; 261 ha of cypress
forest, 60 ha of juniper woodland and 109 ha of maritime
desert scrub. Erosion control actions (17 ha) are focused
only on the cypress forest, in an area with slopes of 27%,
loss of around 75% of the superficial soil layer, and deep
gullies (Ramos Franco 2007), which is considered as
extreme degradation (CONAFOR, 2004). On the other
hand, due to a fire which occurred in 2008 in the cypress
forest, the quantity of accumulated fuel was alarming,
around 110 t/ha on average, with a maximum of 1,000 t/
ha in certain areas (Luna-Mendoza et al. 2016). For this
reason, fire management actions were focused here. The
goal is to carry out fuel reduction (through manual removal
of surface fuels and increasing the height to live crown) in
100 ha and to restore 10 km of firebreaks.
An on-site nursery was built as part of the project
(Fig. 2). The nursery (480 m2) is surrounded by a mouseproof, galvanised steel fence of 50 m × 30 m, as mice
are responsible for the loss of huge amounts of seed and
seedlings at early stages. Around 15 species of native
and endemic species are being produced: Guadalupe pine,
Guadalupe cypress, Guadalupe palm, insular oak, island
hazardia, Guadalupe lupin (Lupinus niveus), Guadalupe
phacelia (Phacelia phyllomanica), island malva (Malva
occidentalis), Guadalupe rock daisy (Perityle incana),
among others. Species produced were chosen based on
their rarity on the island (e.g. Leptosyne gigantea and
Cistanthe guadalupensis); endangerment (e.g. juniper);
propagation material available; potential as nurse plant (e.g.
Sphaeralcea spp.); importance as food or shelter for native
invertebrates and landbirds (e.g. Senecio palmeri) and
effectiveness at retaining soil (e.g. Calystegia macrostegia
ssp. macrostegia). Their allocation was based on historic
information of former vegetation communities as well as
observations of where there has been natural recruitment.
Fig. 2 Plant nursery on Guadalupe Island with the capacity
to produce over 60,000 plants per year. Photo credits:
GECI Archive/J.A. Soriano.
573
Island invasives: scaling up to meet the challenge. Ch 3C Strategy: Outcomes
For example, goat removal allowed the recovery of the
island Ceanothus from seed surviving in the seed bank.
Now the species grows close to the cypress and pineoak forests, but individuals are sparse. However, for this
species, as seed is not capable of long distance dispersal by
itself (Minnich, 1982), it can take a long time to recover its
original coverage. Another species, the Guadalupe lupin,
is spreading at a fast pace, but being a legume with big
seed depends on rainfall to disperse seed downslope. On
Guadalupe the dispersal pathways are limited, as seedeating birds are few and native terrestrial mammals are
absent. In the case of the endemic cistanthe, although the
species is very common on the islets, on the main island
only three individuals have been recorded in the last 10
years. This is one of the lost species of the maritime desert
scrub, and a goal of this project is to reintroduce it to this
vegetation community.
Up to June 2018, 90,000 plants have been produced
in the nursery. The final goal is to produce 160,000
plants, mostly trees and shrubs. We have planted almost
40,000 plants, most of them trees. To date we have nearly
completed reforestation of the pine-oak and cypress areas
(Fig. 3). Some challenges have arisen: logistics linked to
working on an island; lack of plant propagation information
for some species (especially endemics); limited amount of
seed (insular-endemic or Guadalupe-endemic species and
very few individuals left); diseases; and limited amount of
water (relying mostly on the fog). In the last rainy season
there was virtually no rainfall and fog has been very
intermittent and scarce. We are therefore using resources
such as the commercial hydrophilic polymer based on
polyacrylamide, called Lluvia sólida®. We add 1.5 to 2
l of this hydrated polymer to each plant. So far, results
are encouraging, as survivorship of planted individuals
is above 85%. Regarding soil restoration, mechanical
methods to control and prevent erosion, such as check
dams and contour barriers, have been implemented. More
than 1,500 m of contour barriers of rocks and logs have
been built as well as 66 m3 of rock and log check dams. So
far, 27 hectares have been cleared of fuel and 1,500 m3 of
material has been removed.
WHAT’S NEXT?
There still much to be done on Guadalupe Island. A
period of 130 years of feral goats on the island caused
severe ecosystem degradation. However, with collaborative
projects, such as the one described here, we are heading in
the right direction to restore the island’s ecosystem services
and biodiversity. Future projects, conducted in collaboration
with CONANP, are to continue with fuel reduction actions
in the cypress and pine-oak forests, to collect seed of
endemic species to be stored at national seed banks and
to estimate carbon sequestration in forest ecosystems. As
a country, Mexico is now fully committed to the recovery
of its islands, going a step further than eradication actions.
On Socorro Island (Revillagigedo Archipelago; Pacific
Ocean), feral sheep were removed a few years ago. This
island is very similar to Guadalupe and faces the same
challenges: soil degradation and a need to do some active
restoration of the forest, not only for the vegetation itself
but to restore the habitat for many endemic land birds
which were close to extinction. In some cases we cannot
wait for the islands to recover naturally, especially where
other threats are still present (other invasive mammals).
Currently other islands in Mexico, such as Espíritu Santo
Island (Gulf of California), and María Cleofas Island (Las
Marías Archipelago, Pacific Ocean), are being cleared of
herbivores, and hopefully more active restoration actions
could be established on these sites in the near future.
ACKNOWLEDGEMENTS
We would like to thank government agencies, private
corporations and non-governmental associations that have
assisted with this project: Abuloneros y Langosteros,
Agroproductores de Zacualpan, Alianza WWF- Fundación
Carlos Slim, Colegio de Postgraduados-Campus Montecillo,
Comisión Nacional de Áreas Naturales Protegidas,
Comisión Nacional de Áreas Naturales ProtegidasReserva de la Biósfera Isla Guadalupe, Comisión Nacional
Forestal, Comisión Nacional para el Conocimiento y
Uso de la Biodiversidad, Instituto Nacional de Ecología
y Cambio Climático, Marisla Foundation, Secretaría de
Gobernación, Secretaría de Marina, Secretaría de Medio
Ambiente y Recursos Naturales-DGVS and DGGIMAR,
the David and Lucile Packard Foundation, The Nature
Conservancy, US National Park Service and Universidad
Autónoma Chapingo. Thanks also to the editors and
anonymous reviewers whose comments greatly improved
this manuscript.
REFERENCES
Fig. 3 Reforested areas on Guadalupe Island. The polygon
in the north relates to pine-oak forest and the three
polygons in the south are in the cypress forest area.
Shaded (grey) areas are the original areas proposed in
the project and striped areas are the ones completed (up
to June 2018).
574
Aguirre-Muñoz, A., Samaniego-Herrera, A., García-Gutiérrez, C., LunaMendoza, L., Rodríguez-Malagón, M. and Casillas-Figueroa, F. (2005).
‘El Control y la Erradicación de Fauna Introducida Como Instrumento
de Restauración Ambiental: Historia, Retos y Avances en México’.
In: O. Sánchez, E. Peters Recagno, R. Márquez-Huitzil, E. Vega, G.
Portales, M. Valdés and D. Azuara (eds.) Temas sobre Restauración
Ecológica, pp. 215–229. México, D.F.: Instituto Nacional de Ecología.
Aguirre-Muñoz, A., Samaniego-Herrera, A., Luna-Mendoza, L., OrtizAlcaraz, A., Rodríguez-Malagón, M., Méndez-Sánchez, F., FélixLizárraga, M., Hernández-Montoya, J.C., González-Gómez, R., TorresGarcía, F., Barredo-Barberena, J.M. and Latofski-Robles, M. (2011).
‘Island restoration in Mexico: ecological outcomes after systematic
eradications of invasive mammals’. In: C.R. Veitch, M.N. Clout and
D.R. Towns (eds.) Island invasives: eradication and management, pp.
250–258. Occassional Paper SSC no. 42. Gland, Switzerland: IUCN
and Auckland, New Zealand: CBB.
Luna-Mendoza, et al.: Guadalupe Island ten years after goat eradication
AZE. (2010). 2010 AZE Update. Alliance for Zero Extinction. <www.
zeroextinction.org>. Accessed 18 June 2017.
Berglund, H., Järemo, J. and Bengtsson, G. (2009). ‘Endemism predicts
intrinsic vulnerability to nonindigenous species on islands’. The
American Naturalist 174: 94–101.
Brook, B.W., Sodhi, N.S. and Bradshaw, C.J.A. (2008). ‘Synergies
among extinction drivers under global change’. Trends in Ecology and
Evolution 23: 453–460.
Campbell, K. and Donlan, C.J. (2005). ‘Feral goat eradications on
islands’. Conservation Biology 19: 1362–1374.
Camps, J. and Ramos, M. (2012). ‘Grape harvest and yield responses
to inter-annual changes in temperature and precipitation in an area of
north-east Spain with a Mediterranean climate’. International Journal
of Biometeorology 56: 853–864.
Castro, R., Mascarenhas, A., Sánchez-Barba, A., Durazo, R. and GílSilva, E. (2005). ‘Condiciones Meteorológicas en el sur de Isla
Guadalupe’. In: E. Peters and K. Santos Del Prado (eds.) Restauración y
Conservación de la Isla Guadalupe, pp. 27–37. México, D.F.: Instituto
Nacional de Ecología.
Ceceña-Sánchez, M.L. (2014). ‘Estudio Fitosociológico de las
Comunidades Vegetales de los Matorrales de Isla Guadalupe, México’.
MSc thesis. Ensenada: Universidad Autónoma de Baja California.
Chynoweth, M.W., Litton, C.M., Lepczyk, C.A., Hess, S.C. and Cordell,
S. (2013). ‘Biology and impacts of Pacific Island invasive species. 9.
Capra hircus, the feral goat (Mammalia: Bovidae)’. Pacific Science 67:
141–156.
Coblentz, B.E. (1978). ‘The effects of feral goats (Capra hircus) on island
ecosystems’. Biological Conservation 13: 279–285.
CONAFOR. (2004). Protección, Restauración y Conservación de Suelos
Forestales. Manual de Obras Prácticas. Jalisco, México: Comisión
Nacional Forestal, Secretaría del Medio Ambiente y Recursos Naturales.
Courchamp, F., Chapuis, J.-L. and Pascal, M. (2003). ‘Mammal invaders
on islands: Impact, control and control impact’. Biological Reviews 78:
347–383.
Crum, H. and Miller, H.A. (1956). ‘Bryophytes from Guadalupe Island,
Baja California’. The Southwestern Naturalist 1: 116–120.
Delgado-Argote, L.A., García-Abdeslem, J. and Mendoza-Borunda,
R. (1993). ‘Correlación Geológica entre la Batimetría y los Rasgos
Estructurales del Oriente de la Isla Guadalupe, México’. In: L.A.
Delgado-Argote and A. Martín-Barajas (eds.) Contribuciones a la
Tectónica de México, pp. 1–11. México: Monografía No. 1 de la Unión
Geofísica Mexicana.
Garcillán, P.P., Vega, E. and Martorell, C. (2012). ‘The Brahea edulis
palm forest in Guadalupe Island: A North American fog oasis?’ Revista
Chilena de Historia Natural 85: 137–145.
Granda, E., Rossatto, D., Camarero, J.J., Voltas, J. and Valladares, F.
(2014). ‘Growth and carbon isotopes of Mediterranean trees reveal
contrasting responses to increased carbon dioxide and drought’.
Oecologia 174: 307–317.
Hanna, G.D. (1925). ‘Expedition to Guadalupe Island, Mexico, in 1922’.
Proceedings of the California Academy of Sciences, Fourth Series XIV:
217–275.
Holdgate, M.W. (1967). ‘The influence of introduced species on the
ecosystems of temperate oceanic islands’. In: C.E. Conrad and W.C.
Oechel (eds.). Towards a new Relationship of Man and Nature in
Temperate Lands. Part III. Changes due to Introduced Species, pp.
152–177. Lucerne: IUCN.
Huey, L.M. (1925). ‘Guadalupe Island: An object lesson in man-caused
devastation’. Science 61: 405–407.
Jehl Jr., J.R. and Everett, W.T. (1985). ‘History and status of the avifauna
of Isla Guadalupe, Mexico’. Transactions of the San Diego Society of
Natural History 20: 313–336.
Junak, S.J., Keitt, B., Tershy, B., Croll, D., Luna-Mendoza, L. and
Aguirre-Muñoz, A. (2005). ‘Esfuerzos Recientes de Conservación y
Apuntes sobre el Estado Actual de la Flora de Isla Guadalupe’. In: E.
Peters and K. Santos del Prado (eds.) Restauración y Conservación
de la Isla Guadalupe, pp. 83–93. México, D.F.: Instituto Nacional de
Ecología.
León de La Luz, J.L., Rebman, J.P. and Oberbauer, T.A. (2003). ‘On
the urgency of conservation on Guadalupe Island, Mexico: Is it a lost
paradise?’ Biodiversity and Conservation 12: 1073–1082.
León de La Luz, J.L., Rebman, J.P. and Oberbauer, T.A. (2005). ‘El Estado
Actual de la Flora y la Vegetación de Isla Guadalupe’. In: E. Peters
and K. Santos del Prado (eds.) Restauración y Conservación de la Isla
Guadalupe, pp. 55–65. México, D.F.: Instituto Nacional de Ecología.
León Portilla, M. (1989). Cartografía y Crónicas de la Antigua California.
México, D.F.: Universidad Nacional Autónoma de México.
Luna-Mendoza, L., Keitt, B. and Junak, S. (2007). ‘The restoration of
Guadalupe Island revisited’. Fremontia 35: 14–17.
Luna-Mendoza, L., Aguirre-Muñoz, A., Méndez-Sánchez, F., VargasHernández, J.J., Rodríguez-Chávez, O.G., Martínez-Rodríguez, J.M.,
Cosio-Muriel, D., Gómez-Reyes, J., Puebla-Hernández O. y LuvianosColin, S. (2016). Restauración Integral del Ecosistema del Bosque de
Ciprés de la RB Isla Guadalupe. Reporte final. Ejercicio PROCER
2016. GECI-COLPOS-CIPACTLI. Report for the Comisión Nacional
de Áreas Naturales Protegidas-Reserva de la Biosfera Isla Guadalupe.
Mack, R.N., Simberloff, D., Lonsdale, W.M., Evans, H., Clout, M.N. and
Bazzaz, F.A. (2000). ‘Biotic invasions: Causes, epidemiology, global
consequences, and control’. Ecological Applications 10: 689–710.
Meli, P., Holl, K.D., Benayas, J.M.R., Jones, H.P., Jones, P.C., Montoya,
D. and Mateos, D.M. (2017). ‘A global review of past land use, climate,
and active vs passive restoration effects on forest recovery’. PLOS ONE
12: e0171368.
Méndez-Sánchez, F.A. (2012). ‘Co-management and Small-scale
Fisheries in Mexico: The Case of a Fishers' Cooperative in Cedros and
San Benito islands’. MSc thesis. Auckland, NZ: University of Auckland.
Minnich, R. A. (1982). ‘Grazing, Fire and the Management of Vegetation
of Santa Catalina Island, California’. In: C.E. Conrad and W.C. Oechel
(eds.). Proceedings of the Symposium on Dynamics and Management of
Mediterranean-type Ecosystems, pp. 444–449. Berkeley, CA: USDA,
Forest Service, Pacific Southwest Forest and Range Experiment Station,
General Technical Report PSW-58.
Minnich, R.A., Franco Vizcaino, E. and Dezzani, R.J. (2009). ‘The
El Niño/Southern Oscillation and precipitation variability in Baja
California, Mexico’. Atmósfera 13: 1–20.
Moran, R. (1996). The Flora of Guadalupe Island, Mexico, San Francisco,
USA: California Academy of Sciences.
Morgan, L., Maxwell, S., Tsao, F., Wilkinson, T.A.C. and Etnoyer,
P. (2005). Marine Priority Conservation Areas: Baja California to
the Bering Sea. Montreal, Canada: Commission for Environmental
Cooperation of North America and the Marine Conservation Biology
Institute.
Morrone, J.J., Organista, D.E., Zúñiga, C.A. and Bousquets, J.L. (1999).
‘Preliminary classification of the Mexican biogeographic provinces: A
parsimony analysis of endemicity based on plant, insect, and bird taxa’.
The Southwestern Naturalist 44: 507–514.
Morrone, J.J., Espinosa-Organista, D. and Llorente-Bousquets, J. (2002).
‘Mexican biogeographic provinces: Preliminary scheme, general
characterizations, and synonymies’. Acta Zoológica Mexicana 85:
83–108.
Myers, N., Mittermeier, R.A., Mittermeier, C.G., Da Fonseca, G.A. and
Kent, J. (2000). ‘Biodiversity hotspots for conservation priorities’.
Nature 403: 853–858.
Oberbauer, T.A. (2005). ‘La Vegetación de Isla Guadalupe: Entonces y
Ahora’. In: E. Peters and K. Santos del Prado (eds.). Restauración y
Conservación de la Isla Guadalupe, pp. 39–53. México, D.F.: Instituto
Nacional de Ecología.
Parkes, J., Henzell, R. and Pickles, G. (1996). Managing Vertebrate Pests:
Feral Goats. Canberra: Australian Government Publishing Service.
Ramos Franco, C.A. (2007). ‘Propuesta de Manejo de la Erosión Hídrica
para la Restauración de Suelo del Bosque de Ciprés de Isla Guadalupe.’
MSc thesis. Ensenada: Universidad Autónoma de Baja California.
Raven, P.H. (1965). ‘The Floristics of the California Islands’. In: R.N.
Philbrick (ed.) 1st Symposium on the Biology of the California Islands,
pp. 57–71. Santa Barbara, California: Santa Barbara Botanic Garden.
Reaser, J.K., Meyerson, L.A., Cronk, Q., De Poorter, M., Eldrege,
L.G., Green, E., Kairo, M., Latasi, P., Mack, R.N., Mauremootoo,
J., O'Dowd, D., Orapa, W., Sastroutomo, S., Saunders, A., Shine, C.,
Thrainsson, S. and Vaiutu, L. (2007). ‘Ecological and socioeconomic
impacts of invasive alien species in island ecosystems’. Environmental
Conservation 34: 98–111.
Rebman, J.P., Oberbauer, T.A. and León de La Luz, J.L. (2005). ‘La Flora
de Isla Guadalupe y sus Islotes Adyacentes’. In: E. Peters and K. Santos
del Prado (eds.) Restauración y Conservación de la Isla Guadalupe, pp.
67–81. México, D.F.: Instituto Nacional de Ecología.
Rogers, D.L., Matheson, A.C., Vargas-Hernández, J.J. and Guerra-Santos,
J.J. (2006). ‘Genetic conservation of insular populations of Monterey
pine (Pinus radiata D. Don)’. Biodiversity and Conservation 15: 779–
798.
SARH-Colegio de Postgraduados. (2010). Normales Climatológicas
1951–2010. <http://www.cm.colpos.mx/meteoro/progde/norm/norm24/
listd.htm>. Accessed 28 November 2012.
Searcy-Bernal, R., Ramade-Villanueva, M.R. and Altamira, B. (2010).
‘Current status of abalone fisheries and culture in Mexico’. Journal of
Shellfish Research 29: 573–576.
Simberloff, D. (1995). ‘Why do introduced species appear to devastate
islands more than mainland areas?’ Pacific Science 49: 87–97.
Sol, D. (2007). ‘Do Successful Invaders Exist? Pre-Adaptations to
Novel Environments in Terrestrial Vertebrates’. In: W. Nentwig (ed.)
Biological Invasions, pp. 127–141. Berlin, Heidelberg: Springer.
Ugolini, F.C. and Dahlgren, R.A., (2002). ‘Soil development in volcanic
ash’. Global Environmental Research–English Edition 6: 69–82.
Veitch, C.R., Clout, M.N. and Towns, D.R. (eds.) (2011). ‘Island
invasives: eradication and management’. Occassional Paper SSC no.
42. Gland, Switzerland: IUCN and Auckland, New Zealand: CBB.
Vidal, R.M., Berlanga, H. and del Coro Arizmendi, M. (2009). ‘Important
Bird Areas: Mexico’. In: C. Devenish, D.F. Díaz Fernández, R.P.
Clay, I. Davidson and I. Yépez Zabala (eds.) Important Bird Areas
Americas - Priority Sites for Biodiversity Conservation. Quito: BirdLife
International (BirdLife Conservation Series No. 16).
Whittaker, R.J. and Fernández-Palacios, J.M. (2007). Island
Biogeography: Ecology, Evolution, and Conservation. New York:
Oxford University Press.
Wilkinson, T., Wiken, E., Bezaury-Creel, J., Hourigan, T., Agardy, T.,
Herrmann, H., Janishevski, L., Madden, C., Morgan, L. and Padilla, M.
(2009). Marine Ecoregions of North America. Montreal: Commission
for Environmental Cooperation.
575
R. Luxmoore, R. Swann and E. Bell
Luxmoore, R.; R. Swann and E. Bell. Canna seabird recovery project: 10 years on
Canna seabird recovery project: 10 years on
R. Luxmoore1, R. Swann2 and E. Bell3
National Trust for Scotland, Hermiston Quay, 5 Cultins Road, Edinburgh, EH11 4DF, Scotland. <rluxmoore@nts.
org.uk>. 2Highland Ringing Group, 14 St Vincent Road, Tain, Ross-shire, IV19 9JR, Scotland. 3Wildlife Management
International Ltd, P.O. Box 607, Blenheim, Marlborough 7240, New Zealand.
1
Abstract Rats were eradicated in 2005–2006 from the islands of Canna and Sanday, Scotland (total area 1,320 ha).
Poison bait was laid from December 2005 onwards and the last rat was killed in February 2006. An intensive period of
monitoring over the next two years confirmed that no rats remained on the islands. Seabirds have been monitored on
Canna for nearly 50 years and some species have shown good evidence of recovery since the eradication. Other species
have not recovered and this may have been due to mortality caused by food shortages or storm events which have
been impacting seabirds in the region. These regional changes in pressures affecting the seabird populations make the
interpretation of the impacts of the rat eradication programme much more difficult. Atlantic puffins, formerly confined to
offshore stacks, have recolonised sites on the mainland of Canna and a count of over 2,000 was recorded in 2016. Manx
shearwaters, which had ceased nesting in the monitored colony have made a slow recovery to one or two pairs in 2016.
Productivity has also increased from a low of 0.2 chicks per nest in the 1990s to 0.74 in 2017. European shags nesting in
boulder colonies were most susceptible to rat predation. One such colony has recovered from 45 nests in 2005 to 75 in
2016 and productivity increased from less than 0.7 chicks per nest to an average of 1.6 following eradication. Populations
of shags nesting in cliff locations have shown no recovery or have declined. Mew gulls, which nest along the shoreline,
have increased from five to over 30 pairs. Other seabirds, such as common guillemots and black-legged kittiwakes, have
shown no clear trends and are probably affected by other factors. Rabbit populations have increased on both islands,
reaching an estimated 15,500 animals in 2013 that were causing considerable damage through grazing, erosion, and
disturbance of archaeological remains. It is unclear whether the increase in rabbit numbers can be attributed to rat
eradication. An intensive control programme has brought the rabbit population under control. While some seabirds have
responded positively to the rat eradication, the response of some has been slow and others have not responded, probably
as a result of regional pressures on their survival. It is important that monitoring of both seabirds and rabbits continues to
track the success of this important seabird colony.
Keywords: contingency plans, invasive species, monitoring, quarantine, rabbits, rat eradication
INTRODUCTION
The islands of Canna and Sanday, which are connected
at low tide (total area 1,320 ha) are in the Inner Hebrides,
off western Scotland. They are owned and managed by the
National Trust for Scotland and are designated as a Special
Protection Area because of their internationally important
seabird colony. Seabird populations and breeding success
have been monitored on the islands since 1969 by the
Highland Ringing Group, making this one of the best
monitored sites in Scotland. The main species present are
common guillemot (Uria aalge), razorbill (Alca torda),
black-legged kittiwake (Rissa tridactyla), northern fulmar
(Fulmaris glacialis), European shag (Phalacrocorax
aristotelis) and Atlantic puffin (Fratercula arctica). Manx
shearwater (Puffinus puffinus) used to be present in large
numbers (1,500 apparently occupied burrows) but suffered
very poor breeding success and, by 2000, had been
virtually wiped out (Swann, 2002) Other seabirds were
also recorded as declining. Predation by a large population
of brown rats (Rattus norvegicus) was identified as the
likely cause of this decline, from three types of evidence: 1.
Direct observation of increasing numbers of rats foraging
in the seabird colonies and of stashes of predated egg
shells and carcases; 2. Declining numbers and decreasing
breeding success of vulnerable species; and 3. Changing
nesting behaviour of breeding seabirds moving to less
accessible sites. After favourable feasibility studies (Bell
& Bell, 2004), it was decided to eradicate the rats using
poison bait, and funding was obtained from the EU LIFE
fund, Scottish Natural Heritage and the National Trust for
Scotland. The programme objective was to halt declines
in breeding seabird populations on Canna and Sanday and
to facilitate their recovery and long-term protection. It
was carried out under contract by Wildlife Management
International, starting in late 2005. By February 2006 the
last rat sign was detected and, after a two-year period of
intensive monitoring, the island was declared rat-free in
2008 (see Bell, et al., 2011).
Canna and Sanday are inhabited by a population of
15–20 people and are farmed with a mixture of sheep and
cattle. They are served by a ferry service five days a week.
The harbour and all houses are in the eastern portion of
the islands, where there are a number of fenced pastures
and some planted woodlands. There are high cliffs around
much of the coast, particularly to the north and west, and
the higher ground is mostly covered in wet heath. There
is a population of distinctive, large (presumed introduced)
field mice (Apodemus sylvaticus) that were not removed by
the rat poisoning programme and a substantial (introduced)
population of rabbits (Oryctolagus cuniculus). There is
a small number of (introduced) hedgehogs (Erinaceus
europaeus), and regular sightings of European otters (Lutra
lutra), but no other ground predators. Two pairs of whitetailed eagles (Haliaeetus albicilla), one pair of golden eagles
(Aquila chrysaetos), up to two pairs of peregrine falcons
(Falco peregrinus), about 15 pairs of common buzzards
(Buteo buteo) and ravens (Corvus corax) regularly breed
on the island and there are small numbers of great skuas
(Catharacta skua) and great black-backed gulls (Larus
marinus). Predation by eagles, and possibly locally by
otters, may impact populations of northern fulmar, while
ravens may impact shags, particularly on the cliff-nesting
colonies (Swann, 2008; Swann, et al., in press), but none of
these predators exerts substantial pressure on other seabird
populations.
Following the eradication programme, biosecurity
measures were put in place, consisting of continuous
monitoring (wax blocks and kill traps), quarantine and
contingency plans. No incursions of rats have been
detected.
The main post-eradication monitoring has been a
continuation of the long-running seabird programme which
can be used to detect any changes following the eradication
of rats. A rapid expansion in the rabbit population was
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
576
up to meet the challenge, pp. 576–579. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Luxmoore, et al.: Canna seabird recovery
noted in 2011–2013 which caused locally severe grazing
and considerable erosion through collapsed burrows.
This necessitated the introduction of control measures
consisting of a rapid reduction cull in January–March 2014,
followed by continuous lower level culling thereafter.
There have been surveys of vegetation condition (SNH,
2014), grassland fungi (Murfitt & Macdonald, 2012),
invertebrates (Rotheray & Lyszkowski, 2012) and lichens
(Acton, 2011). Any changes in the vegetation are thought
to be due to rabbit grazing or livestock management, rather
than directly related to the removal of rats.
This paper reviews the changes in seabird population
size and breeding success reported in Swann, et al. (in
press) and discusses fluctuations in rabbit populations and
the control measures employed.
MATERIAL AND METHODS
All seabird population and productivity estimates
follow the methodology of Walsh, et al. (1995) and are
described in Swann, et al. (in press). Seabird population
estimates for the years 1995 to 2017, derived from Swann
(2008) and Swann, et al. (in press), were analysed. For
northern fulmar, European shag, black-legged kittiwake,
mew gull, herring gull, lesser black-backed gull and
greater black-backed gull, the population figures represent
the number of apparently occupied territories or apparently
occupied nest sites throughout the two islands. For
common guillemots and razorbills, they are the number
of nest sites at a small number of accessible monitoring
plots. Nest sites were recognised by the presence of an egg,
chick or, in the case of razorbill, a shell or dense mass of
droppings. The data were divided into the 11 years prior
to the eradication (1995–2005) and the 12 years following
(2006–2017). Population trends were determined by fitting
an exponential line to each set of data. The exponents
shown on the graphs represent r in the equation N = ert,
where N is the population size (pairs) and t is the time in
years.
Productivity estimates represent large young/chicks per
occupied nest within all the monitoring plots for northern
fulmar and black-legged kittiwake. For great black-backed
gulls and herring gulls, they represent large chicks per
apparently occupied territory. For European shags, the
breeding success (chicks per occupied nest) is separated
into plots within the boulder colonies and plots on cliff
sites. Productivity of Manx shearwaters was calculated as
the number of large chicks produced per occupied burrow
located.
In June 2013 and 2014 (the latter following a rabbit
reduction programme in January–March 2014) rabbit
populations were estimated by walking over the entire
island using the Modified McLean scale, an 8-point
scale based on the observation of rabbits and pellets
(NPCA, 2012). Densities were estimated within different
topographic or vegetation units and the area of each unit
was calculated using GIS. The approximate population of
rabbits was then calculated by multiplying the area of each
unit by the densities corresponding with the McLean scale.
Rabbit culls were recorded by the trappers on a weekly
basis in different zones.
RESULTS
Seabird population size
Seabird population trends are shown in Fig. 1.
Numbers of breeding pairs of almost all species, except
the black-legged kittiwake were declining prior to 2005
but after 2006, three gull species were stable or slowly
increasing and populations of all other species, except
the northern fulmar, showed a reduced rate of decline. It
should be noted that the correlation coefficients (R2) were
low, showing significant declines for most species prior to
eradication but significant increases only for mew gull and
lesser black-backed gull after eradication.
In the case of Manx shearwater, the population had
already fallen to very low levels prior to 1995 from a high
Fig. 1 Population trends in the numbers of breeding pairs of nine species of seabird on Canna during the period 1995–
2017. Data were split into pre-eradication (1995–2005) and post-eradication (2006–2017) and exponential graphs fitted
to each.
577
Island invasives: scaling up to meet the challenge. Ch 3C Strategy: Outcomes
of 1,500 pairs in the mid-1970s. By 2000, no nests could
be located in the main colony, along Tarbert Road (see Fig.
2). Following the rat eradication in 2006 the first nesting
shearwater was detected again in the Tarbert Road colony,
but by 2017 this had grown to only two nests. A further
four nests were located at accessible locations in the west
of the island and all of these nests were monitored for
breeding success. Based on calling behaviour at night, it
was estimated that there were more nests, possibly 10–20
pairs, in inaccessible locations (M. Carty, pers. comm.)
Mew gulls (Larus canus) nest along the shore on Canna.
Numbers have increased from nine pairs at the time of the
eradication to around 30 pairs in recent years. Other large
gulls, especially herring gulls (Larus argentatus), declined
rapidly in the 1990s and early 2000s similar to colonies
elsewhere in Scotland. Since 2006, populations on Canna
have remained low.
The overall number of breeding European shags
had dropped to about 300 nests by 2005. Different subcolonies have performed differently, with those nesting
under boulders declining most rapidly (Swann, 2005).
One boulder colony, Lamasgor, has subsequently shown
an increase from 45 nests in 2005 to 75 in 2016. Other
sub-colonies, particularly those on cliffs, however, have
remained stable, or declined, so that overall the population
has not increased. Nevertheless, the overall rate of decline
of all of the nests on the island has slowed (Fig. 1).
Atlantic puffin breeding populations are difficult to
monitor, especially because, prior to 2005, they were
virtually confined to two inaccessible stacks. After 2006
they began to spread to sites along the north coast of Canna
at Guegasgor (see Fig. 2). Where more accurate census
methods are not practical, Walsh, et al. (1995) recommend
counting puffins rafting on the sea near the colony. A count
of rafting puffins in 1995 gave 1190 individuals while, in
2016, 2050 were counted. Though not conclusive, this is
consistent with an increase in breeding numbers as well
as the recorded expansion of the puffin colony to colonise
previously unoccupied sites on the mainland of Canna.
Seabird breeding success
Between 1999 and 2004, breeding success of European
shags in monitored nests in boulder colonies had dropped to
<0.7 young per nest. Since the eradication breeding success
has averaged 1.6 young per nest. In contrast, breeding
success at colonies nesting on cliff ledges averaged 0.97
per nest over the years 2006–2017 (Table 1).
Manx shearwater breeding productivity has greatly
improved. In the 1980s it averaged 0.6 young/nest,
dropping to <0.2 young/nest in the mid-1990s. Since 2009.
out of a total of 19 burrows that were known to contain an
egg, 14 successfully produced a chick, an average of 0.74
young/nest.
Rabbit population
The rabbit population on Canna has routinely fluctuated
in response to disease and weather conditions, but numbers
had not been formally monitored. By 2013, rabbit numbers
were causing serious damage to agricultural interests and
Fig. 2 Estimated rabbit density on Canna in July 2013.
Densities relate to the Modified McLean Scale (NPCA,
2012).
were damaging archaeological remains. A rapid assessment
of rabbit population density gave an overall population
estimate of 15,500 (Fig. 2). As a consequence, a rabbit
population reduction exercise was carried out in January–
March 2014, with a total of 8,200 rabbits removed. This
brought the population down to an estimated 7,000 by July
2014. Continuous culling of around 5,000 rabbits a year
has maintained it at a level where agricultural damage is
acceptable.
DISCUSSION
The response of Canna seabirds to the successful rat
eradication in winter 2005/06 was species specific. For
some populations, such as the Atlantic puffin, the mew gull
and the boulder-nesting colonies of European shag, there
have been apparent increases in numbers of breeding pairs.
The colony of European shags on Canna is of international
importance and formerly numbered 1,800 pairs. Many of
them nest under boulders in relatively accessible locations
and these birds were found to be particularly susceptible
to rat predation, declining most rapidly. Simultaneously,
there was a shift whereby a greater proportion of shags
started nesting on more exposed cliff locations where they
were less susceptible to rat predation but more exposed to
avian predation (Swann, 2005). Following the eradication
of rats, some boulder colonies have expanded rapidly but
the cliff-nesting birds have continued to decline. The net
effect has been a continuing decline in the overall shag
population but at a reduced rate.
Similarly, although populations of common guillemots
and razorbills have continued to decline there has been a
slowing in the rate of decline. These two species of auk nest
in similar locations to the European shags and were also
affected by rat predation. Guillemots were badly affected
by a severe period of stormy weather in western Scotland
in late summer 2004 (Swann, 2004) which caused heavy
mortality of both adults and chicks. Ringing studies showed
that adult survival dropped from a long term average of
0.9 to 0.6 between 2004 and 2005. Breeding numbers of
guillemots on Canna were very low in 2005 and breeding
success remained low until 2008 probably as a result of
subsequent food shortages. Breeding success improved in
2009 and the population has started to increase.
Numbers of breeding razorbills showed a sharp jump
in 2006, and this was almost certainly due to a reduction
Table 1 Breeding productivity of Manx shearwaters at Tarbert Road colony and European shags in the boulder and cliff
colonies on Canna, 2001–2017. “-“ = Productivity not monitored.
Year
Manx
shearwater
Shag – boulder
colonies
Shag – cliff
colonies
578
01
02
03
04
05
06
07
08
-
-
-
-
-
-
-
-
1.2
1.4
1.5
1.8
-
1.4
-
0.7
1.0
0.3
0.7
1.5
0.8
1.6
0.1 0.26 0.16 0.01 0.7
-
-
-
1.4
0.7
09
10
11
12
13
14
15
16
17
0
1
1
0.25
0.5
0.2
0.5
1
0.5
1
1.5
1.8
2.2
1.3
1.0
0.7
1.2
1.0
1.2
Luxmoore, et al.: Canna seabird recovery
in predation by rats, eggs appearing in areas that had been
clear of nesting for several years (Swann, 2008). However,
the number of occupied breeding sites then remained
roughly stable at this level until 2016, and the breeding
success remained low, probably as a result of food shortage
(Swann, et al., in press). Since occupied breeding sites in
this species are identified by the presence of an egg or
chick, failure to breed does not necessarily indicate the
absence of adults attempting to breed – they may have laid
an egg and then left again following predation of the egg.
The population increase observed in 2006 may therefore
indicate higher early survival of eggs rather than a greater
number of adults attending the colony.
The large gulls, especially the lesser black-backed gull
and the herring gull, suffered a particularly sharp decline
in the period 2000–2005 and have since shown a slight
increase. Foster, et al. (2017) have analysed this trend and
shown that it is closely correlated with the commercial
landings of fish in the nearby port of Mallaig, with the gulls
feeding extensively on discards from the fishing industry.
Since 2006, numbers have stabilised at a much lower level.
It is unclear whether the subsequent slow increase of lesser
black-backed gulls had anything to do with the reduction
in rat predation.
The number of breeding pairs of other species, such as
the northern fulmar, have continued to decline and there has
been no reduction in the rate of decline. The fulmars nest
largely in cliff locations and would have been less affected
by rat predation, although it is possible that predation by
eagles may be significant (Swann, 2008).
The black-legged kittiwake population has stayed
roughly stable over the whole period, although it too
suffered poor breeding success in the period 2005–2008,
probably as a result of the food shortage experienced by
other seabirds in the region. This species typically nests
on near-vertical cliffs and is therefore probably the least
susceptible to ground predators such as rats. The causes of
any changes in population size or breeding success must
therefore be sought elsewhere.
One species that was expected to benefit strongly
from the removal of rats was the burrow-nesting Manx
shearwater. Although there has been a tiny increase in
breeding numbers and a clear improvement in breeding
productivity, there are still thought to be fewer than 20
pairs nesting on the island. Shearwaters are long-lived and
slow-maturing species, possibly not breeding until eight
years of age and so endogenous growth of the surviving,
relict population would be expected to be slow (Brooke,
et al., 2018). However, Canna lies next to the larger
Manx shearwater colony on Rum (estimated to be around
60,000 pairs) and is regularly visited by (presumably nonbreeding) birds at night. It is apparent that these have not
colonised the former colony on Canna to any great extent.
The shearwater colony on Rum is subject to predation by a
large population of brown rats. The impact of this predation
is unclear and it is not known whether the colony is stable
or declining (Lambert, et al., 2015). If predation is high,
the pressure for emigration from Rum would be lower
than from a colony that was limited by shortage of nest
sites. An attempt was made to attract breeding birds to recolonise Canna by playing recordings of shearwater calls
at night in 2006 and 2007. This was discontinued because
of a lack of obvious success.
It is possible that the growth in the rabbit population is
attributable to the removal of rats as young rabbits would
be likely to have suffered predation by rats. However, large
fluctuations are a characteristic of rabbit populations and
high numbers have been reported in previous years. Thus,
while 15,500 may seem a high population for Canna, and
it undoubtedly caused damage to agricultural interests, it
may not be unprecedented. It has been necessary to control
rabbits on Canna for many years prior to the eradication
of rats although historically the rabbit populations used to
cycle due to outbreaks of myxomatosis. Surprisingly, there
has been no evidence of myxomatosis in recent years. Total
eradication of rabbits has been deemed unfeasible (Bell,
2012) and so it is inevitable that control of rabbit populations
will be necessary for the foreseeable future to prevent the
build-up of excessive numbers. Although vacated rabbit
burrows can provide nest sites for Manx shearwaters, at
high densities there is likely to be competition for burrows
and this will reduce sites available for shearwaters. The
burrows have also caused severe erosion, including large
landslips, in some of the former shearwater colonies (Bell,
2012).
Overall, the removal of rats from Canna has had some
very beneficial impacts on some species of seabirds but
this effect was masked for other species by some very
difficult local conditions in the period 2004–2008, firstly
by storm-related mortality and subsequently by regional
food shortages. The gulls were also impacted by a lack of
fisheries discards following a drop in commercial fisheries.
Because these external factors occurred at approximately
the same time as the rat eradication programme it may
take many years for the full benefits to play out. It is clear
that continued detailed monitoring of seabird populations
and breeding success is vital in unravelling the complex
interactions between local conditions on the breeding
colony and regional changes in the marine ecosystem.
REFERENCES
Acton, A. (2011). Canna and Sanday Lichen Survey. Unpublished report
to the National Trust for Scotland, Edinburgh, Scotland.
Bell, E.A. and Bell, B. (2004). An Assessment of the Proposal to Eradicate
Brown Rats from the Isle of Canna, Scottish Hebrides. Unpublished
report to the National Trust for Scotland, Edinburgh, Scotland.
Bell, E., Boyle, D., Floyd, K., Garner-Richards, P., Swann, B.,
Luxmoore, R., Patterson, A., and Thomas, R. (2011). ‘The groundbased eradication of Norway rats (Rattus norvegicus) from the Isle
of Canna, Inner Hebrides, Scotland’. In: C.R. Veitch, M.N. Clout and
D.R. Towns (eds.) Island invasives: eradication and management, pp.
269–274. Occassional Paper SSC no. 42. Gland, Switzerland: IUCN
and Auckland, New Zealand: CBB.
Bell, E. (2012). Rabbits on Canna: Assessment and Advice on the Rabbit
Situation on Isle of Canna, Inner Hebrides, Scotland. Unpublished
report to the National Trust for Scotland, Edinburgh, Scotland.
Brooke, M.de L., Bonnaud, E., Dilley, B.J., Flint, E.N., Holmes, N.D.,
Jones, H.P., Provost, P., Rocamora, G., Ryan, P.G., Surman, C., and
Buxton, R.T. (2018). ‘Seabird population changes following mammal
eradications on islands’. Animal Conservation 21(1): 3–12.
Foster, S., Swann, R.L. and Furness, R.W. (2017). ‘Can changes in fishery
landings explain long-term population trends in gulls?’ Bird Study 64:
90–97.
Lambert, M., Carlisle, S. and Cain I. (2015). The Role of Brown Rat
(Rattus norvegicus) Predation in Determining Breeding Success
of Manx Shearwaters (Puffinus puffinus) on Rum. Scottish Natural
Heritage Commissioned Report No. 697, Inverness, Scotland.
Murfitt, A. and Macdonald, A. (2012). Grassland Fungi Survey; Isle
of Canna. Unpublished report to the National Trust for Scotland,
Edinburgh, Scotland.
NPCA (2012). “Modified Maclean Scale”. <http://www.npca.org.nz/
index.php/news/84-general/226-rabbit-control-training.html>.
Rotheray, G.E. and Lyszkowski, R. (2012). The Entomological
Significance of Coastal Grassland on Canna and Sanday. Unpublished
report to the National Trust for Scotland, Edinburgh, Scotland.
SNH (2014). Site Condition Monitoring Note, Canna and Sanday, Cycle
3. Unpublished report. Inverness, Scotland: Scottish Natural Heritage.
Swann, R.L. (2002). Canna Seabird Studies 2001. Joint Nature
Conservation Committee Report 327, Peterborough, UK.
Swann, B. (2004). ‘Seabird Wreck, North West Scotland late summer
2004’. Seabird Group Newsletter 98: 12–14.
Swann, R.L. (2005) Canna Seabird Studies 2005. Joint Nature
Conservation Committee Report, No. 474b, Peterborough, UK.
Swann, R.L. (2008). Canna Seabird Studies 2008. Joint Nature
Conservation Committee Report No. 474e, Peterborough, UK.
Swann, R.L., Aiton, D.G., Call, A., Foster, S., Graham, A., Graham, K.
and Young, A. (in press). Canna Seabird Studies 2017. Joint Nature
Conservation Committee Report. Peterborough, UK.
Walsh, P.M., Halley, D.J., Harris, M.P., del Nevo, A., Sim, I.M.W. and
Tasker, M.L. (1995). Seabird Monitoring Handbook for Britain and
Ireland: A Compilation of Methods for Survey and Monitoring of
Breeding Seabirds. Peterborough: JNCC, RSPB, ITE, Seabird Group.
579
J.E. Millett, W. Accouche, J. van de Crommenacker, M.A.J.A. van Dinther, A. de Groene, C.P. Havemann, T.A. Retief, J. Appoo and R.M. Bristol
Millett, J.E.; W. Accouche, J. van de Crommenacker, M.A.J.A. van Dinther, A. de Groene, C.P. Havemann, T.A. Retief, J. Appoo and R.M. Bristol.
Conservation gains and missed opportunities 15 years after rodent eradications in the Seychelles
Conservation gains and missed opportunities 15 years after rodent
eradications in the Seychelles
J.E. Millett1, W. Accouche1, J. van de Crommenacker1,2, M.A.J.A. van Dinther2, A. de Groene3, C.P. Havemann4,
T.A. Retief4, J. Appoo1 and R.M. Bristol5
Green Islands Foundation, PO Box 246, Victoria, Mahé, Seychelles. <jemillett2003@yahoo.co.uk>. 2Frégate Island
Private, PO Box 330, Victoria, Mahé, Seychelles. 3WWF-Netherlands, Driebergseweg 10, 3708 JB Zeist, Netherlands.
4
North Island, P.O. Box 1176, Victoria, Mahé, Seychelles. 5La Batie, Beau Vallon, Mahé, Seychelles.
1
Abstract The Seychelles was one of the first tropical island nations to implement island restoration resulting in
biodiversity gain. In the 2000s a series of rat eradication attempts was undertaken in the inner Seychelles islands which
had mixed results. Three private islands with tourist resorts successfully eradicated rats: Frégate (2000), Denis Island
(2003) and North Island (2005). Frégate Island was successful with the first eradication attempt whereas North and Denis
Islands were initially unsuccessful, and both required second eradication operations. All three islands have developed
conservation programmes including biosecurity, habitat rehabilitation, and species reintroductions, and have integrated
nature into the tourism experience. Conservation actions, including rat and other invasive species eradications, on these
three islands resulted in the creation of 560 ha of mammalian predator-free land, the reintroduction of seven populations
of five globally threatened birds (GTB) and the safeguarding of two existing GTB populations and several reptile and
invertebrate species. However, on these and many other islands in the Seychelles, the potential of this conservation
“model”, where island owners implement conservation programmes largely funded by the tourism businesses in
collaboration with NGOs (Non Government Organisations), has not been fully realised. We review the rehabilitation
on Frégate, Denis and North Islands from inception to the present, and assess factors that have facilitated the subsequent
development of conservation programmes, the presence of receptive businesses and governmental/NGO/donor support
and explore limitations on business-led island rehabilitation.
Keywords: eradications, invasive alien species, island conservation, rehabilitation Seychelles, tourism
INTRODUCTION
Islands harbour much of the world’s endangered
biodiversity (Kaiser-Bunbury, et al., 2010) and island
species are very vulnerable to the impacts of Invasive Alien
Species (IAS). Over the last 500 years, the majority of
documented plant or vertebrate extinctions have occurred
on islands (Tershy, et al., 2015). Causes include habitat
modification and over exploitation; however, IAS have
played a key role. In particular, invasive mammals such as
rats (Rattus spp.) have been implicated in numerous bird
extinctions, extirpations and population declines (Moors
& Atkinson, 1984; Burger & Gochfeld, 1994; Hilton &
Cuthbert, 2010) as well as reptile declines and impacts
on other taxa (Towns, 1991; Harper & Bunbury, 2015;
Thibault, et al., 2016).
The eradication of rats and other invasive mammals,
often with concomitant habitat rehabilitation, was initially
pioneered in New Zealand and other temperate areas but
has become increasingly practiced in tropical regions,
(Russell & Holmes, 2015; Russell & Broome, 2016). The
understanding of the measures required to successfully
execute mammal eradications on tropical islands has
improved (Keitt, et al., 2015). In the late 1990s, the
Seychelles was one of the first tropical island nations to
implement rodent and multispecies eradications (Merton,
et al., 2002)
The Seychelles archipelago in the Indian Ocean extends
over an Exclusive Economic Zone of 1,374,000 km2 (Fig.
1). The ancient “inner” islands are situated approximately
4ᵒ S and 54ᵒ E and are composed of continental rock,
while the much more recently formed “outer” islands are
formed from raised reefs and sand cays (Stoddart, 1984)
scattered for approximately 1000 km to the south-west
of the inner islands. The Seychelles have high endemism
(Stoddart, 1984) and the inner islands are an Endemic Bird
Area (EBA100); supporting 11 endemic species of bird
(BirdLife International, 2017).
Since the human colonisation of the Seychelles in
the late 18th century, IAS have caused range reductions,
population declines, and extinctions of native species
(BirdLife International, 2000). Alien predators are
considered the most destructive species (Harper &
Bunbury, 2015): most inner islands have had populations
of black rat (Rattus rattus) and feral cat (Felis catus), and
some have had brown rat (R. norvegicus). Only a few inner
islands remained free of mammalian predators, and in the
1980s only four islands larger than 20 ha remained free of
rats (Aride, Cousin, Cousine and Frégate), although feral
cats were on Cousine and Frégate, and house mice (Mus
musculus) on Aride and Frégate. Construction projects on
islands resulted in the introduction of rats including black
rat to Bird Island in 1968 and brown rat to Frégate in 1995.
Four Seychelles’ bird species endemic to the inner
islands were listed as Critically Endangered when at their
lowest known population size: Seychelles magpie-robin
(Copsychus sechellarum), Seychelles white-eye (Zosterops
modestus), Seychelles paradise-flycatcher (Terpsiphone
corvine), Seychelles scops-owl (Otus insularis). Four
species were listed as Vulnerable: Seychelles warbler
(Acrocephalus sechellensis), Seychelles fody (Foudia
sechellarum), Seychelles kestrel (Falco araea) and
Seychelles swiftlet (Aerodramus elaphrus) (BirdLife
International, 2017) (Table 1).
Three species of bird that had been historically
widespread (Gaymer, et al. 1969) became restricted to
black rat-free islands: the Seychelles magpie-robin on
Frégate (Gaymer, et al., 1969; Burt, et al., 2016), Seychelles
warbler on Cousin (Komdeur, 2003), and Seychelles fody
on Cousin, Cousine and Frégate (Vesey-Fitzgerald, 1940).
The Seychelles white-eye had a small population in the
uplands of Mahé and a larger population on Conception
Island which had brown rats, but no black rats (Rocamora,
1997). Aride remained free of rats but did not retain
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
580
up to meet the challenge, pp. 580–587. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Millett, et al.: Fifteen years after rodent eradications in the Seychelles
Table 1 Number of island populations and threat categories of birds endemic to the Granitic Seychelles Endemic
Bird Area (EBA) since 1994.
Species
Seychelles magpie-robin
Smallest documented
Current no of
number of populations populations
1
5
Seychelles white-eye
2
4
Seychelles paradise-flycatcher
Seychelles scops-owl
1
1
2
1
Seychelles warbler
1*
5
Seychelles fody
3
6
Seychelles kestrel
Seychelles swiftlet
Seychelles black parrot
1
3
2
2
3
2
IUCN Threat
category since 1994
Endangered
Critical
Vulnerable
Endangered
Critical
Critical
Endangered
Critical
Near threatened
Vulnerable
Near threatened
Vulnerable
Vulnerable
Vulnerable
Vulnerable
Year
2005–2018
1994–2005
2016–2018
2005–2015
1994–2005
1994–2018
2004–2018
1994–2004
2012–2018
1994–2012
2004–2018
1994–2004
1994–2018
1994–2018
2014–2018
* First reintroductions of Seychelles warblers undertaken prior to 1994 endangered species categories being applied.
populations of endemic birds which may be attributed to
forest loss and cat predation (eradicated in 1930s; Warman
& Todd, 1984). The distributions of these endemic
bird species suggest black rats were an important factor
contributing to decline and extirpation of populations.
Initial conservation efforts focused on the purchase
and protection of two rat-free islands (Cousin, 1968;
Aride, 1973) by NGOs (Non Government Organisations).
Successful attempts were made to reintroduce the
Seychelles warbler to Aride and Cousine (Richardson,
2001; Komdeur, 2003). Attempts to introduce Seychelles
magpie-robin to Cousin and Cousine met with success, but
several introduction attempts to Aride were unsuccessful
(Watson, 1978; Lucking & Ayrton, 1995).
Further progress was achieved with cat eradications
on Frégate and Cousine during the 1980s (Rocamora &
Henriette, 2015). Rodent eradications were not attempted in
the Seychelles until 1996, when black rats were eradicated
from Bird Island, and later in the early 2000s when a a
series of rat and multispecies eradications were initiated
that included privately owned Frégate, Denis and North
Islands. Subsequent habitat rehabilitation and endemic
bird reintroductions were implemented (Thorsen, et al.,
2000; Merton, et al., 2002; Samways, et al., 2010).
This paper reviews the conservation programmes
on three privately owned islands and the conservation
outcomes.
ISLAND DESCRIPTIONS: NORTH, DENIS, AND
FRÉGATE
North Island (Ile du Nord)
Fig. 1 The Seychelles showing islands mentioned in the
text.
The native vegetation of North Island (201 ha) was
replaced in the early 19th century by a coconut (Cocos
nucifera) plantation, which was abandoned in the 1970s,
and guano excavation left pits that are still present today.
The island harboured black rats, cats, and feral cattle
(Bos taurus). Hill (2002) identified the island as having
a high rehabilitation potential; small enough to eradicate
mammals, sufficiently isolated to manage reinvasion risk
and a proportionally large coastal plateau that is likely to
support rehabilitated forest suitable for Seychelles magpie581
Island invasives: scaling up to meet the challenge. Ch 3C Strategy: Outcomes
robins and Seychelles paradise-flycatchers (Currie, et al.,
2003). The island was privately purchased in 1997 by an
eco-tourism company, which opened an exclusive resort in
2005, with the intention to ultimately fund the rehabilitation
of the island. Conservation is promoted through all aspects
of the tourism operation, whereby guests and staff are
educated and encouraged to participate in environmental
activities such as guided hikes, presentations, and data
collection.
Denis Island
Denis is a coralline island of 140 ha located 80 km north
of the capital island of Mahé. Early descriptions mention
abundant land tortoises and seabirds (Bradley, 1940). The
original vegetation of the island, as described in 1773, was
of open grassy areas and forest; probably Pisonia grandis
(Stoddart & Fosberg, 1981). Extensive guano deposits
indicated the historical presence of seabirds. The island
has been altered profoundly, first through the cultivation
of coconuts from around 1890 (Stoddart & Fosberg, 1981),
then through guano extraction in the 1930s (Baker, 1963),
followed by the replanting of coconuts in the 1940s.
In 1975, a new owner built an airstrip and a small hotel
and abandoned the coconut plantation. In 1998 the island
changed ownership again and today it is managed as a
luxury tourist resort with 30 villas.
Denis was considered a priority site for rehabilitation
due to the large area of flat land conducive to rehabilitation,
existing native woodland, and an owner who supported
conservation, if black rats and feral cats were removed
(Hill, 2002). The island appears to be in the natural range
of some endemic birds as, in 2004, a Seychelles magpierobin from Aride flew to Denis (Burt, et al., 2016) and in
2009 a Seychelles sunbird (Cinnyris dussumieri) flew from
Bird Island to Denis (R. Bristol, pers. obs., 2009).
Frégate Island
The original vegetation of Frégate Island (219 ha) was
removed to make way for spice and coconut plantations,
that were abandoned in the 1980s, leaving coconutdominated forest and several areas of the introduced tree
sandragon (Pterocarpus indicus). The original vegetation
is unknown; however, a few relict plants including
Pandanus balfourii and Euphorbia pyrifolia survived on
rocky glacis areas, indicating some of the vegetation that
existed before plantations (J. Millett, pers. obs., 2000). The
availability of canopy-forming sandragon forest and the
absence of rats contributed to the survival of Seychelles
fody, the last Seychelles magpie-robin population, and
rich assemblages of reptiles, amphibians, and invertebrates
including two single-island endemics: a beetle Polposipus
herculeanus, and a snail Pachnodus fregatensis (Canning
2011b; Gerlach, 2006). Frégate was re-developed in
1995–1999 as an exclusive resort, and brown rats were
accidentally introduced in 1995 at the time of hotel
construction (Thorsen, et al., 2000; Merton, et al., 2002).
Management of the island's biodiversity was initially not
highly prioritised, but renewed interest in conservation was
created through increased awareness amongst stakeholders
over how nature can contribute to tourism. In 2003 the
hotel appointed its own environment staff, who conducted
biodiversity management and monitoring (J. Millett, pers.
obs., 2003).
Vertebrate eradications
The islands of Frégate, Denis and North were amongst
a series of islands that had rodents and /or cats eradicated
between 1982 and 2005. In total six islands over 10 ha in
size had black rats successfully removed, and three had
brown rats successfully removed (Rocamora & Henriette,
2015). Some of the eradications were implemented in
multi-island projects; however, the work was not planned
as a phased programme. On the three islands central to this
paper, the rodent eradications were, in part, motivated by
business interests, and two required second attempts. The
eradication of rodents has been an iterative (and at times
faltering) process but was ultimately successful on most
islands (Table 2).
Interest in eradicating rodents in the Seychelles was
stimulated in 1995, when the introduction of brown rats to
Frégate Island raised national and international concerns
over the impact on Seychelles magpie robins (Merton,
1996). A proposal, led by the Ministry of Environment and
Table 2 Black rat, brown rat, mouse and cat eradications in the Seychelles.
Island
Species
Bird
Size
(ha)
101
Frégate
219
Curieuse
286
Denis
143
Denis
143
North
201
D’Arros
150
Anonyme
North
10
201
Brown rat
Mice
Cats
Black rat
Cats
Black rat
Cats
Black rat
Mice
Black rat
Cats
Brown rat
Cats
Mice
Black rat
Black rat
582
Black rat
Invasion
date
1960s
Eradication
date
1995
1995
?
?
?
?
?
2001
?
~1784
?
?
?
?
?
?
Method
Outcome
Ground Application
Success
2000
2000
1982
2000
2000
2000
2000
2002
Aerial Application
Aerial Application
Ground Application
Aerial Application
Ground Application
Aerial Application
Ground Application
Ground Application
Success
Success
Success
Reinvaded Poor
Success
Reinvaded Poor/ None
Success
Success
Poor/Medium
2003
2003
2003
2003
2003
2003
2005
Aerial Application
Ground Application
Ground Application
Failed
Success
Success
Success
Failed
Success
Success
Ground Application
Aerial Application
Prevention
measures
Medium/
Good
Medium/
Good
Poor/Medium
Good
Medium/Poor
Medium/
Good
Millett, et al.: Fifteen years after rodent eradications in the Seychelles
Transport (MET), to eradicate rats and other mammals on
several islands in a combined operation led to the eventual
attempted eradication of black rats on Denis and Curieuse,
brown rats on Frégate, and cats on Denis and Curieuse.
House mice present on Frégate and Denis were not specific
targets of eradication , but they were eradicated from the
former during the operation.. The eradication operational
costs on the private islands (Denis and Frégate) were
financed by the island owners, and on Curieuse (state
owned) it was funded by a grant from the Dutch Trust
Fund (DTF) that also covered consultancy costs for the
three islands (Merton, et al., 2002; United Nations, 2002;
Rocamora & Henriette, 2015; John Nevill, pers. comm.
2018).
Rodent eradication operations commenced in June 2000
with two aerial applications of brodifacoum bait totalling
18 kg/ha applied to Denis with a nine day interval, and
three aerial applications (23 kg/ha) to Frégate at five and
25 day intervals (Merton, et al., 2002). Areas that could
not be covered through the aerial application, including
buildings, work yards and hydroponic green houses were
hand baited. The third application was in response to a
lactating female rat trapped in an agricultural plot after
the second application (J. Millett, pers. obs., 2000) after
which no further rats were observed and the eradication
on Frégate was successful. Cat eradication on Denis
proceeded one week after the second rat-bait application
using trapping and baiting with Compound 1080 (Merton,
et al., 2002). The last cat on Denis was killed 14 months
after the eradication started. On the same day the last cat
was killed, black rats were confirmed as being present
again and breeding on Denis (J. Millett, pers. obs., 2001).
It was not possible to conclude if the population arose from
survivors or reintroduction. However, given the short time
duration between eradication and discovery and better
understanding of factors influencing tropical island rodent
eradications, eradication survival is likely (Rocamora &
Henriette, 2015; Keitt, et al., 2015). Subsequently it was
discovered the eradication attempt on Curieuse had also
been unsuccessful (G. Climo, pers. comm., 2001), possibly
due to reinvasion and/or survival (Rocamora & Henriette,
2015).
The owners of Denis Island decided to undertake a
second eradication attempt to eradicate rats and mice.
This proceeded with a ground-based operation in 2002,
using brodifacoum poison in bait stations on a 40 m grid.
Monitoring indicated that rats were killed quickly, but
mice persisted for several weeks around the livestock
farm where alternative food sources were available (G.
Climo, pers. comm., 2002). Both species were eradicated
successfully within two months.
A black rat eradication was attempted on North Island
in 2003 with an aerial baiting operation using three aerial
applications of brodifacoum. In March 2004 black rats
were still present (G. Climo, pers. comm., 2004; Rocamora
& Henriette, 2015). Cats were eradicated successfully at
this time with a combination of poisoning with Compound
1080 and trapping. A second attempt to remove rats was
made in 2005 with four aerial applications and a grid of
bait stations on the whole plateau and in the vicinity of
housing (Climo & Rocamora, 2006). In response to a rat
being captured four days after the third application a fourth
application was conducted (Climo & Rocamora, 2006)
which ultimately resulted in the eradication of rats.
Not only mammals have proved to pose problems for
endemic island species: introduced Indian myna birds
(Acridotheres tristis) attack some native birds and compete
for nest sites with Seychelles magpie-robin (Burt, et al.,
2015; Feare, et al., 2017). An attempt to eradicate mynas
on Frégate in 2000–2003 by shooting was unsuccessful
(Millett, et al., 2005) but eradication succeeded using traps
in 2011 (Canning, 2011a). Eradication on Denis Island
using an avicide (Starlicide) and shooting commenced in
2000 but was unsuccessful. A subsequent attempt used
trapping with follow-up shooting, which succeeded in 2015
(Feare, et al., 2017). On North Island, in 2006, an attempt
to eradicate mynas with Starlicide was unsuccessful due
to difficulties importing a rifle to start the shooting phase
when poisoning had reduced the population to fewer than
100 birds; shooting was finally conducted in 2008-2009
but the population had recovered and was too numerous
to be effective (Rocamora & Henriette, 2015). It was
reattempted with a decoy trapping campaign from May
2016 to March 2017, followed by shooting. This reduced
the population to three individuals by June 2018 with the
eradication attempt ongoing (Havemann, pers. obs., 2018).
Overall experiences on Denis, North and Frégate
indicate that sustained trapping programmes using small
decoy traps located in areas frequented by foraging
mynas, followed by shooting with an experienced hunter,
is effective. Shooting as a standalone measure and using
avicides appear to create aversion and have not worked
well. Disruption or cessation of culling results in a
population recovery (Millett, et al., 2005; Feare, et al.,
2017).
BIOSECURITY
Biosecurity controls have been implemented on each
of the three islands since undertaking rodent eradication.
North Island has rigorous biosecurity with pre-departure
inspections of all cargo on Mahé, inspections on arrival,
the processing of cargo and baggage through a pest
containment room, and fumigation and permanent bait
stations locate close to landing areas, human habitation and
beaches (North Island, 2015). Frégate Island has a rodent
abatement protocol which includes cargo inspection and
controls as well as permanent rat bait stations (Rocamora,
2015). There is also a rodent-proof fence around the
harbour, made of steel mesh set in to the ground and topped
with a smooth metal strip. However, maintenance of the
structure has remained a challenge, especially where the
ends of the fence meet the water and are influenced by
wave action (J. Millett, pers. obs., 2018). Denis Island has
a rodent prevention protocol (GIF, 2015). The protocol is
focussed on rodent control with measures including baiting
on boats, baiting arrival points and contingency measures
to respond to an incursion: Denis Island still brings cargo
to the island using a beach-landing barge, which increases
reintroduction risk.
On all three islands, visiting vessels need to be in
possession of a rat-free certificate which is obtained after
a thorough check for rats on board prior to departure
from Mahé. All of the protocols have been implemented
voluntarily and devote greater effort to inspection and
containment of rodents on the islands and less on loading
controls at departure. Although all plans concentrate on
rodent prevention, they are likely to be effective at reducing
wider biosecurity risks.
FOREST REHABILITATION
Endemic birds rely on forest (Vega, 2005; Njoroge,
2002). Most other native vertebrates and invertebrates
are also forest dwelling species; some have been able to
adapt to gardens and plantations. Invertebrate densities and
diversity on foliage tended to be higher for native trees,
yielding greater food availability for species such as the
Seychelles warbler and Seychelles paradise-flycatcher
(Komdeur, 1991; Komdeur, 1992; Richardson, 2001; Hill,
2002). Accordingly, rehabilitation of native forest has been
583
Island invasives: scaling up to meet the challenge. Ch 3C Strategy: Outcomes
a prerequisite to restoring endemic bird populations. The
original vegetation on Frégate, Denis and North Islands is
uncertain but evidence from remnant species and vegetation
of similar, less modified islands suggests that the original
contained Pisonia grandis and other native coastal trees
including Thespesia populnea, Heritiera littoralis and
Calophyllum inophyllum (Hill, 2002). The objective of
rehabilitation has been to create habitat for native species,
not to recreate pre-human forest.
North Island
Vegetation rehabilitation started in 2001 with the
removal of invasive plant species such as Lantana camara,
planting of native species on the coastal plateaux (including
Terminalia catappa, Barringtonia asiatica, Heritiera
littoralis) and attempts to rehabilitate vegetation on the
hills by planting Pyrostria bibracteata, Dodonaea viscosa
and other robust native shrubs. By 2017, approximately
60 ha of the coastal plateau was a native-dominated forest
with T. catappa and C. inophyllum being the most abundant
species. The establishment of native species on the hills
has been slower and more labour intensive with <2 ha
restored. The current area of native-dominated woodland
is approximately 30% of the island's total surface area.
Denis
In 2001, approximately 20 ha of coconut plantation
that was naturally reverting to native forest dominated
by T. catappa, was cleared of coconuts and planted with
native tree species (Hill, 2002). In 2007–2008, 12.5 ha
were rehabilitated with the aim of creating habitat for
Seychelles paradise-flycatchers (Bristol, et al., 2009). The
rehabilitation involved removing coconut, Nephrolepis
biserrata fern and other introduced weeds and replanting
with tree species including Terminalia catappa, C.
inophyllum, Thespesia populnea, Cordia subcordata, B.
asiatica, Ficus lutea, Guettarda speciosa, Hernandia
nymphaeifolia, H. littoralis, Ochrosia oppositifolia,
Pandanus balfourii, P. grandis, Ficus reflexa, Hibiscus
tiliaceus and Morinda citrifolia. In 2013–14, a further 2.5
ha area was cleared of coconut and Casuarina equisetifolia
and replanted with C. inophyllum and Mimusops
sechellarum and ca. 18 ha of T. catappa woodland were
weeded. The current area of native-dominated woodland
is approximately 40 ha (Bristol, 2014), comprising 29% of
the island's total surface area.
Frégate
A small amount of native tree planting was undertaken
in the 1990s to benefit Seychelles magpie-robins and,
in 1998, the hotel development used mostly native tree
species for landscaping. A wilt disease caused by Fusarium
oxysporum (Boa & Kirendall, 2004) killed all the Sandragon
trees on Frégate in the early 2000s. Most of the sandragon
forest was on the hills and these areas were replanted with
native species, mostly Ficus reflexa, F. lutea, Premna
serratifolia and Tabernaemontana coffeoides. Some
further coastal areas were replanted with T. catappa and
Guettarda speciosa, which has resulted in approximately
30 ha of native-dominated forest, comprising 15% of
the island's total surface area. In addition, quite a lot of
non-native forest was under-planted with native species
but, unfortunately, the habitat rehabilitation is not well
documented.
ENDANGERED SPECIES RECOVERY AND
INTRODUCTIONS
The eradication of vertebrate predators on three islands
with a total area of 560 ha, and associated improvement
in forest habitat for native birds, has contributed to the
recovery of several endangered bird species by increases
in existing populations or by reintroducing populations
(Table 3).
North Island
In 2007, 25 Seychelles white-eyes were introduced to
North Island from Conception (Rocamora & HenriettePayet, 2008). The population established and in 2017 was
estimated at between 127 and 140 birds using direct census
of groups with colour-ringed and unringed birds (Pietersen,
2017).
Denis Island
Four bird species have been introduced; 47 Seychelles
fodies were translocated from Frégate in February 2004
(Bristol, 2005), and the current population is estimated
at 600 individuals (van de Crommenacker, pers. obs.,
Table 3 Species conservation outcomes – an estimate of the number of endangered birds on rehabilitated
private islands.
Frégate
Species
Island
Seychelles magpie145c
robin
Seychelles warbler
North
Island
0
0
Denis
Island
Other populations
76a
Cousin
46
Cousine
32
Aride
10
d
400
Cousin
320
Cousine
210
Aride
1,850
0
Conception 3,140g
Mahé
25-35h
La Digue
ca. 400
84a
0
600c
209e
0
Seychelles white200c, g
eye
Seychelles Paradise0
flycatcher
Seychelles fody
1,182f
134b
Cousin
Cousine
Aride
D’Arros
1,000
430
ca. 500
250
Percentage of Seychelles
population on Frégate,
Denis & North Islands.
72
21
50
21
35
Data: Birdlife International, 2017 except: a Bristol & Gamatis, 2017; b Pietersen, 2017; c van de Crommenacker, pers.
obs., 2017; d Lopera-Doblas, et al., 2015; e Gala, 2017; f Vega, 2005 ; g Rocamora & Henriette, 2015; h Rocamora,
pers. obs., 2017.
584
Millett, et al.: Fifteen years after rodent eradications in the Seychelles
2016) using a grid point count with 58 counting points
located every 150 m. Fifty-eight Seychelles warblers
were introduced from Cousin in 2004, and the most recent
population estimate is 400 individuals (Lopera-Doblas,
et al., 2015; van de Crommenacker, pers. obs., 2016).
Twenty Seychelles magpie-robins were introduced in June
2008, 16 from Frégate and four from Cousin (Burt, et al.,
2016), and the population in June 2017 was 76 individuals,
estimated by monitoring colour ringed birds (Bristol
& Gamatis, 2017). Twenty-three Seychelles paradiseflycatchers were introduced from La Digue in November
2008 and the population in June 2017 was 84 individuals,
surveyed in a direct count Bristol & Gamatis, 2017).
Frégate Island
The population of Seychelles magpie-robin prior to
conservation efforts was very small, with as few as 39 in
2000 (Burt, et al., 2016). Habitat for this territorial species
was limiting (López-Sepulcre, et al., 2010). Increased
habitat area and quality allowed the population to rise to 137
by 2015 (Burt, et al., 2016) and approximately 145 in 2017.
This was estimated by an ongoing programme to colourring as many birds as possible to allow identification in
the field. Then, by searching the whole island for presence
of birds and their nest locations, group associations and
behaviour, a territory map was constructed along with a
status list with the identity of all birds within each territory
(van de Crommenacker, pers. obs., 2017). The Seychelles
fody population was estimated to have a population of
1,182 using mark and re-sight methods (Vega, 2005).
Frégate was considered suitable to reintroduce
Seychelles white-eyes because of the abundant fruiting
trees, including the non-native cinnamon (Cinnamon
vernum). Reintroduction was undertaken between 2001
and 2003, with 37 birds from Conception (Henriette &
Rocamora, 2011). The most recent estimate is at least 200
individuals, based on point counts (van de Crommenacker,
pers. obs., 2017). The habitat suitability for Seychelles
warbler was investigated (Hammers & Richardson, 2011)
and found to be suitable. Accordingly, 59 individuals were
translocated from Cousin in 2011 (Wright, et al., 2014); the
population was estimated to be at least 209 individuals in
2017 (Gala, 2017).
Other species appear to have benefitted from island
rehabilitation, including the endemic beetle Polposipes
herculeanus which showed a dramatic decline between
1995 and 2000, probably due to rat predation, but appears
to have subsequently recovered (Lucking & Lucking,
1997; Canning, 2011b).
FACTORS THAT INFLUENCED ISLAND
REHABILITATION
The Seychelles endemic birds’ crisis in the 1970s
and 1980s resulted in interventions by international
organisations including the Royal Society for Protection of
Birds (RSPB), International Council for Bird Preservation
(ICBP), BirdLife International, and the Royal Society for
Nature Conservation (RSNC), initially by direct funding
and deployment of staff and later through the establishment
and support of the local NGOs Nature Seychelles and
Island Conservation Society (ICS). The investment, at
its height, contributed several hundred thousand British
pounds each year, and facilitated the involvement of
technical expertise from New Zealand in an advisory role.
The potential benefits of rodent eradication for tourism,
farming and nature inspired island owners to finance
eradications. At the same time a proactive approach was
taken by the Seychelles government which wanted to
promote eradication programmes and donor funding was
available. International funding to help finance eradications
and reintroduction operations was obtained by NGOs and
Government through Global Environment Facility, Dutch
Trust Fund and Fonds Français pour l’Environnement
Mondial, among others.
As such, an enabling policy context (where a national
biodiversity plan was in place and being implemented),
international support, private sector interest, motivators
with a “can do” approach, and finance all came together to
facilitate change. At the time, the risk of failure in tropical
rodent eradications was not estimated (Keitt, et al., 2015),
and was therefore not perceived as a constraint.
The results of island rehabilitation, in particular rodent
eradications, have not only been sustained but enhanced
by hotel businesses who value it as part of their tourism
product. Each island has a conservation manager and a small
team of conservation staff and volunteers who implement
biodiversity monitoring, biosecurity, habitat rehabilitation
and education and awareness activities, including activities
for hotel guests. These businesses have been able to access
funds to support conservation management including
directing Corporate Social Responsibility Tax (CSRT)
to conservation programmes, donations from clients and
paying volunteers. Each island has used independent
approaches and methods to sustain this work, and
cooperation improved when North Island, Denis Island
and Frégate Island began working in partnership with the
Green Islands Foundation (GIF), which was established
in 2005 with the objective of improving cooperation and
conservation work on islands. GIF has been able to assist
with coordinating conservation programmes, apply for
funds and manage projects on behalf of island conservation
programmes and, importantly, act as a representative and
advocate at national meetings related to the environment.
Limitations to biodiversity conservation on private
islands
Recent years have seen a cessation of rodent eradications
and species reintroductions, with no rodent eradications
undertaken since 2010 (Rocamora & Henriette, 2015).
International partners progressively reduced support to
the Seychelles before ceasing funding, mainly due to the
reduced threat to endemic birds and Seychelles being no
longer considered a low-income country. National policy
still supports island rehabilitation and species conservation
(Nevill, et al., 2014), but specific actions are not being
proactively promoted. Major donors’ priorities have
shifted and the national project portfolio is dominated
by climate change adaptation and energy (Programme
Coordination Unit, 2017). In principle, funding is available
for island rehabilitation, but it is not being requested by the
government.
Very few native animal species that are not birds
have been introduced to Frégate, Denis or North Islands.
Moreover, predator-free North Island’s potential to support
populations of endangered species has not yet been realised.
One of the reasons is that habitat rehabilitation requires a
number of years to produce a canopy-forming forest that is
suitable for endemic birds and the forest has only recently
become suitable for species such as Seychelles black
paradise flycatcher (Bristol, 2017). Moreover, consensus
building for species reintroduction takes time and may
be influenced by views that are not necessarily evidencebased or pro-conservation.
There are several inner and outer islands with
conservation potential for rehabilitation that have not been
subject to eradication of invasive predators. For example,
the inner island Félicité has potential for rehabilitation
(Hill, 2002) and is currently suitable for the reintroduction
of Seychelles paradise flycatchers, a species tolerant of
rats, (Bristol, 2017) and potentially more endemic species
585
Island invasives: scaling up to meet the challenge. Ch 3C Strategy: Outcomes
if cats and rodents were eradicated. Several outer islands
already have hotels that support conservation programmes
run by ICS and four islands are proposed as protected areas
(UNDP, 2016). Rodent and cat eradications, if feasible, are
likely to be beneficial for seabird populations and other
biodiversity and well as tourism (Millett, et al., 2016).
Globally, many island-based businesses have nature
orientated tourism, and many islands have undertaken rat
eradications. However, comparable examples, whereby
tourism businesses have undertaken, often with the support
of local NGOs, invasive species eradication, habitat
rehabilitation and endangered species reintroductions
have not been observed in other island regions such as the
Pacific or Caribbean. There may be reasons for this: for
example, islands may have complex traditional ownership,
species that are conservation priorities may not be suited
to smaller islands or eco-tourism may be less valued by
tourism sectors in other regions.
CONCLUSION AND RECOMMENDATIONS
The contribution of private islands to national
conservation objectives is substantial, with 560 ha of
predator-free land on three islands supporting nine
populations of five species of globally threatened birds.
The self-financing private sector, the enabling role of the
Seychelles government and the contribution played by
NGOs in facilitation, information exchange and advocacy
are important. Whilst the priorities and contributions of
international networks and some NGOs have focussed
on other areas of work, others, notably GIF and ICS, are
still attempting to increase networking and cooperation
between these and other islands.
Lessons and recommendations for future work are:
1.
Local NGOs should work more closely together
and with business to improve knowledge exchange,
build capacity, and enhance rehabilitation
programmes;
2.
Opportunities for the restoration of species on
predator-free islands should be taken to the full,
notably on North Island;
3.
Develop a shared biosecurity facility on Mahé to
reduce the risk of invasive species reintroduction;
4.
Promote mammal eradications and habitat
rehabilitation on suitable inner islands including
Félicité, the proposed protected areas in the outer
islands of the Seychelles;
5.
6.
7.
Government should translate national policies
including the NBSAP into implementation plans
for species and sites;
The allocation of resources for island rehabilitation
should be advocated by the government of
Seychelles to international donors for large-scale
national projects;
The approach adopted by Frégate, North and
Denis Islands should be promoted as good practice
internationally by organisations that facilitate
collaboration and information sharing between
small island states.
REFERENCES
Boa, E. and Kirendall, L. (2004). Sandragon Wilt Disease, Seychelles.
Unpublished report for Government of Seychelles, Mahé, Seychelles:
CABI.
BirdLife International. (2000). Threatened Birds of the World. Barcelona
and Cambridge, UK: Lynx Edicions and BirdLife International.
586
BirdLife International. (2017). ‘Country Profile: Seychelles’. <www.
birdlife.org/datazone/country/seychelles>. Accessed 25 July 2017.
Baker, B.H. (1963). ‘Geology and mineral resources of the Seychelles
Archipelago’. Memoir of the. Geological Survey Kenya 3: 1–140.
Bradley, J.T. (1940). The History of Seychelles., Victoria, Seychelles:
Clarion Press.
Bristol, R. (2005). ‘Conservation introductions of Seychelles fody and
warbler to Denis Island, Seychelles’. Re-introduction News 24: 35–36.
Bristol, R., Barra, A. and Rose, M. (2009). Denis Island Habitat
Rehabilitation Final Report. Unpublished report for Green Islands
Foundation, Mahé, Seychelles: Darwin Initiative Project 15-009.
Bristol, R. (2014). Denis Island Habitat Rehabilitation Plan 2015–2019.
Unpublished report for Green Islands Foundation, Mahé, Seychelles.
Bristol, R. (2017). Proposal to Translocate Seychelles Paradise
Flycatchers to Félicité and Curieuse Islands. Unpublished Report for
Government of Seychelles. Mahé, Seychelles: Darwin Initiative project
23-006.
Bristol R. and Gamatis I. (2017). Seychelles Paradise-flycatcher
Population Census on Denis Island, June 2017. Unpublished report for
Green Islands Foundation, Mahé, Seychelles.
Burger, J. and Gochfeld, M. (1994). ‘Predation and Effects of Humans
on Island-nesting Seabirds’. In: D.N. Nettleship, J. Burger, M. and
Gochfeld (eds) Seabirds on Islands: Threats, Cases Studies and Action
Plans, pp. 39–67. Cambridge: BirdLife International.
Burt, A.J., Gane, J., Olivier, I., Calabrese, L., de Groene, A., Liebrick,
T., Marx, D. and Shah, N. (2016). ‘The history, status and trends of
the endangered Seychelles magpie-robin Copsychus sechellarum’. Bird
Conservation International 1: 1–19.
Canning, G. (2011a). ‘Eradication of the invasive common myna
(Acridotheres tristis) from Frégate Island, Seychelles’. Phelsuma 19:
43–53.
Canning, G. (2011b). ‘Population assessment of the Frégate Island giant
tenebrionid beetle (Polposipus herculeanus)’. Phelsuma 19: 69–78.
Climo G. and Rocamora G. (2006). The Successful Eradication of Black
Rats from North Island (Seychelles) in September 2005 and Proposed
Strategies to Reduce the Risks of Reinvasions. Unpublished report.
Réhabilitation des Ecosystèmes Insulaires (FFEM) Project. Seychelles:
Island Conservation Society and North Island.
Currie, D., Bristol, R., Millett, J., Hill, M., Bristol, U., Parr, S.J. and Shah,
N.J. (2003). ‘Habitat requirements of the Seychelles black paradise
flycatcher Terpsiphone corvina: A re-evaluation of translocation
priorities’. Ibis 145: 624–636.
Feare, C.J., van der Woude, J., Greenwell, P., Edwards, H.A., Taylor, J.A.,
Larose, C.S., Ahlen, P.A., West, J., Chadwick, W., Pandey, S., Raines,
K., Garcia, F., Komdeur, J. and de Groene, A. (2017). Eradication of
common mynas Acridotheres tristis from Denis Island, Seychelles. Pest
Management Science 73(2): 295–304.
Gala, C. (2017). ‘Exploring the Effect of an Invasive Lepidopteran on an
Endemic Insectivorous Bird’. MSc thesis. University of East Anglia:
Faculty of Biological Sciences.
Gaymer, R., Blackman, R. A. A., Dawson, P. G., Penny, M. and Penny, C.
M. (1969). ‘The endemic birds of the Seychelles’. Ibis 111: 157–176.
Gerlach, J. (2006). Terrestrial and Freshwater Mollusca of the Seychelles
Islands. Leiden: Backhuys.
Green Island Foundation. (GIF) (2012): Denis Island Rodent Protocol.
Seychelles: Green Islands Foundation.
Harper, G. and Bunbury N. (2015). ‘Invasive rats on tropical islands: Their
population biology and impacts on native species’. Global Ecology and
Conservation 3: 607–627.
Hammers, M. and Richardson, D.S. (2011). Assessment of the Suitability
of Frégate Island for Seychelles Warblers Acrocephalus sechellensis
2011. Unpublished report. University of Groningen and University of
East Anglia.
Henriette, E. and Rocamora, G. (2011). ‘Comparative use of three
methods for estimating the population size of a transferred island
endemic: the endangered Seychelles white-eye Zosterops modestus’.
Ostrich: Journal of African Ornithology 82(2): 87–94.
Hill, M. J. (2002). ‘Biodiversity surveys and conservation potential of
inner Seychelles islands’. Atoll Research Bulletin 495: 1–272.
Hilton, G. M. and Cuthbert, R. J. (2010). ‘The catastrophic impact of
invasive mammalian predators on birds of the UK Overseas Territories:
a review and synthesis’ Ibis 152: 443–458.
Millett, et al.: Fifteen years after rodent eradications in the Seychelles
Kaiser-Bunbury, C.N., Traveset, A. and Hansen, D.M. (2010).
‘Conservation and restoration of plant–animal mutualisms on
oceanic islands’. Perspectives in Plant Ecology, Evolution and
Systematics 12(2): 131–143.
Keitt, B., Griffiths, R., Boudjelas, S., Broome, K., Cranwell, S. and
Millett J. (2015). ‘Best practice guidelines for rat eradication on tropical
islands’. Biological Conservation 185: 17–26.
Komdeur, J. (1991). ‘Cooperative Breeding in the Seychelles Warbler’.
PhD thesis. Cambridge: University of Cambridge.
Komdeur, J. (1992). ‘Importance of habitat saturation and territory quality
for evolution of cooperative breeding in the Seychelles warbler’. Nature
358: 493–495.
Komdeur, J. (2003). ‘Adaptations and maladaptations to island living in
the Seychelles warbler’. Ornithological Science 2(2): 79–88.
Lopera-Doblas, L., McClelland, S., van der Woude, J. and van de
Crommenacker, J. (2015). Seychelles Warbler Population Census on
Denis Island August–October 2015. Unpublished report. Seychelles:
Green Island Foundation.
López-Sepulcre, A., Kokko, H. and Norris, K. (2010). ‘Evolutionary
conservation advice for despotic populations: Habitat heterogeneity
favours conflict and reduced productivity in Seychelles magpie-robins’.
Proceedings of the Royal Society of London Series B 277: 3477–3482.
Lucking, R.S. and Ayrton, V. (1995). The Seychelles Magpie Robin
Recovery Plan. Aride Translocation: Summary Report, November
1992–November 1994. Unpublished report. Mahé, Seychelles: BirdLife
International.
Lucking, R. and Lucking, V. (1997). ‘Preliminary results of monthly
surveys of Polposipus herculeanus’. Phelsuma 5: 5.
Merton, D. (1996). Frégate Island Rat Crisis – Implications,
Recommendations, Eradication Plan, Indicative Budgets 1996 and 1997.
Republic of Seychelles: Unpublished report. BirdLife International and
Division of Environment, Ministry of Environment and Transport.
Merton, D., Climo, G., Laboudallon, V., Robert, S. and Mander, C. (2002).
‘Alien mammal eradication and quarantine on inhabited islands in the
Seychelles’. In: C.R. Veitch and M.N. Clout (eds.) Turning the tide:
the eradication of invasive species, pp. 182–198. Occasional Paper SSC
no. 28. IUCN SSC Invasive Species Specialist Group. IUCN, Gland,
Switzerland and Cambridge, UK.
Millett, J., Climo, G. and Shah, N.J. (2005). ‘Eradication of common
mynah Arcidotheres tristis population in the granitic Seychelles:
Successes, failures and lessons learned’. Advances in Vertebrate Pest
Management 3: 169–183.
Millett, J, Bristol, R., Feare, C., Duhec, A., Risi, M., Nogués, P, Morgan,
M. and Blumel, J. (2016). Seabird Management Plans for the Outer
Islands of the Seychelles. Unpublished report. Mahé, Seychelles: Island
Conservation Society.
Rocamora, G. and Henriette-Payet, E. (2008). Conservation introductions
of the Seychelles White-eye on predator-free rehabilitated islands of
the Seychelles archipelago, Indian Ocean. In: Soorae, P. (ed.) Global
Reintroduction Perspectives. Reintroduction case studies from around
the globe, pp. 121–129. Gland, Switerland: IUCN.
Rocamora G. (2015). Biosecurity Protocols for the Transportation of
Vessels, Cargo and People Between Islands, with Special Reference
to Protected Areas and Islands of High Biodiversity Value. Report
for project ‘Mainstreaming Prevention and Control Measures for IAS
into Trade, Transport and Travel across the Production Landscape’.
Seychelles: Government of Seychelles-UNDP-GEF.
Rocamora, G. and Henriette, E. (2015). Invasive Alien Species in
Seychelles: Why and How to Eliminate Them? Identification and
Management of Priority Species. Seychelles: Island Biodiversity &
Conservation centre, University of Seychelles; Mèze: Biotope Editions,
and Paris: Muséum National d’Histoire Naturelle (Inventaires &
Biodiversité series).
Russell, J.C. and Holmes, N.D. (2015). ‘Tropical island conservation: Rat
eradication for species recovery’. Biological Conservation 185: 1–7.
Russell, J.C. and Broome, K.G. (2016). ‘Fifty years of rodent eradications
in New Zealand: Another decade of advances’. New Zealand Journal of
Ecology 40(2): 197–204.
Samways, M.J., Hitchins, P.M., Bourquin, O. and Henwood, J. (2010).
Restoration of a tropical island: Cousine Island, Seychelles. Biodiversity
and Conservation 19: 425–434.
Stoddart, D.R. (1984). Biogeography and Ecology of the Seychelles
Islands. Monographiae Biologicae Volume 55. The Hague/Boston/
Lancaster: Dr. W. Junk Publishers.
Stoddart, D.R. and Fosberg, F.R. (1981). ‘Bird and Denis Islands,
Seychelles’. Atoll Research Bulletin 252: 1–46.
Tershy, B.R., Shen, K., Newton, K.M., Holmes, N.D. and Croll, D.A.
(2015). ‘The importance of islands for the protection of biological and
linguistic diversity’. Bioscience 20: 1–6.
Thibault, M., Brescia, 1F., Jourdan, H. and Vidal. E. (2016). ‘Invasive
rodents, an overlooked threat for skinks in a tropical island hotspot of
biodiversity’. New Zealand Journal of Ecology 41(1): 74–83.
Thorsen, M., Shorten, R., Lucking, R. and Lucking, V. (2000). ‘Norway
rats (Rattus norvegicus) on Frégate Island, Seychelles: The invasion;
subsequent eradication attempts and implications for the island's
fauna’. Biological Conservation 96: 133–138.
Towns, D.R. (1991). ‘Response of lizard assemblages in the Mercury
Islands, New Zealand, to removal of an introduced rodent: The kiore
(Rattus exulans). Journal of the Royal Society of New Zealand 21:
119–136.
United Nations (2002). ‘Seychelles Country Profile’. <www.un.org/esa/
agenda21/natlinfo/wssd/Seychelles.pdf>. Accessed 4 July 2017.
Moors, P.J. and Atkinson, I.A.E. (1984). ‘Predation on seabirds by
introduced animals, and factors affecting its severity’. ICBP Technical
Publication 2: 667–690.
UNDP. (2016). Seychelles’ Protected Areas Finance Project: Project
Document. GEF ID 5485 / Project ID 00095320. Seychelles: United
Nations Development Programme.
Nevill, J., Prescott, J., Shah, N.J. and Jeremie M. (2014). Seychelles
Biodiversity and Action Plan 2015 – 2020. Seychelles: Government of
Seychelles.
Vega, L.B. (2005). ‘The Ecology and Conservation of the Seychelles
Fody or Toc-toc Foudia sechellarum’. PhD thesis. UK: University of
Reading.
Njoroge, P. (2002). ‘The Seychelles Magpie-robin Copsychus
sechellarum: Territory Quality, Demography and Conservation of a
Threatened Species’. PhD thesis. UK: University of Reading.
Vesey-Fitzgerald, D. (1940). ‘The birds of the Seychelles 1. The endemic
birds’. Ibis 4: 482–489.
North Island (2015). Manual for the Conservation Work, Environmental
Management and Sustainability Efforts of North Island. Seychelles:
Wilderness Collection.
Pietersen, D.W. (2017). Assessment of the Seychelles White-eye, Zosterops
modestus Newton 1867, Population on North Island, Seychelles. North
Island Annual Report 2017. Seychelles: Wilderness Collection.
Programme Coordination Unit. (PCU) (2017). ‘Ongoing Projects’.
<http://www.pcusey.sc/index.php/pcu-projects/ongoing>. Accessed 4
July 2017.
Watson, J. (1978). The Seychelles Magpie Robin. WWF Project
1590: Endangered land birds, Seychelles. Final Report 1(a). Gland,
Switzerland: World Wildlife Fund. Mimeo.
Warman, S. and Todd, D. (1984). ‘A biological survey of Aride Island
Nature Reserve, Seychelles’. Biological Conservation 28: 51–71.
Wright, D.J., Shah, N.J. and Richardson, D.S. (2014). ‘Translocation of
the Seychelles warbler Acrocephalus sechellensis to establish a new
population on Frégate Island, Seychelles’. Conservation Evidence 11:
20–24.
Richardson, D. S. (2001). Species Conservation Assessment and Action
Plan: Seychelles Warbler (Timerl Dezil). Unpublished report. Norwich:
Nature Seychelles and University of East Anglia.
Rocamora, G. (1997). Rare and Threatened Species, Sites and Habitats
Monitoring Programme in Seychelles: Monitoring Methodologies
and Recommended Priority Actions. Unpublished Report. Seychelles:
Government of Seychelles.
587
G. Rocamora
Rocamora, G. Eradication of invasive animals and other island restoration practices in Seychelles: achievements, challenges and scaling up perspectives
Eradication of invasive animals and other island restoration practices in
Seychelles: achievements, challenges and scaling up perspectives
G. Rocamora
Scientific Director & Chair, Island Biodiversity & Conservation Centre, University of Seychelles, Anse Royale Campus,
Box 1348, Victoria, Seychelles. <IBC@unisey.ac.sc>.
Abstract In recent decades Seychelles has accumulated extensive experience in the management of invasive species
and other island restoration practices. Non-government organisations (NGOs), governmental, parastatal and private
stakeholders have conducted successful programmes to control and eradicate invasive animals and plants, particularly
on small islands of high biodiversity value. Biosecurity protocols have been implemented to prevent (re)infestations.
With at least 50 vertebrate populations (33 mammal, 16 bird and one reptile) from 14 different species successfully
eradicated, Seychelles is the third country in the world after Australia and the USA for invasive vertebrate eradications
from tropical islands, and the seventh when considering all countries. Twenty-four islands have benefited from invasive
vertebrate eradications and other ecosystem restoration processes to create refuges for native biodiversity. About 470
ha of woodland have been rehabilitated through replanting and recovery of native vegetation, and at least 36 successful
island translocations of native birds and reptiles have been conducted. This includes 16 conservation introductions or
reintroductions of six endemic land birds (all but one threatened), two of a terrapin species and 18 of Aldabra giant
tortoises. Recovery of native species and natural recolonisations have occurred on islands where invasive predators have
been removed. As a result, four globally threatened endemic land birds have been down-listed in the IUCN Red List and
dozens of other native species including seabirds, land birds, reptiles, invertebrates and plants have also benefited. Future
challenges include increasing the proportion of the country’s land area free of rats and cats from 3.9% to potentially
15.4%, mainly in the outer islands, and 50% in the long term if Aldabra and Cosmoledo are considered. Factors limiting
future eradications and translocations are discussed. Alternative conservation approaches such as ‘mainland-islands’ are
recommended for large islands, and the development of partnerships with nature-based tourism is encouraged to help
fund further restoration.
Keywords: ecosystem recovery, habitat rehabilitation, invasive birds, invasive mammals, reintroductions, species
recovery, species translocations, vegetation restoration
INTRODUCTION
MATERIALS AND METHODS
The Republic of Seychelles comprises 115 main islands
totalling 445 km2 of land area within a marine Exclusive
Economic Zone of 1,374,000 km2. These are classified
into the ‘inner islands’ archipelago, of granitic substrate
(ca. 45), and the remote, coralline ‘outer islands’ (c.70)
to the south and south-east, that include the Amirantes,
Providence-Farquhar and Aldabra groups (Fig 1). Aldabra
atoll, a nature reserve and World Heritage Site, represents
about one third of the country area.
Rocamora (2015) used information from publications
(Beaver & Mougal, 2009; Nevill, 2009), internal reports and
newsletters, and unpublished information from personal
knowledge, to construct a database recording all attempts
made in Seychelles to eradicate vertebrate populations.
For each eradication attempt the database records: island
name, area, animal species, year(s), methods used, and
the final outcome of the overall eradication programme
but not to the immediate result of each method employed.
This information was checked in 2014 and made consistent
with the Database of Island Invasive Species Eradications
(DIISE) managed by Island Conservation, to allow
comparison of Seychelles’ performance in eradication with
that of other countries (DIISE, 2017). This information
base was updated in 2017. No new eradication attempts
have taken place since 2014. One operation formerly
classified as control was re-classified as ‘eradication’ as the
target species, the crested-tree lizard (Calotes versicolor),
was eradicated. The status of the five operations that were
ongoing in 2015, and finalised by 2017, was updated.
In recent decades, the restoration of small islands has
been an effective conservation tool in Seychelles to create
sanctuaries for native biodiversity (Rocamora, 1997;
Nevill, 2001; Shah, 2001; Merton, et al., 2002; Shah,
2006; Asconit & ICS, 2010; Rocamora, 2010a; Samways,
et al., 2010b; Nevill, 2011). This has been achieved
by eradicating or controlling invasive alien predators
and competitors. Native habitats have been restored by
eliminating invasive alien plants and replanting native
vegetation. Globally threatened species of endemic birds
and other native wildlife have been translocated to these
rehabilitated islands, contributing to their subsequent
recovery (Kömdeur & Pels, 2005; Richardson, et al.,
2006; Rocamora & Henriette-Payet, 2008; Shah, 2008).
This paper updates the inventory of island restoration
achievements in Seychelles documented in Rocamora
(2015) and discusses future perspectives and challenges. It
considers only actions for nature conservation purposes and
excludes eradications of invasive species for agricultural
(two declared; National Biosecurity Agency, pers. comm.
2017) or public health purposes. Names of islands are
the ones normally used by the islanders as listed in the
Constitution of Seychelles, with the exception of ‘Ile du
Nord’ also referred to as ‘North Island’.
Rocamora (2015) also gathered information on the
area of natural habitat rehabilitated (from reports, or
estimates by island owners, managers or conservation
staff) as a result of removing invasive plants and
propagating and planting native vegetation. He also
documented translocations of native species that occurred
in Seychelles, i.e. reintroductions to islands where the
species was formerly present, ‘conservation introductions’
to islands outside a species’ known historical range (IUCN,
2013), and historical introductions or reintroductions of
Aldabra giant tortoises (Aldabrachelys gigantea). For each
translocation he recorded: species, island, year, type of
translocation (reintroduction, conservation introduction),
and outcome. He then analysed how island restoration
practice has developed in Seychelles, together with nature-
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
588
up to meet the challenge, pp. 588–599. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Rocamora: Island restoration practices in Seychelles
based tourism, and how this has benefited the conservation
of native biodiversity.
RESULTS
Eradication of introduced predators and competitors:
the first step to ecosystem recovery
By September 2017, 68 attempts to eradicate invasive
animals had been made on 24 islands. Three of these
operations are still in progress (Table 1).
Most island eradications conducted in Seychelles
have targeted mammals (44 attempts, 68%) and birds
(22 attempts, 29%) on 22 islands of at least 10 ha, plus
two mammal eradications on two islets smaller than 1 ha.
In five of the 65 completed eradications, species that
were not the main target (feral cats and barn owls) also
disappeared following the removal of rats and on three
occasions island populations of feral goats and chickens
died out following control. Of the remaining 55 eradication
attempts completed (excluding the two small islets), 40
(72.7%) succeeded and 15 (27.3%) had a failed outcome
(i.e. survival or recolonisation before the island could be
certified pest-free; see DIISE, 2017). When including the
rat eradications on two islets (< 1ha), success rate is 73.7%
and failure 26.3% (n=57).
By the end of 2017, 50 alien vertebrate (33 mammal, 16
bird and one reptile) populations had been eradicated from
islands in Seychelles. One operation targeting common
myna (Acridotheres tristis) on Ile du Nord is almost finished
and one (the ring-necked parakeet, Psittacula krameri, on
Mahé) is in the final phase of monitoring. Fig. 2 lists the 14
species of vertebrates eradicated from islands in Seychelles
(plus one yet to be confirmed) and gives the outcomes of
eradication attempts. Success rates of eradication attempts
vary: 33% for house mouse, 56% for common myna, 57%
for feral goat, 75% for black rat (Rattus rattus) and brown
rat (R. norvegicus) (excluding the two islets), and 100% for
other species.
Domestic pigs (Sus scrofa) and cows (Bos taurus) were
also removed from Cousine and Ile du Nord. These were
small numbers: some of the animals were not completely
wild and may have still depended on supplementary food
from humans, so they were easy to catch, and it is unclear
if some of them were reproducing in the wild (Samways, et
al., 2010a; Bruce Simpson/North Island Ltd; pers. comm.;
Victorin Laboudallon, pers. comm. 2015). Feral cats (Felis
catus) died out on Picard after the 1970s, with no control or
eradication programme involved. These pig, cow and cat
cases are not included in the calculations as eradications,
but we did include the reported eradication of goats on
Aride by shooting before 1920 (Warman & Todd, 1984),
the removal of feral goats from the Aldabra atoll islands
of Polymnie and Ile Esprit in the 1970s in response to
localised control (Nancy Bunbury/SIF, pers. comm.), and
the extinction of feral chickens on Desnoeufs in 2007 by
local staff for consumption (Roland Nolin, pers. comm.).
The numbers of eradication attempts and success rates
have varied over time (Table 2): 73% (over 11 attempts)
before 1995, 64% (over 28 attempts) during 1995–2004,
and after 2004 up to 89% (over 18 attempts finalised by
2017).
Removal of invasive alien plants and replanting native
vegetation
Control of invasive plants and habitat rehabilitation has
been important for restoring ecosystems and protecting
native biodiversity (Table 3). Significant areas (over 60
ha) have been rehabilitated since the 1990s on Praslin
(National Park) and Mahé (Morne Seychellois National
Park). Most invasive plant control and native species
replanting activities have taken place on small and medium
sized granitic islands, as part of programmes to restore
Fig. 1 Islands of the Republic of Seychelles.
Fig. 2 Number of eradication attempts (n = 68), and
success outcomes for the 15 species of invasive
vertebrates targeted in Seychelles. Species are listed
in increasing order of successful attempts. Success or
failure refers to the final outcome of the operations (see
text).
589
(1 from
control)
_
_
1
(1 from
control)
Red-whiskered
bulbul
Pycnonotus jocosus
Madagascar fody
Foudia
madagascariensis
Total attempts
Success (Outcome)
1
_
House mouse
Mus musculus
_
_
_
_
_
Brown rat
Rattus norvegicus
_
_
_
_
_
(died out
by 1976
after
control)
(died out
by 1976
after
control)
_
51
475
Black rat
Rattus rattus
Feral chicken
Gallus gallus
Feral cat
Felis catus
Feral goat
Capra hircus
Area (ha)
Outer Islands
Ile
chronological order Polymnie, Esprit,
of eradications
Aldabra Aldabra
1
1
_
_
_
_
_
_
(died out
after1970s)
1993–95
940
Picard,
Aldabra
1
2
_
_
_
_
_
_
_
1993–95
1987–88
failed
2,680
2
3
_
_
2003
failed
2003
_
_
2003
_
140
2
1 (+ 1
indirect)
(1
harvested
for food)
_
_
_
_
2007
_
(2008)
_
143
1
_
_
_
_
_
(died out
by 2007
after
harvest)
_
_
35
1 (+ 1
indirect)
2
_
_
_
_
2007
_
(2007)
_
21
1
1
_
_
_
_
2007
_
_
_
<1
2
4
2012–15
_
(inc. 2012–14)
_
_
_
_
_
2007–12
93–97 failed
87–88 failed
11,610
2
2
2012–15
2012–14
_
_
_
_
_
_
1,171
Grand
Malabar,
Grande Ile,
Polyte, Petit Polyte, Grande Terre,
Aldabra D'Arros Desnoeufs Cosmoledo Cosmoledo Cosmoledo
Aldabra
Assomption
20
2200
2
1
1
1
3
1
3
8
Total
attempts
590
11 (+ 2
indirect & 3
after control/
harvest)
2
1
0
1
3
(+harvest)
1 (+ 2
indirect)
3 (+ 2 control)
Successful
outcomes
Table 1. Attempts to eradicate invasive alien animals from islands in Seychelles. Total number of attempts = 68 (65 completed; 3 in progress). Successful attempts (in bold) = 50 (42
direct; + 5 indirect eradications induced by rat removal and 3 from control, in brackets). Success or failure refers to the outcome of the operation. ‘Failed’ attempts involved animals
that either survived or immediately reinvaded after a technically successful eradication phase. Only the active removal of individuals that had established a substantial wild breeding
population is considered a genuine eradication attempt. Removal of small numbers of wild or semi-feral domestic animals (cattle, pigs, chickens), or populations of the same that
died out naturally are not included here. * = eradications almost finished or in monitoring phase, to be confirmed in 2018. Occ. inc. = occasional incursions. Species and islands are
listed in chronological order of their first eradication.
Island invasives: scaling up to meet the challenge. Ch 3C Strategy: Outcomes
_
before
1920
1930s
Feral goat
Capra hircus
Feral cat
Felis catus
Success (Outcome)
Total attempts
2
2 (+ 1
ongoing)
4 (+ 1
ongoing)
4
_
2012–
ongoing*
_
2014–
ongoing*
_
Ring-necked
parakeet
Psittacula krameri
Black-headed ant
Pheidole
megacephala
_
Crested-tree lizard
Calotes versicolor
2003–04
+occ. inc.
_
_
_
_
_
_
House sparrow
Passer domesticus
House mouse
Mus musculus
Brown rat
Rattus norvegicus
Black rat
Rattus rattus
_
_
1996
+occ. inc.
Barn owl
Tyto alba
_
_
Feral rabbit
Oryctolagus
cuniculus
_
_
_
1993–94
+occ. inc.
Common myna
Acridotheres tristis
1977–94
occ. inc.
Feral chicken
Gallus gallus
_
House crow
Corvus splendens
_
15252
73
Area (ha)
Mahé
Aride
chronological
order of
eradications
Inner Islands
_
1983–85
_
26
Cousine
4
6
_
_
_
_
3
3
(control
2008)
_
_
_
_
_
2000
1996 failed
2000
_
_
_
1996
_
_
_
_
2010–11
(98–02 failed) 2001–02
+occ. inc.
+occ. inc.
_
1981–82
_
219
Frégate
2
3
_
_
_
_
1996–97
failed
_
1996–97
_
1996–97
_
_
_
_
_
101
Bird
Island
1
1
_
_
_
_
_
_
_
_
_
_
2000–02
+occ. inc.
_
_
_
29
1
3
_
_
_
_
2000
failed
4
7
_
_
_
1
1
_
2
2
_
_
_
2003–
14
_
_
_
1
1
_
_
_
_
_
_
2006
2003 &
2006
_
_
_
_
_
_
_
_
<1
_
_
_
_
_
_
_
10
_
_
2002
2000
failed
_
_
_
_
_
_
2002
2000failed
2000
failed
_
_
_
_
_
2010–15
2000–01
failed
_
_
_
_
219
1
1
3 (+ 1
indirect)
_
_
_
_
_
2007
_
_
_
_
_
_
_
_
69
6 (+ 1
ongoing)
_
_
_
_
_
_
2005
2003 failed
(2005)
occ. inc.
2003
2006–09
failed. 2012–
ongoing*
2003
_
201
1 (+ 2
indirect)
4
_
_
_
_
_
_
2010
(2010)
occ. incv.
2011
failed
(2010)
_
85
1
1
_
_
_
_
_
_
2010
_
_
_
_
_
_
_
35
Sainte
Ile aux North Island
Grande Petite
Anne Anonyme rats (Ile du Nord) Conception Sœur Sœur
_
2000
_
143
Denis
_
_
_
_
_
2000
_
289
Cousin Curieuse
Total
attempts
45 (+ 3
ongoing)
(1
ongoing)
(1
ongoing)
1
1
5
3
11
3
1
2
9 (+1
ongoing)
1
7
1
1
Successful
outcomes
31 (+ 3
indirect)
1
1
2
2
8
1 (+ 2
indirect)
1
2
5
1
6 (+ 1
indirect)
Rocamora: Island restoration practices in Seychelles
591
Island invasives: scaling up to meet the challenge. Ch 3C Strategy: Outcomes
Table 2 Temporal distribution and success outcomes of
attempts to eradicate invasive animals in Seychelles (n
= 68). The few eradications declared successful after
2015 had their last individuals eliminated on this year or
before.
Pre1995
Direct eradications
8
Indirect eradications
2
Failed outcomes
3
Ongoing attempts
Total attempts
13
1995– 2005– Total
2004 2015 attempts
18
16
42
6
8
10
2
15
3
3
28
27
68
Table 3 Approximate areas rehabilitated (see explanation
in text) on islands where vegetation management
(replanting and spontaneous native woodland recovery
after exotic species removal) has been undertaken.
Frégate
Ile du Nord
Félicité
Denis
Praslin
Mahé
Curieuse
Aride
Cousin
Cousine
Conception
Bird (Island
aux Vaches)
Area of
Area of
Planting
Woodland /
after exotic Shrubland
sp. removal rehabilitation
(ha)
(ha)
60
45–50
40
35
25 (inc. 10
20
bare land)
15–20
18–20
7
?
10
2 (coconut removal)
Area of
Woodland
recovery
(Pisonia
dominated)
(ha)
4
62
27
16
>1
<1
St Anne
1
Silhouette
<1
Moyenne
0.5
Anonyme
0.5
Inner Islands ca. 228
Desroches
12
D'Arros
11
Aldabra
(sisal
removal)
Alphonse
1
Outer Islands ca. 72
20
<5
35
ca. 145
-
<5
-
TOTAL
25
ca. 145
592
ca. 300
abandoned coconut plantations and lowland coastal forests
previously dominated by invasive species.
Habitat rehabilitation was initiated on Aride and Cousin
in the 1970s (Warman & Todd, 1984; Kömdeur & Pels,
2005), and since the mid-1990s has been implemented
on Frégate, Ile du Nord, Denis, Curieuse, Cousine and
Félicité as well as to a minor extent on other granitic
islands. Very little vegetation restoration has taken place
in the outer islands. On D’Arros, some 11 ha of former
coconut plantations have progressively been replaced
by plantations of native trees since 2009 (von Brandis,
2012; von Brandis, pers. comm. 2015). On Alphonse and
Desroches, since 2006 and 2009 respectively, small areas
have been cleared of exotics and replanted. At Aldabra,
rehabilitation activities to control sisal (Agave sisalana)
have taken place since the 1970s on Picard, Polymnie and
Ile Michel and, since 2013, to eradicate it (van Dinther, et
al., 2015).
In Seychelles, control and clearing of exotic plants
has mostly been done physically, using machetes and
chainsaws for woody plants, pulling by hand for creepers,
and sometimes using heavy machinery, as on Frégate or
Ile du Nord. Chemical treatments have rarely been used to
eradicate invasive plants, although some trials have been
conducted on several islands (Kaiser-Bunbury, et al., 2015).
Elimination of coconut trees inland of the beach crest has
been done on most of the rehabilitated islands that had, in
the past, been exploited as coconut plantations. These have
been replaced by forests dominated by native and endemic
species, through natural regrowth or replanting.
Although precise figures are not available for all
islands, we estimate that at least 220 ha have been actively
cleared of alien invasives, replanted with native trees and
maintained. This reaches 300 ha when including areas
partially restored and ca. 470 ha when including natural
recovery of native woodland. Rehabilitated vegetation now
covers 17% (405/2,480 ha) of middle-sized and small inner
islands, but only a tiny proportion (c.1%) of the country
area.
Nurseries were established by successive ministries
and associated public authorities responsible for the
environment on Mahé and Praslin. On private islands such
as Frégate (early 1990s), Ile du Nord (early 2000s) and
Félicité (early 2010s), nurseries dedicated to propagating
Seychelles native plants and trees have been created. These
have successfully multiplied most of the 85 endemic plants
of Seychelles and have produced tens of thousands of
saplings that have been used in island rehabilitation. Based
on the areas rehabilitated at the average density of 1,000
plants/ha normally used in Seychelles (Kueffer & Vos,
2004), we estimate that a minimum of 220,000 native trees
have been planted in Seychelles over the last 50 years.
Species translocations to rehabilitated islands
Species translocations to predator-free islands with
suitable habitats also contribute to the process of island
restoration. Table 4 lists 20 documented translocations of
eight rare and threatened species and one common species
that have taken place to date on ten rehabilitated islands.
This includes six species of Seychelles endemic land birds,
one species of reptile and one very rare insect. Ninety
percent of these translocations were successful (including
nine reintroductions and 10 conservation introductions
of rare and threatened species). In addition, a common
land bird was successfully transferred to Bird Island
(Ile aux Vaches). The two translocations which failed
were of Seychelles leaf-insects, Phyllum bioculatum,
to Conception and of Seychelles white-eyes, Zosterops
NT
EN
(VU)
(LC)
CR
(CR)
(LC)
Seychelles warbler
Acrocephalus sechellensis
Seychelles magpie-robin
Copsychus sechellarum
Seychelles white-eye
Zosterops modestus
Aldabra rail
Dryolimnas (cuvieri) aldabranus
Sey. black paradise flycatcher
Terpsiphone corvina
Seychelles black-mud terrapin
Pelusios subniger parietalis§
Seychelles leaf insect
Phyllum bioculatum
LC
Number of translocations per island (all
native species)
Seychelles sunbird
Cynniris dussumieri
Number of translocations of rare and
threatened species
NT
Seychelles fody
Foudia sechellarum
IUCN
Threat
Status
1
1
1994–95
X
X
Cousin
-
1990
X
3
4
2012
2
3 (1
failed)
2007
2002* 1995–96
1988
2002
Aride Cousine
©
©
2
2
2001
X
2011
-
Frégate
©®
4
4
2008
2008
2004
2004
Denis
©®
2
2
2008
2007
-
Ile du
Nord
©®
0
1 (failed)
2010
X
-
Conception
®
1
1
1968
1
1
1999
-
1
2006
0
Picard
D'Arros (Aldabra) Bird
®©
©
®
20 (2
failed)
1
19 (2
failed)
1
2
1
1
3
4
4
3
Total
Table 4 Translocations of rare and threatened species (other than Aldabra tortoises) and common species to rehabilitated islands in Seychelles. X = naturally present; Normal
case: reintroduction; Underlined: conservation introduction; Italic: failed attempt; *: initial transfer trials to Aride between 1978 and 1995 were unsuccessful; ©: cats eradicated;
®: rats eradicated; species are listed per chronological order of translocation; IUCN threat status between brackets likely to change due to status or taxonomic re-assessment;
§ endemicity questioned by Fritz, et al. (2013).
Rocamora: Island restoration practices in Seychelles
593
Island invasives: scaling up to meet the challenge. Ch 3C Strategy: Outcomes
modestus, to Cousine (Galman, 2011; Julie Gane / Cousine
Island, pers. comm. 2015). The islands with the highest
number of successful translocations are Denis (four) and
Aride (three), then Cousine, Frégate and Ile du Nord (two),
all other islands having benefited from only one species
translocation.
Outer Islands
Grande Ile (Cosmoledo)
D'Arros
St Joseph atoll*
Bancs du Sud (Providence)*
Marie-Louise*
Desnoeufs*
Ile du Sud-Ouest
(Cosmoledo)*
Bancs Africains*
Goëlettes (Farquhar)*
Grand Polyte (Cosmoledo)
St Francois (Alphonse)*
Ile du Nord (Cosmoledo)*
Ile du Nord-Est (Cosmoledo)*
Banc de Sable (Farquhar)*
Pagode (Cosmoledo)*
Goëlettes (Cosmoledo)*
594
219
201
143
101
85
73
69
35
29
26
20
10
9
5
2
1
143
140
122
Cats
eradicated
Inner Islands
Frégate
North Island (Ile du Nord)
Denis
Bird Island (Ile aux Vaches)
Grande Sœur
Aride
Conception
Petite Sœur
Cousin*
Cousine
Ile aux Récifs*
Anonyme
Mamelles*
Ile aux Vaches Marines*
Ile aux Cocos*
Ile aux Rats
Rats
eradicated
Islands free of rats and cats
Area (ha)
Table 5 Main islands free of rats and cats in Seychelles.
Islands are listed in order of decreasing size. Small
islands of less than 10 ha in italics. Islands naturally free
of rats and cats are also marked by *. Note: Conception
was found to have been recolonised by rats in late 2017,
while writing up this paper.
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Including the reintroductions of giant tortoises, the
total number of successful species translocations between
islands in Seychelles is 36.
DISCUSSION
With 50 island populations of invasive vertebrates (of
14 species) eradicated from islands, Seychelles stands as a
world leader. In 2014, it was ranking third after Australia
and the USA for tropical islands, and seventh when all
islands are considered (DIISE, 2017; Rocamora, 2015).
Despite more eradication attempts during the period
1995–2004, a lower success rate (64%) was recorded
compared to the following decade (89%). This may be a
result of improving project selection, field implementation,
and post-eradication biosecurity measures to prevent
reinvasions.
Global conservation impacts of Seychelles island
restoration
X
X
X
X
X
X
X
71
53
35
30
31
25
21
17
11
9
7
6
5
Translocations of Aldabra giant tortoises (IUCN Red
List category ‘Vulnerable’) have been accounted for
separately as many are ancient and/or poorly documented
(dates uncertain; possible failures not accounted for). After
the giant tortoises naturally present on most of the granitic
islands had been overexploited and driven to extinction
(Fauvel, 1909), 18 successful translocations of Aldabra
giant tortoises have taken place. Eight granitic islands
have been repopulated since 1850, including Frégate,
Curieuse and Cousin where they were reintroduced
before 1950, Moyenne (probably in the 1970s), and Ile
du Nord, Cousine, Grande Sœur and Silhouette that were
last repopulated during the period 1993–2012 (Gerlach,
et al., 2013). Aldabra giant tortoises have also been
introduced or reintroduced to 10 coralline islands (Bird,
Denis, D’Arros, Desroches, Rémire, Alphonse, Farquhar,
Providence, Assomption, Cosmoledo) during the past
25–50 years, although some of these populations are small
and of uncertain long-term viability (Gerlach, et al., 2013).
Four of these translocations were to rat and cat free islands
(Cousin, Cousine, Frégate and Bird). The reintroduction
of giant tortoises to Aride in 1933–34 is not counted, as
the animals were removed in 1951 and brought to Cousin
(Warman & Todd, 1984); however, some are planned to be
reintroduced from Frégate Island in 2018.
Population translocations to islands that have benefited
from predator eradications and habitat rehabilitation have
improved the conservation status of endemic species
threatened with global extinction in Seychelles (Henriette,
2011; Nevill, 2011; Russell, et al., 2016). Island restoration
has allowed 17 successful reintroductions or conservation
introductions of eight rare and threatened species and
the down-listing of four globally threatened birds on the
IUCN Red List: the Seychelles warbler (Acrocephalus
sechellensis) from Critically Endangered to Near
Threatened; the Seychelles magpie-robin (Copsychus
sechellarum) and Seychelles white-eye (Zosterops
modestus) from Critically Endangered to Endangered and
Vulnerable respectively; and the Seychelles fody (Foudia
sechellarum) from Vulnerable to Near Threatened.
The Seychelles black paradise flycatcher (Terpsiphone
corvina), which was transferred to Denis Island, is still
considered Critically Endangered. The Aldabra giant
tortoise (Vulnerable) has also benefited from 18 successful
translocations (Gerlach, et al., 2013).
Ecosystem recovery
The recovery of native fauna and flora on rehabilitated
islands where introduced predators and competitors have
been eradicated has already been observed on many
islands around the world (Mulder, et al., 2011; Veitch, et
Rocamora: Island restoration practices in Seychelles
al., 2011; Russell & Holmes, 2015). This is also occurring
in Seychelles, where monitoring of birds, reptiles,
invertebrates and plants has been undertaken and casual
observations collected (Rocamora & Henriette, 2015). After
the eradication of introduced predators and competitors,
some species that had become inconspicuous started to
reappear (e.g. giant millipedes Sechelleptus sechellarum
and endemic snails Stylodonta unidendata on Conception;
Galman, 2011). Five species of seabirds (Ardenna pacifica,
Gygis alba, Anous tenuirostris, Phaethon lepturus and
Sula dactylatra) have (re)established nine new breeding
populations on seven rehabilitated islands. Populations that
already existed have increased, as observed on most other
islands where invasive mammals have been eradicated
around the world (Brooke, et al., 2017). Reptiles and land
birds have typically shown increasing or stable trends and
some (e.g. Seychelles blue pigeon Alectroenas pulcherrima,
common moorhen Gallinula chloropus) have recolonised
islands, whereas invertebrates showed mixed responses,
including strong decreases for some groups. This is
probably linked to the increase in native land birds, reptiles
and large invertebrates that had previously been preyed
upon by rats and cats (Galman, 2011). As part of a global
study to demonstrate the impact of mammal eradications
on native wildlife (Jones, et al., 2016), 67 populations of
26 native vertebrates (13 land birds, eight seabirds, five
reptiles) were identified as having benefited from these
operations in Seychelles (Rocamora & Henriette, 2015).
This illustrates how important it is to undertake ecosystem
monitoring before and after eradications to measure and
understand the ecological changes that occur on islands
under rehabilitation.
The reintroduction of giant tortoises, which dominated
the terrestrial ecosystems of Seychelles for millions of
years, is an essential step in the island restoration process.
These animals fill an important (but still poorly known)
role in the ecosystem by dispersing and promoting
the germination of seeds, fertilising native plants, and
influencing soil invertebrate communities through their
dung. These mega-herbivores are used as ecological
analogues to replace extinct tortoises and help restore
island ecosystems in Mauritius and Rodrigues (Griffiths,
et al., 2010; Hansen, et al., 2010; Griffiths, 2014). The
(re)introduction of Seychelles white-eyes (Rocamora &
Henriette-Payet, 2008), which disseminate many native
berry-producing trees, also contributes to the restoration
process. Future challenges include a more integrated
‘ecosystem approach’, aiming at rehabilitating entire
habitats and communities (including invertebrates), rather
than focusing only on ‘flagship’ threatened species (Asconit
& ICS, 2010; Kaiser-Bunbury, et al., 2010; Galman, 2011).
Seabirds play a critical role in ecosystem recovery as
they boost soil nutrients thereby assisting the development
of the ground microfauna (Mulder, et al., 2011). Seabird
recolonisation can be slow, although decoys and sound
recordings to attract passing adults can speed this process
(Jones & Kress, 2012). In Seychelles, this has only been
done for the sooty tern (Onychoprion fuscatus) on Denis
Island, with little success (Feare, et al., 2015). Seabird
translocations may also be tried in future in Seychelles; this
technique has been employed successfully in the Pacific
(Kappes & Jones, 2014) and is being trialled in Mauritius
(Carl Jones & Nik Cole, pers. comm., 2016).
Scaling up eradication projects and increasing the
rat and cat free area of Seychelles to create more
biodiversity refuges
Since the 1970s, ecosystem restoration has taken place
on 25 small and medium sized islands of Seychelles (i.e. <
2,000 ha; see Fig. 2). As a result, island refuges for native
biodiversity and particularly for rare species threatened
with extinction have multiplied. This process was started
on NGO-owned islands in the 1970s, then followed on
government and privately-owned islands, the public
trust Seychelles Island Foundation and more recently
on government islands managed by the parastatal Island
Development Company and associated partners (private
hotel/villa owners and the Island Conservation Society).
With about 30 small and medium-sized islands free
of rats and cats (see Table 5), Seychelles probably has
proportionally more territory (3.9%) free of invasive
predatory mammals than most island countries. Rats and
cats have been removed from 11 islands larger than 10
ha. Between 1996 and 2011, the number of islands of ≥10
ha free of rats has increased from four (Aride, Cousin,
Cousine, Ile aux Récifs) to 12, and the total rat-free area
of Seychelles has more than tripled from 581 ha to 1,757
ha (Fig. 3).
Nevertheless, there is scope for more eradications
to benefit both wildlife and humans. This would require
scaling up the size of islands tackled for eradications. Rats
Table 6 Additional islands in Seychelles where rodents and
cats could be eradicated and their reinvasion prevented
with currently available techniques. X = presence.
Moyenne, Longue and Ronde (Ste Anne group) would
require a combined operation owing to their mutual
proximity.
Area Black House
(ha) rat
mouse
Inner Islands
Curieuse
Félicité
Marianne
Bird (Ile aux Vaches)
Aride
Ronde (Praslin)
Thérèse
Longue
Moyenne
Ronde (Mahé)
Outer Islands
Assomption
Coétivy
Astove
Ile du Sud (Farquhar)
Desroches
Ile du Nord
(Farquhar)
Poivre
Alphonse
Providence
D'Arros
Platte
Rémire
Manahas (Farquhar)
Marie-Louise
Desnoeufs
Feral
cat
286
230
100
101
73
19
74
17
9
2
X
X
X
?
1,171
931
660
400
394
X
X
X
X
X
X
X
X
X
X
300
255
174
157
140
54
27
10
53
35
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
?
X
X
X
X
X
X
X
X
X
X
X
595
Island invasives: scaling up to meet the challenge. Ch 3C Strategy: Outcomes
and cats could be removed from another 22 islands with
currently available techniques (Table 6). This includes five
more granitic (inner) islands, plus three small islands in the
Ste Anne group which would require a permanent grid of
rat bait stations owing to their proximity to other infested
islands.
The outer islands have greater restoration potential,
with 14 islands where rats (and cats) could be eradicated.
However, reinvasion may be difficult to prevent
through strict biosecurity on three of the larger islands
(Coëtivy, Assomption, Desroches) depending on future
developments envisaged. The maximum potential area that
could be cleared of rats and cats is currently ca. 7,000 ha or
15.4% of the total area of the country (see Fig. 3 for inner
and outer islands totals). Clearing rats and cats from these
additional 19–22 islands would open huge possibilities
for ecosystem rehabilitation and population recovery of
many species (land birds, seabirds, reptiles, amphibians,
invertebrates and plants). The eradication of rats from
Aldabra and Cosmoledo (Menai Island) atolls, which have
large extensions of mangroves, is not considered currently
feasible. However, this may change if new techniques
become available in future, in which case up to half
(49.7%) of the country area could be made rat and cat free
in the long term.
Eradication operations should now be extended to
invertebrates such as invasive ants, moths or snails that
also have a high negative impact on native biodiversity.
However, the ability to apply permanent biosecurity
protocols will be critical for these islands to retain their
pest free status. In the low-lying coralline outer islands,
options for translocation of native species may be limited
by sea-level rise in future decades.
Developing partnerships associating nature-based
tourism to fund ecosystem restoration
Seychelles provides some examples of collaborations
between private island owners, parastatals, government
agencies and NGOs to achieve successful control or
eradication of invasive species and to develop ecosystem
rehabilitation programmes (Asconit & ICS, 2010;
Government of Seychelles, 2011; Kueffer, et al., 2013).
By creating synergies, such partnerships can speed-up the
long-term process of ecosystem rehabilitation, and can
help meet the financial, technical and ecological challenges
of these complex operations (Rocamora, 2010a). Many
islands are engaged in tourism activities that can help
Fig. 3 Cumulated number of small and medium sized
islands that have benefited from invasive species
management and other forms of ecological restoration
in Seychelles since the 1970s, and type of management.
596
conservation funding (Rocamora & Payet 2002; Nevill,
2004; Skerrett, 2010). The successful record of Seychelles
in invasive species management and ecological restoration
of many small islands is attributable, at least in part, to the
fact that these operations can make economic sense for
private owners or investors wishing to generate revenue
through ecotourism operations. High densities of rats are
incompatible with tourism, and eradication and biosecurity
procedures are more cost-effective than long-term pest
control. The progressive rehabilitation of an island
ecosystem to recreate a wildlife sanctuary with its original
fauna can be marketed as an attraction (Rocamora, 2010b;
Samways, et al., 2010a). The development of partnerships
associating sustainable nature-based tourism with funding
ecosystem restoration must be strongly encouraged.
Perspectives and challenges in invasive species
management and restoration
The creation of island refuges where invasive species
management enables the (re)introduction of species that
cannot survive in the presence of invasive predators
has proven to be extremely effective in improving the
conservation status of various species, including endemic
birds of Seychelles that had come very close to global
extinction. However, several factors limit the further
development of this strategy.
Availability of additional islands suitable for restoration
and preservation
Although Seychelles still has considerable potential
to increase its area free of alien predatory mammals, the
number of islands where such operations can be conducted
is limited. In the inner islands, there are currently five
to eight islands left which could be made, and kept, free
of rats and cats. Most of these have actual or planned
development projects or do not presently fulfil the required
strict biosecurity conditions to prevent reinvasion.
Challenges to eradicate rats from larger islands in a
humid, tropical climate
Techniques currently available to eradicate rats are
more successful in temperate and sub-Antarctic climates
than in tropical environments, particularly humid ones
(Russell & Holmes, 2015). Here, rains (that can seriously
affect the attractiveness and palatability of rat pellets) and
abundant natural food (which reduces the likelihood that
rats will eat the bait) can be present for much of the year
(Varnham, 2010; Keitt, et al., 2015). Whereas rats (Rattus
rattus and R. norvegicus) have been eradicated from
islands of over 10,000 ha outside of the tropics, the largest
tropical eradication of black rats to date is Cayo Centro,
Chinchorro Bank (539 ha, Mexico) and for brown rats
Frégate Island (219 ha, Seychelles) (DIISE, 2017).
Mangroves are also a limiting factor and the main
obstacle to a large-scale rat eradication on Aldabra atoll
(15,380 ha; c.1,300 ha of mangroves). Although small
areas of mangroves can be dealt with by placing bait
stations or tying rodenticide blocks to trees (SamaniegoHerrera, et al., 2015; Samaniego-Herrera, et al., 2017),
using ‘collars’ or ‘bolas’ (Harper, et al., 2015; Rocamora &
Henriette, 2015), efficient methods to eradicate rats from
large tropical islands are not yet sufficiently well developed
(Russell & Holmes, 2015).
Rat eradications may prove challenging on large
islands with high densities of coconut trees, where nuts
provide abundant food for rats both in the trees and on
the ground, (Climo & Rocamora, 2006). This requires
bait to be available at high densities and for a long-time
period. Unpredictable rainfall and the year-round high
primary productivity of Seychelles ecosystems add further
challenges to conducting rat eradications. Abundance of
Rocamora: Island restoration practices in Seychelles
bait-eating crabs can also cause problems (Griffiths, et al.,
2011; Wegmann, et al., 2011; Keitt, et al., 2015) but this
has not so far been a major problem in Seychelles.
Suitable habitats on restored islands non-existent or too
limited for some species
Some rare and threatened species require very specific
habitats that may not be found on small to medium sized
islands. Examples include the Critically Endangered
Seychelles sheath-tailed bat (Coleura seychellensis) and
the Vulnerable Seychelles swiftlet (Aerodramus elaphrus),
both of which occur only on the larger granitic islands,
breeding in caves and feeding on flying insects. Such
limitations also apply to endemic plants and animals
(reptiles, amphibians, invertebrates) found exclusively
at altitudes above 300–400 m on Mahé and Silhouette
where much of the terrestrial diversity of Seychelles
is concentrated (Senterre, et al., 2013). Most of these
species would probably not survive on the low-lying small
islands, where climatic conditions differ from the more
humid and colder high altitudes. Some species may also
require large expanses of specific habitats that could not
be made available on small islands, such as the Vulnerable
Seychelles black parrot (Coracopsis barklii), which
requires extents of palm-dominated forests (Rocamora &
Laboudallon, 2013).
Increased interspecific interactions on small islands with
multiple (re)introductions
The number of species that can coexist on a given island
is limited by the quality and diversity of habitats available
on the island (MacArthur & Wilson,1967), which is
influenced by island area and characteristics. The survival of
any small newly (re)introduced population will depend on
interactions with the species already there (Blondel, 1979).
This factor may partly account for the failed translocation
of Seychelles leaf insects to Conception Island, many of
which were preyed on shortly after their release (Galman,
2011). On Cousine Island (26 ha), the 23 Seychelles whiteeyes translocated in 2008 established a small breeding
population, but predation of nests and fledglings and high
adult mortality did not allow this population to grow despite
considerable efforts. Young fledglings had to be caged and
fed through the mesh by adults as they were repeatedly
preyed on by another introduced species, the Seychelles
magpie-robin (Rocamora, 2013). Such problems were
not observed after the introduction of white-eyes to two
larger islands, Ile du Nord (201 ha), and Frégate (219 ha)
where a large population of Seychelles magpie-robins
was present. This suggests a limit to the number of (re)
introduced species that a small island can host. In other
words, it will become more difficult to ‘squeeze in’ new
species into small rehabilitated islands as their ecosystems
become increasingly saturated.
The need for alternative conservation approaches on
large islands
Fig. 4 Time progression of the rat and cat free area in
the inner islands, the outer islands and the whole of
Seychelles (reported in hectares and as % of the land
surface). The total land surface that could potentially be
freed of rats and cats with currently available techniques
is also indicated.
In Seychelles, the availability of many small islands
suitable for rehabilitation, and the presence of private
island owners willing to develop ecotourism has favoured
the in situ approach. In other countries, such as Mauritius
or New Zealand, more intensive and costly ex situ
techniques, which require the additional step of readapting
the captive reared animals into the wild, have also been
used (Jones & Merton, 2012). In view of the limitations of
the ‘small island restoration’ model, ecosystem restoration
programmes may be developed on the large islands of
Seychelles through the creation of “mainland islands” where
invasive species are controlled or excluded to enable native
species to thrive, as in New Zealand, Australia and Hawaii
(Innes & Saunders, 2011). Predator control and habitat
rehabilitation programmes are being developed at large
scales in Mauritius and La Réunion for the conservation of
threatened land birds (Vikash Tatayah & Marc Salamolard,
pers. comm., 2016). Similar operations could be conducted
in selected priority sites on the largest granitic islands of
Seychelles. Innovative management techniques such as
predator-proof fences, self-resetting traps, more effective
or target-specific bait, etc. will be key to success. The
only Seychelles example to date is permanent rat control
using grids of bait-stations at the main breeding areas of
the Endangered Seychelles white-eye on Mahé (25–40
ha) since 2006 (Rocamora & Henriette, 2015). More such
projects could bring some of the rarest birds of Seychelles,
now restricted to remote small island sanctuaries, back to
the main islands where they once lived. By providing better
access to these species and native wildlife in general, such
‘mainland islands’ would benefit environmental education
programmes for the public and school children. This in
turn would increase awareness of, and hopefully support
for, pest management programmes.
597
Island invasives: scaling up to meet the challenge. Ch 3C Strategy: Outcomes
CONCLUSION
Seychelles’ achievements with invasive species
management and other island restoration practices are
remarkable. This includes a minimum of 50 island
eradications of invasive vertebrate populations, the
rehabilitation of ca. 470 ha of natural habitats, and at least
36 successful island translocations of native species. The
rehabilitation of small and medium sized islands has made
possible the down-listing of four globally threatened land
birds in the IUCN Red List and the recovery of many other
native animals and plants.
Scaling up the size of islands for eradications is now
required in Seychelles. Factors limiting rat eradication on
larger islands include high densities of coconut trees and
the presence of mangroves, especially on Aldabra atoll.
Invasive predators such as rat and cats could be eradicated
from 19–22 more islands with existing techniques, mainly
in the outer (coralline) islands. As a result, the proportion of
the country’s land area free of rats and cats would increase
from 3.9% to 15.4%, but new techniques will be needed to
remove rats from Aldabra and bring this proportion to 50%.
Making half of Seychelles rat and cat free by 2030–2050
could be a commitment made by Seychelles government
and the main stakeholders involved as part of the Honolulu
Challenge on Invasive Alien Species, launched at the 2016
IUCN Word Conservation Congress in Hawaii.
Eradication operations need to be extended to
invertebrates such as invasive ants, moths or snails that
also have a high negative impact on native biodiversity.
Apart from the availability of islands free of invasive
predators, limiting factors to further translocations of rare
and threatened species include lack of suitable habitats and
increased interspecific interactions on small islands with
multiple (re)introductions. Because of global warming and
sea-level rise, the long-term relevance of island restoration
and species translocations to outer low-lying coralline
islands is questioned.
Local partnerships and financial support from naturebased tourism have been key to past successes. We
recommend for these to be enhanced and alternative
conservation approaches such as ‘mainland-islands’ be
developed on large islands. Most importantly, biosecurity
protocols will be critical to prevent (re)invasion of invasive
species, as lack of vigilance and poor biosecurity could
undo so much of what has already been achieved.
ACKNOWLEDGEMENTS
I wish to thank J.P. Siblet (MNHN), J.Y. Kernel
(Biotope), and A. Grieserjohns (Project Coordinating Unit
UNDP-GEF-Government of Seychelles) for authorising
the publication of updated figures and modified short
text sections initially published in Rocamora & Henriette
(2015), the English of which had been reviewed by Jeanne
Mortimer (IBC centre of UniSey) Two anonymous referees
reviewed the manuscript and provided much appreciated
fruitful comments and improvements.
REFERENCES
Asconit and ICS. (2010). France – Seychelles. Rehabilitating Island
Ecosystems. Eradication of invasive exotic species and Re-introduction
of threatened endemic species. Asconit Consultants, ICS and FFEM,
Paris
and
Seychelles.
<http://www.globalislands.net/userfiles/
seychelles6.pdf>. Accessed 15 September 2017.
Beaver, K. and Mougal, J. (2009). Review and Evaluation of Invasive
Alien Species (IAS) Control and Eradication Activities in Seychelles.
Mainstreaming Prevention and Control Measures for Invasive Alien
Species into Trade, Transport and Travel across the Production
Landscape. Consultancy Report, GoS-UNDP-GEF Project. Victoria,
Mahé, Seychelles: Plant Conservation Action Group (PCA). <http://
www.pcaseychelles.org/uploads/1/2/3/6/12369400/seychelles_ias_
control-erad_review_pca_report_only.pdf>. Accessed 15 September
2017.
598
Blondel, J. (1979). Biogéographie et Écologie. Paris: Masson.
Brooke, M. de L., Bonnaud, E., Dilley, B.J., Flint, E.N., Holmes, N.D.,
Jones, H.P., Provost, P., Rocamora, G., Ryan, P.G., Surman, C. and
Buxton, R.T. (2017). ‘Seabird population changes following mammal
eradications on islands’. Animal Conservation 21: 3–12.
Climo, G. and Rocamora, G. (2006). The Successful Eradication of
Black Rats from North Island (Seychelles) in September 2005 and
Proposed Strategies to Reduce the Risks of Reinvasions. Réhabilitation
des Ecosystèmes Insulaires (FFEM) Project. North Island: Island
Conservation Society.
DIISE. (2017). The Database of Island Invasive Species Eradications.
Island Conservation, Coastal Conservation Action Laboratory UCSC,
IUCN SSC Invasive Species Specialist Group. University of Auckland
and Landcare Research New Zealand. <http://diise.islandconservation.
org>. Accessed 31 December 2017.
Fauvel, A.A. (1909). Unpublished Documents on the History of the
Seychelles Islands Anterior to 1810. Victoria: Seychelles Government.
Feare, C.J., French, G.C.A., Nevill, J.E.G., Pattison-Willits, V.S., Wheeler,
V., Yates, T.L., Hoareau, C. and Prescott, C. (2015). ‘Attempted reestablishment of a sooty tern Onychoprion fuscatus breeding colony on
Denis Island, Seychelles’. Conservation Evidence 12: 19–24.
Fritz, U., Branch, W.R., Gehring, P-S, Harvey, J., Kindler C., Meyer, L.,
Du Preez, L., Široký, P., Vieites, D.R. and Vences, M. (2013). ‘Weak
divergence among African, Malagasy and Seychellois hinged terrapins
(Pelusios castanoides, P. subniger) and evidence for human-mediated
oversea dispersal’. Organisms Diversity and Evolution 13: 215–224.
Galman, G. (2011). Suivi Spatio-temporel des Communautés
d’Arthropodes, Effets de l’Eradication des Rats et Tentative
de Réintroduction d’un Insecte Rare dans des Îles en Cours de
Réhabilitation des Seychelles. Paris: Muséum National d’Histoire
Naturelle.
Gerlach, J., Rocamora, G., Gane, J., Jolliffe, K. and Vanherck, L. (2013).
‘Giant tortoise distribution and abundance in the Seychelles islands:
Past, present and future’. Chelonian Conservation and Biology 12(1):
70–83.
Government of Seychelles. (2011). The National Invasive Alien Species
(Biosecurity) Strategy for Seychelles 2011–2015. Victoria, Seychelles:
Ministry of Home Affairs, Environment and Transport & Ministry of
Industry, Natural Resources and Industry.
Griffiths, C.J. (2014). ‘Rewilding in the Western Indian Ocean’. In:
J. Gerlach. Tortoises of the Western Indian Ocean, pp. 325–349.
Manchester, UK: Siri Scientific Press.
Griffiths, C.J., Jones, C.G., Hansen, D.M., Puttoo, M., Tatayah, R.V.,
Müller, C.B. and Harris, S. (2010). ‘The use of extant non-indigenous
tortoises as a restoration tool to replace extinct ecosystem engineers’.
Restoration Ecology 18: 1–7.
Griffiths, R., Miller, A. and Climo, G. (2011). ‘Addressing the impact
of land crabs on rodent eradications on islands’. Pacific Conservation
Biology 17: 347–353.
Hansen, D.M., Donlan, C.J., Griffiths, C.J. and Campbell, K.J. (2010).
‘Ecological history and latent conservation potential: Large and giant
tortoises as a model for taxon substitutions’. Ecography 33: 272–284.
Harper, G.A., Van Dinther, M., Russell, J.C. and Bunbury, N. (2015).
‘The response of black rats (Rattus rattus) to evergreen and seasonally
arid habitats: Informing eradication planning on a tropical island’.
Biological Conservation 185: 66–74.
Henriette, E. (2011). ‘Conservation Introductions as a Tool for the
Recovery of Endangered Island Species: Territoriality and Demography
of the Seychelles White-eye Zosterops modestus on Frégate Island’.
Docteur en Ecologie Thesis. Paris: Muséum National d’Histoire
Naturelle.
Innes, J. and Saunders, A. (2011). ‘Eradicating multiple pests: an
overview’. In: C.R. Veitch, M.N. Clout and D.R. Towns (eds.) Island
invasives: eradication and management, pp. 177–181. Occassional
Paper SSC no. 42. Gland, Switzerland: IUCN and Auckland, New
Zealand: CBB.
IUCN. (2013). Guidelines for Reintroductions and Other Conservation
Translocations. Version 1.0. Gland, Switzerland: IUCN Species
Survival Commission.
Jones, C. and Merton, D. (2012). ‘A Tale of Two Islands: The Rescue
and Recovery of Endemic Birds in New Zealand and Mauritius’. In: J.
Ewen, D. Armstrong, K. Parker and P. Seddon (eds.). Reintroduction
Biology: Integrating Science and Management, pp. 33–72. Oxford:
Blackwell Publishing Ltd.
Jones, H.P. and Kress, S.W. (2012). ‘A review of the world's active
seabird restoration projects’. Journal of Wildlife Management 76: 2–9.
Jones, H.P., Holmes, N.D., Butchart, S.H.M., Tershy, B.R., Kappes,
P.J., Corkery, I., Aguirre-Muñoz, A., Armstrong, D.P., Bonnaud, E.,
Burbridge, A.A., Campbell, K., Courchamp, F., Cowan, P., Cuthbert,
R.J., Ebbert, J., Genovesi, P., Howald, G.R., Keitt, B.S., Kress, S.W.,
Miskelly, C.M., Oppel, S., Poncet, S., Rauzon, M.J., Rocamora, G.,
Russell, J.C., Samaniego-Herrera, A., Seddon, P.J., Spatz, D.R., Towns,
D.R. and Croll, D.A. (2016). ‘Invasive mammal eradication on islands
results in substantial conservation gains’. Proceedings of the National
Academy of Sciences 113 (15): 4033–4038.
Rocamora: Island restoration practices in Seychelles
Kaiser-Bunbury, C.N., Traveset, A. and Hansen, D.M. (2010).
‘Conservation and restoration of plant–animal mutualisms on oceanic
islands’. Perspectives in Plant Ecology and Evolution Systematics. 12:
131–143.
Kaiser-Bunbury, C.N., Mougal, J., Valentin, T., Gabriel, R. and Blüthgen,
N. (2015). ‘Herbicide application as a habitat restoration tool: Impact
on native island plant communities’. Applied Vegetation Science 8 (4):
650–660.
Kappes, P. and Jones, H.P. (2014). ‘Integrating seabird restoration and
mammal eradication programs on islands to maximize conservation
gains’. Biodiversity Conservation 23: 503–509.
Keitt, B., Griffiths, R., Boudjelas, S., Broome, K., Cranwell, S., Millett,
J., Pitt, W. and Samaniego-Herrera, A. (2015). ‘Best practice guidelines
for rat eradication on tropical islands’. Biological Conservation 185:
17–26.
Kömdeur, J. and Pels, M.D. (2005). ‘Rescue of the Seychelles warbler on
Cousin Island, Seychelles: The role of habitat restoration’. Biological
Conservation 124 (1): 15–26.
Kueffer, C., Beaver, K. and Mougal, J. (2013). ‘Management of Novel
Ecosystems in the Seychelles’. In: R.J. Hobbs, E.S. Higgs and C. Hall.
Novel Ecosystems: Intervening in the New Ecological World Order, pp.
228–238. Oxford, UK: Wiley-Blackwell.
Kueffer, C. and Vos, P. (2004). Case Studies on the Status of Invasive
Woody Plant Species in the Western Indian Ocean: 5. Seychelles. Forest
Health & Biosecurity Working Papers FBS/4-5E. Rome, Italy: Forestry
Department, Food and Agriculture Organization of the United Nations.
MacArthur, R.H. and Wilson, E.O. (1967). The Theory of Island
Biogeography. Princeton, NJ: Princeton University Press.
Merton, D., Climo, G., Laboudallon, V., Robert, R. and Mander, C.
(2002). ‘Alien mammal eradication and quarantine on inhabited islands
in the Seychelles’. In: C.R. Veitch and M.N. Clout (eds.) Turning the
tide: the eradication of invasive species, pp. 182–198. Occasional Paper
SSC no. 28. IUCN SSC Invasive Species Specialist Group. IUCN,
Gland, Switzerland and Cambridge, UK.
Mulder, C., Anderson, W., Towns, D. and Bellingham, P. (2011). Seabird
Islands. Ecology, Invasion, and Restoration. New York: Oxford
University Press.
Nevill, J. (2001). ‘Case Study 5.34. Ecotourism as a Source of Funding
to Control Invasive Species’. In: R. Wittenberg and M.J.W. Cock (eds.).
Invasive Alien Species: A Toolkit of Best Prevention and Management
Practices, p. 202. Wallingford, Oxon, UK: CAB International.
Nevill, J. (2004). ‘Ecotourism as a source of funding to control invasive
species: the case of Seychelles’. Insula, International Journal of Island
Affairs. Special Issue February 2004: 99–102.
Nevill, J. (2009). Mainstreaming Prevention and Control Measures for
Invasive Alien Species into Trade, Transport and Travel across the
Production Landscape. National IAS Baseline Report. GOS-UNDP –
GEF. Victoria, Seychelles: Government of Seychelles/UNDP/GEF.
Nevill, J. (2011). ‘Case Study 6. Ecosystem Rehabilitation. The Case of
Small Islands in Seychelles – A Government, NGO and Private Sector
Partnership’. In: Government of Seychelles. Fourth National Report
to the United Nations Convention on Biological Diversity, pp. 84–87.
Victoria, Republic of Seychelles: Department of Environment.
Richardson, D.S., Bristol, R. and Shah, N.J. (2006). ‘Translocation of
the Seychelles warbler Acrocephalus sechellensis to establish a new
population on Denis Island, Seychelles’. Conservation Evidence 3:
54–57.
Rocamora, G. (1997). Rare and Threatened Species, Sites and Habitats
Monitoring Programme in Seychelles. Project G1 EMPS. Volume 1.
Monitoring methodologies and recommended priority actions. Victoria,
Seychelles & Cambridge, UK: Ministry of Environment/BirdLife
International/EU.
Rocamora, G. and Payet, R. (2002). ‘Sustainable Ecotourism as a Financing
Mechanism for the Conservation of Cosmoledo Atoll’. Conference on
Sustainable Development of Ecotourism in Small Islands Developing
States (SIDS) and Other Small Islands. Preparatory Conference for the
International Year of Ecotourism, Mahé (Seychelles), 8–10 December
2001. Final Report. Annexed Case Study (unpaged). Madrid & Victoria:
World Tourism Organisation & Seychelles Ministry of Tourism.
Rocamora, G. and Henriette-Payet, E. (2008). ‘Conservation Introductions
of the Seychelles White-eye on Predator-free Rehabilitated Islands of
the Seychelles Archipelago, Indian Ocean’. In: P.S. Soorae. Global Reintroduction Perspectives: Re-introduction Case Studies from Around
the Globe, pp. 121–129. Abu Dhabi, UAE: IUCN/SSC Re-introduction
Specialist Group.
Rocamora, G. (2010a). ‘Island Restoration’. In: A. Skerrett, P. Thelma
and J. Skerrett. Outer Islands of Seychelles. Zil Elwannyen Sesel, pp.
178–183. Nairobi, Kenya: Camerapix.
Rocamora, G. (2010b). ‘Oiseau-lunettes et autres espèces menacées:
Bilan des actions menées aux Seychelles’. Conservation des Espèces et
Populations Animales (CEPA) Magazine 2: 19–23.
Rocamora, G. (2013). ‘Zosterops modestus Seychelles White-eye’. In:
R.J. Safford and A.F.A. Hawkins (eds.) The Birds of Africa. Volume
VIII: The Malagasy Region, pp. 791–795. London: Christopher Helm.
Rocamora G. (2015). ‘Restoring Islands to Save Species’. In: G. Rocamora
and E. Henriette. Invasive Alien Species in Seychelles: Why and How to
Eliminate Them? Identification and Management of Priority Species,
pp. 55–95. Biotope, Mèze : Island Biodiversity & Conservation Centre,
University of Seychelles and Paris : Muséum National d’Histoire
Naturelle.
Rocamora G. and Henriette E. (2015). Invasive Alien Species in
Seychelles: Why and How to Eliminate Them? Identification and
Management of Priority Species. Biotope, Mèze: Island Biodiversity
& Conservation Centre, University of Seychelles and Paris: Muséum
National d’Histoire Naturelle (Inventaires & Biodiversité series).
Rocamora G. and Laboudallon V. (2013). Seychelles Black Parrot
Coracopsis barklyi. In: R. Safford and F. Hawkins (eds.). The birds of
Africa, vol. 8: The Malagasy region: Madagascar, Seychelles, Comoros,
Mascarenes, pp. 529–531. London: Christopher Helm.
Russell, J.C. and Holmes, N.D. (2015). ‘Tropical island conservation: Rat
eradication for species recovery’. Biological Conservation 185: 1–7.
Russell J., Cole N., Zuel N. and Rocamora G. (2016). ‘Introduced
mammals on Western Indian Ocean islands: Species distribution,
eradication attempts, and long-term island restoration prospects’.
Global Ecology and Conservation 6: 132–144.
Samways, M., Hitchins, P., Bourquin, O. and Henwood, J. (2010a).
Tropical Island Recovery. Cousine Island, Seychelles. Chichester, UK:
Wiley-Blackwell.
Samways, M.J., Hitchins, P.M., Bourquin, O. and Henwood, J. (2010b).
‘Restoration of a tropical island: Cousine Island, Seychelles’.
Biodiversity and Conservation 19(2): 425–434.
Samaniego-Herrera, A., Aguirre-Muñoz A., Méndez-Sánchez, F., RojasMayoral, E. and Cárdenas-Tapia, A. (2015). ‘Pushing the Boundaries
of Rodent Eradications on Tropical Islands: Ship Rat Eradication on
Cayo Centro, Banco Chinchorro, Mexico’. Poster presented at the 27th
International Congress for Conservation Biology. Montpellier, France,
2–6 August 2015.
Samaniego-Herrera, A., Aguirre-Muñoz, A. Bedolla-Guzmán, Y.,
Cárdenas-Tapia A., Félix-Lizárraga, M., Méndez-Sánchez, F., ReinaPonce, O., Rojas-Mayoral, E. and Torres-García, F. (2017). ‘Eradicating
invasive rodents from wet and dry tropical islands in Mexico’. Oryx
52(3): 559–570.
Senterre, B., Henriette, E., Chong Seng, L., Gerlach, J., Mougal, J. and
Rocamora, G. (2013). Seychelles Key Biodiversity Areas - Output 6:
Patterns of Conservation Value in the Inner Islands. Unpublished
consultancy report, Victoria, Seychelles: Ministry of EnvironmentUNDP-GEF project.
Shah, N.J. (2001). ‘Eradication of alien predators in Seychelles: An
example of conservation action on tropical islands’. Biodiversity and
Conservation 10(7): 1219–1220.
Shah, N.J. (2006). ‘Ecological restoration of islands in the Seychelles’.
Conservation Evidence 3: 1–2.
Shah, N.J. (2008). ‘Managing expectations and outcomes for successful
re-introductions and multiple stakeholders in Seychelles’. Avian Biology
Research 1(1): 31.
Skerrett, A. (2010). ‘Outer Island Conservation’. In: A. Skerrett, P.
Thelma and J. Skerrett. Outer Islands of Seychelles. Zil Elwannyen
Sesel, pp. 174–175. Nairobi, Kenya: Camerapix.
van Dinther, M., Bunbury, N. and Kaiser-Bunbury, C.N. (2015). ‘Trial of
herbicide control methods for sisal (Agave sisalana) in the arid island
environment of Aldabra Atoll, Seychelles’. Conservation Evidence 12:
14–18.
Warman, S. and Todd, D. (1984). ‘A biological survey of Aride Island
Nature Reserve, Seychelles’. Biological Conservation 28: 51–71.
Varnham, K. (2010). Invasive Rats on Tropical Islands: Their History,
Ecology, Impacts and Eradication. RSPB Research Report No. 41.
Sandy, Bedfordshire, UK: Royal Society for the Protection of Birds.
Wegmann, A., Buckelew, S., Howald, G., Helm, J. and Swinnerton,
K. (2011). ‘Rodent eradication campaigns on tropical islands: novel
challenges and possible solutions’. In: C.R. Veitch, M.N. Clout and
D.R. Towns (eds.) Island invasives: eradication and management, pp.
239–243. Occassional Paper SSC no. 42. Gland, Switzerland: IUCN
and Auckland, New Zealand: CBB.
Veitch, C.R., Clout, M.N. and Towns, D.R. (2011). Island invasives:
eradication and management. Occassional Paper SSC no. 42. Gland,
Switzerland: IUCN and Auckland, New Zealand: CBB.
Von Brandis, R.G. (2012). ‘Rehabilitation of abandoned coconut
plantations at D’Arros Island, Republic of Seychelles’. Ocean &
Coastal Management 69: 1–7.
599
A. Wegmann, G. Howald, S. Kropidlowski, N. Holmes and A.B. Shiels
Wegmann, A.; G. Howald, S. Kropidlowski, N. Holmes and A.B. Shiels. No detection of brodifacoum residues in the
marine and terrestrial food web three years after rat eradication at Palmyra Atoll, Central Pacific
No detection of brodifacoum residues in the marine and terrestrial food
web three years after rat eradication at Palmyra Atoll, Central Pacific
A. Wegmann1, G. Howald1, S. Kropidlowski2, N. Holmes1 and A.B. Shiels3
Island Conservation, 2100 Delaware Ave, Suite 1, Santa Cruz, CA 95060, USA. <alex.wegmann@tnc.org>. 2U.S. Fish
and Wildlife Service, Pacific Remote Islands Marine National Monument, Palmyra Atoll National Wildlife Refuge, USA.
3
USDA, APHIS, WS, National Wildlife Research Center, Fort Collins, CO 80521, USA.
1
Abstract Invasive alien species represent one of the greatest threats to native plants and animals on islands. Rats
(Rattus spp.) have invaded most of the world’s oceanic islands, causing lasting or irreversible damage to ecosystems
and biodiversity. To counter this threat, techniques to eradicate invasive rats from islands have been developed and
applied across the globe. Eradication of alien rats from large or complex island ecosystems has only been successful
with the use of bait containing a rodenticide. While effective at eradicating rats from islands, rodenticide can persist in
the ecosystem longer than the time required to eradicate the target rat population and can potentially harm non-target
species. However, the persistence of rodenticides in ecosystems following rat eradication campaigns is poorly understood,
though predictions can be made based on the chemical properties of the rodenticide and the environment it is applied
in. Brodifacoum, a relatively persistent second-generation anticoagulant, was used to successfully eradicate rats from
Palmyra Atoll. With this study, we evaluated the persistence of brodifacoum residues in terrestrial and marine species at
Palmyra Atoll (Northern Line Islands) three years after rat eradication. We collected 44 pooled samples containing 121
individuals of the following: mullet (Moolgarda engeli), cockroaches (Periplaneta sp.), geckos (Lepidodactylus lugubris),
hermit crabs (Coenobita perlatus), and fiddler crabs (Uca tetragonon). Despite detection of brodifacoum residue in all
five of the species sampled in this study 60 days after the application of bait to Palmyra Atoll in 2011, brodifacoum
residue was not found in any of the pooled samples collected three years after bait application. Our study demonstrates
how brodifacoum residues are unlikely to persist in the marine and terrestrial food web, in a wet tropical environment,
three years after rat eradication.
Keywords: aerial rodenticide broadcast, best practice, brodifacoum anticoagulant rodenticide, land crabs, Rattus rattus,
risk assessment, tropical island
INTRODUCTION
Invasive alien species represent a key threat to native
plants and animals on islands (Tershy, et al., 2015). In
particular, invasive rodents are known to have widespread
negative impacts following introduction to islands (Towns,
et al., 2006), and rodents have been introduced to most
of the world’s island groups (Atkinson, 1985). In prior
decades, techniques to eradicate invasive rodents from
islands have been developed and applied across the globe,
most using anticoagulant rodenticides (Howald, et al.,
2007). Demonstrable conservation benefits are common
following successful eradication (Jones, et al., 2016;
Brooke, et al., 2017).
To date, rat (Rattus spp.) eradications on tropical
islands experience a lower success rate than those in
temperate regions (Russell & Holmes, 2015). Lack of
seasonality and warm temperatures in tropical latitudes can
provide year-round breeding opportunities and a consistent
abundance of alternative food sources that rodents may
choose instead of the offered bait. Tropical regions also
host land crab populations which readily compete with rats
for bait (Wegmann, et al., 2011; Holmes, et al., 2015). In
2011, Palmyra was the site of a successful eradication of R.
rattus (US Fish and Wildlife Service, 2011). The planning
and implementation of the rat eradication required novel
techniques, including direct baiting of the tree canopy,
and two aerial broadcast applications, each at rates of 75
and 85 kg/ha, of bait containing brodifacoum (0.0025%)
(Wegmann, et al., 2012). Ecotoxicology monitoring
undertaken during and after the project detected residual
brodifacoum in soil, water and biota (Pitt, et al., 2015).
Sampling ceased 60 days after the bait application before
undetectable levels of brodifacoum were reached (Pitt,
et al., 2015). Resources to continue the monitoring were
not secured until three years after the bait application for
rat eradication, providing the opportunity to investigate
longer-term persistence of brodifacoum within the Palmyra
food web.
METHODS
Study site and animals
Palmyra Atoll (5°53’ N, 162°05’ W) is located at the
northern end of the Line Islands in the Central Pacific
Ocean. Palmyra is a wet atoll containing approximately 235
ha of emergent land primarily covered in thick rainforest.
The atoll is an incorporated, unorganised territory of the
United States that is managed in partnership by The Nature
Conservancy (TNC) and the US Fish and Wildlife Service
(USFWS). TNC’s preserve includes Cooper/Menge (94.3
ha) and Barren (4.6 ha) islands. Most of the remaining
emergent land is owned and managed by USFWS as
Palmyra Atoll National Wildlife Refuge, which includes
all marine habitats to 12 nm offshore.
Palmyra’s islets support a regional flora that is typical
of Central Pacific wet forests (Wester, 1985). Heavily
influenced by the Intertropical Convergence Zone, Palmyra
receives an average of 450 cm of rain each year. Palmyra
is a refuge for 11 species of seabirds and is home to a
robust community of land crabs comprised of nine species.
Black rats (Rattus rattus) were inadvertently brought to
Palmyra during WWII. In 2011, Palmyra’s rat population
was eradicated through two strategic applications of
compressed-grain bait containing the second-generation
anticoagulant rodenticide, brodifacoum, at 0.0025% (25
ppm) (Wegmann, et al., 2012). Pitt et al. (2015) collected
and analysed fifty-one animal samples representing
15 species of birds, fish, reptiles, and invertebrates for
brodifacoum residue out to 60 days after the initial bait
application.
Environmental monitoring methodology
We followed the sampling methods outlined in Pitt et
al. (2015) to assess brodifacoum residue concentrations
three years after bait application in cockroaches
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
600
up to meet the challenge, pp. 600–603. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Wegmann, et al.: No brodifacoum residues at Palmyra Atoll
(Periplaneta sp.), fiddler crabs (Uca tetragonon), hermit
crabs (Coenobita perlatus), and geckos (Lepidodactylus
lugubris). Limited time and resource restrictions did not
allow sampling of black-spot sergeant fish (Abudefduf
sordidus) or ants, as undertaken in 2011; however, we
harvested mullet (Moolgarda engeli)), which were
opportunistically collected as carcasses in 2011 following
the eradication and their tissues were found to contain
brodifacoum (Pitt, et al., 2015). All biological samples
were collected at Palmyra Atoll between 4 and 19 June
2014. Biological samples were frozen immediately after
collection.
Sampling site selection (Fig. 1) was determined by ease
of access to the target species. All emergent land at Palmyra
has relatively similar characteristics and vegetation and
was treated with the same baiting prescription during the
2011 eradication campaign. We therefore assumed that site
location would not be an influential factor in brodifacoum
residue concentrations three years after bait was applied.
Biological samples were collected at least 500 m from The
Nature Conservancy’s research station where rodenticide
bait is maintained in bait stations for biosecurity when
planes and ships arrive.
All biological samples were collected with gloved-hands
and segregated in sterile sample bags. Captured hermit
crabs were placed in a freezer (-4 C) for 24 hours and then
removed from their gastropod shells and stored in sterile
sample bags. Mullet were collected by dip-nets and fencenets from several shoreline locations around Palmyra’s
central lagoon. Geckos and cockroaches were captured
at night from the leaves of Scaevola taccada shrubs, and
fiddler crabs were collected from lagoon flats at low tide.
American Veterinarian Medical Association guidelines for
euthanasia were followed with all collections. All samples
were pooled (Table 1) to increase probability of detecting
brodifacoum within the funding limits of this project and
to ensure minimum amounts of sample material were
provided for analysis (e.g. cockroach samples required two
individuals to achieve the 2 g minimum for brodifacoum
residue analysis). Samples were shipped frozen to US
Department of Agriculture’s National Wildlife Research
Center (NWRC) in Fort Collins, Colorado, for brodifacoum
residue analysis. Samples were prepared and analysed
according to methods established by USDA NWRC
for detection of brodifacoum in animal tissue, and these
methods, as well as the laboratory conducting the analyses,
were the same as those used in Pitt et al. (2015). Samespecies pooled carcasses’ samples were homogenised for
analysis.
Brodifacoum residue analysis methodology
The whole bodies of geckos and fish were homogenised.
Cockroaches (whole bodies), as well as fiddler crabs and
hermit crabs were homogenised in a liquid nitrogen freezer
mill and 0.25 g of homogenate was placed into 25 ml glass
tubes for further extraction and analysis following methods
of Pitt, et al. (2015). Aliquots (0.5 g) of each homogenised
gecko and fish sample were placed in MARS vessels for
microwave extraction (Pitt, et al., 2015). Samples were
clarified by centrifugation prior to HPLC analysis.
Brodifacoum analyses were performed with Agilent
1100 and 1200 HPLC systems (Pitt, et al., 2015).
Brodifacoum concentrations were determined from
the peak area ratio of brodifacoum to surrogate in each
extracted sample and were compared to the average peak
area ratio from replicate injections of a working standard.
Samples with analytical concentrations above the linear
range were re-diluted into the linear region.
RESULTS
We collected 44 pooled samples containing 121 total
individuals (Table 1). Brodifacoum residues were not
detected (detection levels reported in Table 1) in any of
the pooled samples of mullet, geckos, cockroaches, hermit
crabs, or fiddler crabs.
DISCUSSION
Ecotoxicology monitoring is uncommon for rodent
eradication projects using rodenticides, but future projects
are dependent on the collective knowledge gained
from toxicological monitoring efforts. The Palmyra
rat eradication used substantially higher rodenticide
application rates compared to other rodenticide-based
rodent eradication projects on islands and provided a
unique opportunity to follow residue persistence in the
environment over time. Brodifacoum residues were
detected in soil, water and biota up to 60 days after the first
aerial broadcast application (Pitt, et al., 2015) but were no
longer detectable in the range of biota studied three years
later, indicating rodenticides break down in this ecosystem
over time. Resource availability did not allow complete
repetition of the 2011 sampling, thus we chose to sample
animals with known residue concentrations, as this had the
most biologically useful outcome for management.
The use of second generation anticoagulant
rodenticides can pose significant risks to non-target species
(Howald, et al., 2007), particularly birds and mammals.
However, knowledge gaps exist, particularly for taxa less
sensitive to rodenticides, such as reptiles and invertebrates
(Hoare & Hare, 2006). The distribution and longevity of
rodenticide residue within a food web will be a function of
rodenticide properties and how it is applied, environmental
Table 1 Biological samples analysed in 2014 for
brodifacoum residue analysis following the 2011
eradication of rats from Palmyra Atoll. “Pooled”
represents the number of individuals contained
in each sample; “MLOD” is the mean level of
brodifacoum detection.
Fig. 1 Locations of sample collections that were used to
investigate persistence of residual brodifacoum three
years after the implementation of the 2011 eradication
of rats from Palmyra Atoll.
Organism
Mullet
Gecko
Cockroach
Hermit crab
Fiddler crab
Samples
analysed
9
5
15
5
10
Pooled
2–3
5
1–2
3
3
MLOD
(μg/g)
0.013
0.011
0.011
0.0057
0.0057
601
Island invasives: scaling up to meet the challenge. Ch 3C Strategy: Outcomes
compartments it ultimately resides within (e.g. soil,
animals), open pathways to transfer residue (e.g. scavenger
consumption of poisoned carcasses), and exposure to
environmental conditions (e.g. temperature, precipitation,
ultraviolet radiation, and fungi) that impact its persistence.
Ultimately, the breakdown of rodenticides is believed to
be accelerated in soil rich in organic matter with healthy
populations of microbiological organisms. Different island
ecosystems can be expected to have different timescales of
residue longevity, and we expect our results will transfer
most closely to other wet tropical atolls and low islands,
rather than dry and/or temperate island environments.
changes to native species populations following the
removal of rat impacts are emerging, including increased
seedling recruitment of several native tree species and the
non-native coconut palm (Cocos nucifera), the elimination
of a non-native mammal-biting mosquito population
(Aedes albopictus), as well as the discovery of two newto-Palmyra land crab species (Geograpsus grayi and
Ocypode cordimanus). These short and long-term changes
are consistent with management expectations, and the rat
eradication has proven to be a baseline restoration activity
to advance natural resource management goals.
Rodenticides are known to temporarily infiltrate
the food web when undertaking rat eradications as
happened with the Palmyra rat eradication. Brodifacoum
residues were found in ocean water, soil, and marine
and terrestrial biota within 60 days of the initial baiting,
indicating diverse food web integration (Pitt, et al., 2015).
Other studies document brodifacoum residues in various
compartments of the food web after brodifacoum bait was
applied to eradicate rats from islands (e.g. Dowding, et al.,
1999; Masuda, et al., 2014; Masuda, et al., 2015; Pitt, et
al., 2015; Siers, et al., 2015; Rueda, et al., 2016; Shiels,
et al., 2017). Although few studies include long-term (>1
year) sampling for residues after brodifacoum application,
there are three recent studies that report residues in
animals two years (Rueda, et al., 2016), three years
(Siers, et al., 2015), and four years (Shiels, et al., 2017)
post-application. Brodifacoum persisted in lava lizards
(Microlophus duncanensis) in the Galápagos Islands for
2.1 years (Rueda, et al., 2016), where liver residue levels
were <0.200 μg/g (mean level of detection [MLOD] =
0.010 μg/g). On Wake Island in the Pacific Ocean, three
years after rat eradication (Siers, et al., 2015), two out of 69
fish samples had detectable levels of brodifacoum in their
livers, with concentrations 0.0038 μg/g and 0.0086 μg/g
(MLOD = 0.0035 μg/g); the two fish were caught within
an intermittently land-locked pond. Finally, on Desecheo
Island, Puerto Rico, detectable levels of brodifacoum were
found in seven animal samples (three endemic lizards, two
black rats, one forest bird, and one cockroach sample [18
individuals]) four years after bait application (Shiels, et al.,
2017). The range of brodifacoum residues in these seven
samples was 0.027-0.134 μg/g (MLOD = 0.0054-0.012
μg/g, depending on species. Desecheo, Wake, and the
Galápagos islands receive less rainfall than Palmyra (e.g.
Desecheo = 1,020 mm/yr, Wake = 906 mm/yr; Pinzon,
Galápagos = <1,100 mm/yr; Palmyra = 3,500 mm/yr),
and this may contribute to the lack of detectable levels
of brodifacoum in the Palmyra food web three years after
bait application. We hypothesise that warmer and wetter
environments, and soils with more diverse microbiological
communities support microbiological processes breaking
down residues faster. This remains an important research
avenue, including decomposition experiments in a
laboratory setting.
ACKNOWLEDGEMENTS
Undertaking eradications of invasive species from
islands should only proceed where expected benefits
outweigh expected costs (Broome, et al., 2014), including
consideration of the environmental impacts of the method
used (Empson & Miskelly, 1999). Potential non-target
impacts were anticipated as part of the environmental
impact assessment for the Palmyra rat eradication, but the
decision to proceed was based on negative impacts ceasing
shortly after the bait application and positive benefits
accruing over a longer time-span (US Fish and Wildlife
Service, 2011). Immediately following bait application,
brodifacoum residues were detected within multiple
levels of the food web, and were attributed to mortality of
birds, fish, and crabs (Pitt, et al.., 2015). Our results show
undetectable levels of residue three years later, suggesting
this short-term impact is no longer present. Longer-term
602
The authors thank Coral Wolf (Island Conservation),
Eric VanderWerf (Pacific Rim Conservation), Stacie
Hathaway (US Geological Survey), and Marie Hellen
(Simon Frasier University) for assistance with sample
collection. We also thank the staff of The Nature
Conservancy’s Palmyra Program for operational support.
This study was funded by a grant from the Packard
Foundation. The field activities and sample collections
were authorised under permit number 12533-14033 by
the Palmyra Atoll National Wildlife Refuge, U.S. Fish and
Wildlife Service, Department of the Interior.
REFERENCES
Atkinson, I.A.E. (1985). ‘The Spread of Commensal Species of Rattus to
Oceanic Islands and their Effects on Island Avifaunas’. In: P.J. Moors
(ed.) Conservation of Island Birds. pp. 35–81. Cambridge: International
Council for Bird Preservation.
Brooke, M.d.L., Bonnaud, E., Dilley, B.J., Flint, E.N., Holmes, N.D.,
Jones, H.P., Provost, P., Rocamora, G., Ryan, P.G., Surman, C. and
Buxton, R.T. (2017). ‘Seabird population changes following mammal
eradications on islands’. Animal Conservation 21(1): 3–12.
Broome, K., Cox, A., Golding, C., Cromarty, P., Bell, P. and McClelland,
P. (2014). Rat Eradication Using Aerial Baiting. Current Agreed Best
Practice Used in New Zealand (Version 3.0). Wellington, New Zealand:
New Zealand Department of Conservation internal document.
Dowding, J.E., Murphy, E.C. and Veitch, C.R. (1999). ‘Brodifacoum
residues in target and non-target species following an aerial poisoning
operation on Motuihe Island, Hauraki Gulf, New Zealand’. New
Zealand Journal of Ecology 23: 207–214.
Empson, R.A. and Miskelly, C.M. (1999). ‘The risks, costs and benefits of
using brodifacoum to eradicate rats from Kapiti Island, New Zealand’.
New Zealand Journal of Ecology 23(2): 241–254.
Hoare, J.M. and Hare, K.M. (2006). ‘The impact of brodifacoum on nontarget wildlife: Gaps in knowledge’. New Zealand Journal of Ecology
30: 157–167.
Holmes, N.D., Griffiths, R., Pott, M., Alifano, A., Will, D., Wegmann, A.S.
and Russell, J.C. (2015). ‘Factors associated with rodent eradication
failure’. Biological Conservation 185: 8–16.
Howald, G., Donlan, C.J., Galvan, J.P., Russell, J.C., Parkes, J.,
Samaniego, A., Wang, Y., Veitch, D., Genovesi, P., Pascal, M., Saunders,
A. and Tershy, B. (2007). ‘Invasive rodent eradication on islands’.
Conservation Biology 21(5): 1258–1268.
Masuda, B.M., Fisher, P. and Jamieson, I.G. (2014). ‘Anti-coagulant
rodenticide brodifacoum detected in dead nestlings of an insectivorous
passerines’. New Zealand Journal of Ecology 38: 110–115.
Masuda, B.M., Fisher, P. and Beaven, P. (2015). ‘Residue profiles of
brodifacoum in coastal marine species following an island rodent
eradication’. Ecotoxicology and Environmental Safety 113: 1–8.
Jones, H.P., Holmes, N.D., Butchart, S.H.M., Tershy, B.R., Kappes,
P.J., Corkery, I., Aguirre-Muñoz, A., Armstrong, D.P., Bonnaud, E.,
Burbidge, A.A., Campbell, K., Courchamp, F., Cowan, P.E., Cuthbert,
R.J., Ebbert, S., Genovesi, P., Howald, G.R., Keitt, B.S., Kress, S.W.,
Miskelly, C.M., Oppel, S., Poncet, S., Rauzon, M.J., Rocamora, G.,
Russell, J.C., Samaniego-Herrera, A., Seddon, P.J., Spatz, D.R., Towns,
D.R. and Croll, D.A. (2016). ‘Invasive mammal eradication on islands
results in substantial conservation gains’. Proceedings of the National
Academy of Sciences 113: 4033–4038.
Pitt, W.C., Berentsen, A.R., Shiels, A.B., Volker, S.F., Eisemann,
J.D., Wegmann, A.S. and Howald, G.R. (2015). ‘Non-target species
mortality and the measurement of brodifacoum rodenticide residues
after a rat (Rattus rattus) eradication on Palmyra Atoll, tropical Pacific’.
Biological Conservation 185: 36–46.
Wegmann, et al.: No brodifacoum residues at Palmyra Atoll
Rueda, D., Campbell, K.J., Fisher, P., Cunninghame, F. and Ponder, J.B.
(2016). ‘Biologically significant residual persistence of brodifacoum
in reptiles following invasive rodent eradication, Galapagos Islands,
Ecuador’. Conservation Evidence 13: 38.
Russell, J.C. and Holmes, N.D. (2015). ‘Tropical island conservation: Rat
eradication for species recovery’. Biological Conservation 185: 1–7.
Shiels, A.B., Witmer, G.W., Samra, C., Moulton, R.S., Ruell, E.W.,
O’Hare, J.R., Eisemann, J.D., Volker, S.F. and Goldade, D.A. (2017).
Assessment of Bait Density, Bait Availability, and Non-target Impacts
During an Aerial Application of Rodenticide to Eliminate Invasive Rats
on Desecheo Island, Puerto Rico. Final Report QA 2588. Ft Collins,
CO: USDA, APHIS, WS, NWRC.
Siers, S.R., Shiels, A.B., Goldade, D.A., Volker, S.F., McAuliffe, T.W.,
Coad, H.L. and Pitt, W.C. (2015). Wake Atoll Fish Tissue Sampling and
Analysis Three Years After an Island-wide Rodenticide Application.
Final Report QA 2241. Hilo, HI: USDA, APHIS, WS, NWRC.
Tershy, B.R., Shen, K.-W., Newton, K.M., Holmes, N.D. and Croll, D.A.
(2015). ‘The importance of islands for the protection of biological and
linguistic diversity’. Bioscience 65(6): 592–597.
Towns, D.R., Atkinson, I.A.E. and Daugherty, C.H. (2006). ‘Have
the harmful effects of introduced rats on islands been exaggerated?’.
Biological Invasions 8(4): 863–891.
US Fish and Wildlife Service (2011). Palmyra Atoll National Wildlife
Refuge Rat Eradication Project Final Environmental Impact Statement.
Honolulu, Hawaii: US Fish and Wildlife Service.
Wegmann, A., Buckelew, S., Howald, G., Helm, J. and Swinnerton, K.
(2011). ‘Rat eradication campaigns on tropical islands: novel challenges
and possible solutions’. In: C.R. Veitch, M.N. Clout and D.R. Towns
(eds.) Island invasives: eradication and management, pp. 239–243.
Occassional Paper SSC no. 42. Gland, Switzerland: IUCN and
Auckland, New Zealand: CBB.
Wegmann, A., Flint, E., White, S., Fox, M., Howald, G., McClelland, P.,
Alifano, A. and Griffiths, R. (2012). ‘Pushing the Envelope in Paradise:
A Novel Approach to Rat Eradication at Palmyra Atoll’. In: R.M. Timm
(ed.) Twenty Fifth Vertebrate Pest Conference, pp. 48–53. Davis, CA:
University of California.
603
P. Bell, H. Nathan and N. Mulgan
Bell, P.; H. Nathan and N. Mulgan. ‘Island’ eradication within large landscapes: the remove and protect model
‘Island’ eradication within large landscapes: the remove
and protect model
P. Bell, H. Nathan and N. Mulgan
Zero Invasive Predators Ltd, 39 Waiapu Road, Karori, Wellington 6012, New Zealand <phil@zip.org.nz>.
Abstract New Zealand has been the world leader in the eradication of invasive mammalian predators from offshore
islands. Today, the focus for invasive predator management is shifting to larger landscapes; big inhabited islands or
the mainland itself. The most cost-effective approach in the long term will be to eradicate the predators from those
areas, ensuring permanent freedom for vulnerable and threatened native biodiversity to recover or be reintroduced. Island
eradication technologies cannot always be employed on the mainland (e.g. aerial brodifacoum), so a new approach
is required. Zero Invasive Predators Ltd (ZIP) is a not-for-profit research and development entity, established in New
Zealand through public, private, and philanthropic funding, to pioneer a novel predator management model for landscapescale application – a model known as ‘Remove and Protect’. ZIP is developing the tools and technologies to both enable
the complete removal of rats, possums, and stoats from large areas of mainland New Zealand, and then protect those areas
from reinvasion. Among the innovations being tested is the ‘virtual barrier’, essentially converting large peninsulas into
islands without the use of traditional predator fencing (which is expensive and impractical in some terrain); and a ‘minimal
infrastructure’ detection system for automated early warning of any predator incursions. We review the transformative
predator management model ZIP is developing and how it could help to pave the way towards large-scale predator-free
landscapes.
Keywords: detection, invasive predators, mainland, Rattus, response, Trichosurus vulpecula, virtual barrier, 1080
INTRODUCTION / CONTEXT
New Zealand is a global biodiversity hotspot (Myers, et
al., 2000), yet more than 3,000 native taxa are threatened
or at risk of extinction (Hitchmough, 2013). It is generally
agreed that there are three mammalian predator species
that cause most of the ecological damage in New Zealand:
possums (Trichosurus vulpecula), ship rats (Rattus rattus),
and stoats (Mustela erminea) (Brown, et al., 2015). From
here on, the term ‘predators’ refers to these three species
plus Norway rats (Rattus norvegicus). The house mouse
(Mus musculus) is specifically excluded as a predator in
the context of this paper and is not a target species for
ZIP. Aside from the estimated 25 million native birds they
kill each year (Russell, et al., 2015), predators cost New
Zealand hundreds of millions of dollars annually, both in
terms of revenue lost and in control costs (Clout, 2011), and
they impact the country’s primary production base through
the transmission of diseases such as bovine tuberculosis
(Coleman & Caley, 2000).
New Zealand has an impressive track record in the
eradication of invasive mammalian predators from offshore
islands for the protection of native biodiversity. Since the
first successful eradication in 1964 (Towns & Broome,
2003), 134 islands have been completely freed from
invasive mammals (Parkes, et al., 2017a). Although costs
vary widely from island to island, the initial eradication
cost is in the order of NZ$300/ha (Parkes et al., 2017b); and
the ongoing biosecurity surveillance costs of these islands
typically ranges from NZ$17 to NZ$160/ha per annum (New
Zealand Department of Conservation (DOC) unpublished
data, 2017). These costs exclude incursion response. For
example, the stoat incursion response on Kapiti Island in
2010−2011 cost approximately NZ$600,000 (NZ$305/
ha) (King, et al., 2014). These predator-free islands are
considered to be the ‘jewels’ of the conservation crown;
however, they represent only 58,921 ha, or <0.01% of the
land area of New Zealand (Parkes, et al., 2017a).
For most of the New Zealand mainland, where restricting
the reinvasion of predators is currently not possible, the
management model used is the ongoing suppression of
predator populations. Currently the main tool used by the
major predator management agencies (DOC, TB Free New
Zealand, Regional Councils) for large scale (up to 100,000
ha) predator control is repeated pulsing of aerially applied
sodium fluoroacetate (1080) toxin, typically every three to
five years (Brown, et al., 2015; Elliott & Kemp, 2016). The
current annualised cost of this is approximately NZ$10/ha
(Brown, et al., 2015). The benefits of this technique are of
limited duration without ongoing sustained control, because
not all individuals are removed from the treatment area,
and immigration is uncontrolled so predator populations
are able to recover (Griffiths & Barron, 2016).
The alternative, ground-based predator control methods
rely on either a knockdown of the resident predator
population, followed by ongoing suppression to low levels,
or seasonal control to realise biodiversity benefits (e.g. for
the native bird breeding period). This work is relatively
labour intensive (via trapping or toxins in bait stations) and
is presently undertaken over areas of up to 50,000 ha (e.g.
Murchison mountains stoat trapping; Hegg, et al., 2013).
The current annualised cost of this work is in the order of
NZ$25 to NZ$60/ha depending on the scale and intensity
of the control efforts and target predator species (Brown,
et al., 2015).
Predator exclusion fencing, a physical mesh fence with
a solid steel capping, is also used to recreate eradication-like
conditions on the mainland (colloquially, New Zealand’s
North and South Islands) by providing a physical barrier
to halt reinvasion (Burns, et al., 2012). Predator fencing is
scale-limited by terrain and cost, with the cost of recently
constructed fences ranging from NZ$253−NZ$461/linear
metre (Curnow & Kerr, 2017), with ongoing maintenance
costs estimated to be 4% of capital costs per annum for the
life of the fence (Norbury, et al., 2014), and eradication
costs additional. Debate continues on the ecological, social
and financial return on investment for predator fencing
(Scofield, et al., 2011; Scofield & Cullen, 2012; Innes, et
al., 2012; Norbury, et al., 2014).
To dramatically improve the status of New Zealand’s
biodiversity, a step change is required in the ability to
manage predators, and the cost of doing so. The New
Zealand Government has declared the goal of a predatorfree New Zealand by 2050 (Cabinet, 2016). In order to
In:
604C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 604–610. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Bell, et al.: The remove and protect model
achieve this ambitious goal, the country will need to heed
the call of The Royal Society of New Zealand (2014), for
urgent action to develop novel approaches and to improve
existing tools to protect the country’s environment and
economy.
Remove and protect model
One novel approach being investigated is the ‘remove
and protect’ model, entailing complete removal of predators
from an area and then protection against reinvasion. In
essence, this creates permanent ‘island’ eradications within
large landscapes of the New Zealand mainland. A research
and development entity, Zero Invasive Predators Ltd (ZIP;
founded in 2015), has been established with the purpose of
developing the ‘toolbox’ to enable this model.
The remove and protect model involves three streams
of research and development:
Initial removal of target predators
The most common and most successful technique
for island eradication has been the aerial application of
the toxin brodifacoum (Howald, et al., 2007; Parkes, et
al., 2011). However, the use of this technology in New
Zealand is governed by a Code of Practice (Epro Ltd.
2006) that limits its current use to offshore islands and
stock-free areas of the mainland behind predator fences
– preventing its immediate application in the remove and
protect model. As a result, eradicating predators on the
New Zealand mainland will likely require new techniques
to be developed or novel refinement of the application of
existing tools – refer to Case Study 1 for one such example.
Defending a line to protect against reinvasion
Implementing a campaign of the scale of predatorfree New Zealand by 2050 (Cabinet, 2016) will require
the ability to divide the country up into manageable land
parcels for progressive removal operations. Predator
fencing has allowed small areas to be treated as ‘islands’ on
the mainland but has limited application because of rugged
terrain and/or social acceptance (Clapperton & Day, 2001;
Burns, et al., 2012). Dividing up the country will require
additional approaches; the creation of a virtual barrier is
one such approach – refer to Case Study 2.
Detecting and removing invaders before they significantly
impact on the predator-free area
Traditionally, in the island eradication context,
biosecurity surveillance consists of intensive networks
of passive devices to find individual invaders (Russell,
et al., 2008). In order to ensure the remove and protect
model is scalable, and to protect any significant predatorfree investment, there is a need to develop a minimal
infrastructure detection system that can facilitate timely
incursion response before significant ecological damage
is incurred – refer to Case Study 3 for detection concepts
being explored.
Changing the cost model
Eradication is the most cost-effective methodology
for predator management (Pascal, et al., 2008), so long as
long-term biosecurity costs are manageable, as the upfront
costs of removal only need to be found once. However,
on the mainland, where reinvasion into management
sites is typically not controllable, the most cost-efficient
technique at present is to aim for predator suppression
over as large a land area as affordable, in the knowledge
that it will need to be repeated ad infinitum to maintain
the gains achieved. In New Zealand, where a relatively
modest budget for predator control (given the scale of the
issue at hand) is largely static year-on-year, the cyclical
pattern of suppression means that only a limited land area
can be managed and that cannot expand without increased
investment.
The remove and protect model seeks to change that cost
structure. By treating blocks of land like island eradications,
i.e. removing all predators and managing reinvasion to zero,
those gains can be secured, and the predator management
programme can be expanded to treat new land areas. Due
to the greater expected biodiversity outcomes derived from
complete predator absence in the long term (Ismar, et al.,
2014; Towns, et al., 2016), i.e. a larger ecological return
on investment, the initial management costs can be greater
than those currently afforded for suppression, especially as
they are a one off cost. However, for this cost structure to
be feasible, the remove and protect model must achieve
similar cost profiles to those of island or fenced sanctuary
eradications in both the removal and maintenance phases.
The initial targets ZIP is currently working to are: initial
predator removal costs of NZ$100/ha (cf. NZ$300/ha
for island eradications; Parkes, 2017b); NZ$200/m for
installation of a virtual barrier (cf. NZ$253−NZ$461/m
for predator fencing; Curnow & Kerr, 2017); and NZ$50/
ha/annum for detection and response (cf., for example,
NZ$160/ha per annum for biosecurity surveillance on Ulva
Island; DOC unpublished data, 2017). All costs exclude
Goods and Services Tax (GST).
A focussed approach: Zero Invasive Predators Ltd
(ZIP)
The opportunity to establish a public-philanthropic
partnership presented itself when the NEXT Foundation
approached DOC to invest in ‘transformative change’
for conservation. In what is a first for DOC, the decision
was made to ‘spin out’ of Government and establish ZIP
as a limited liability company (with NEXT Foundation as
the sole shareholder). Founded in 2015, the intention was
that ZIP would be tightly focussed on the core challenge
of developing a new model for predator management; the
equivalent of taking a specialist research and development
unit and sheltering it from the rest of a business until the
problem is ‘solved’. It was further considered that freedom
from Government would provide the best environment in
which to remain agile and innovative.
While ZIP has a business structure, it does not have
commercial motives. Any self-generated Intellectual
Property is held for New Zealand, effectively making it
openly available to those in New Zealand who want to use
or build upon it. The founding constitution confirms this
‘not for profit’ stance, with any products to be sold at the
most accessible price point in New Zealand (while reserving
the right to profit from international sales), with any profit
to be reinvested in conservation, rather than returned as
a dividend to shareholders. ZIP is also recognised as a
Registered Charity by the Charities Commission (the
governing body in NZ). This charitable status has aided in
securing further philanthropic investment (beyond NEXT
Foundation) as donations, which are tax deductible in New
Zealand.
Some of the high-level goals of ZIP, such as removal
of possums and a reduced reliance on cyclic toxin
applications, have also attracted support from New Zealand
dairy companies, who share those intentions (F. Eggleton,
Fonterra Co-operative Group, pers. comm). This support
includes non-shareholding investment in the research and
development programme, thereby further enhancing the
unique public-philanthropic-private investment positioning
of ZIP.
605
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
Operating culture – try, sense, respond
REMOVE AND PROTECT CASE STUDIES
Ecological systems are usually complex and therefore
the development approach of ZIP is to ‘try, sense, and
respond’. Potential solutions are suggested, techniques and
tools are rapid-prototyped and placed in the field as soon as
possible, impacts are measured, and prototypes are refined
as soon as required. The ‘try, sense, respond’ approach
allows rapid learning about real world constraints, which in
turn informs the next iteration of development and testing.
Case study 1: Removal – ‘1080 to Zero’
This operating style aims to recognise failure quickly,
to expose what we don’t know, and to maximise the return
on effort and resources. Supporting this ‘fast fail’ approach,
field trials of prototypes typically begin at small scale,
i.e. less than five units, in the expectation that limitations
will be exposed and the prototype redesigned. Once the
prototype shows sufficient promise, the trial is scaled-up in
stages, going from, for example, 50 to 100 units, then many
hundreds of units, etc. to test if the statistical performance
holds as the scale increases. Alternatively, if the prototype
fails catastrophically at the small scale, and no practicable
alternatives are found, the trial is shut down to minimise
loss of investment.
This operating culture is strengthened by a diverse,
highly-skilled team, purpose-built for research and
development. Scientists and engineers co-design field trials
and technologies, field rangers actively test prototypes,
with timely data analysis by a specialist modeller. Input
from all aspects of the team feeds into each step of the
development process, enabling rapid evolution of the
project. All team members spend time at the field site(s) to
remain grounded in the challenge.
Development in the field
ZIP, under permission from DOC (the land manager),
has established a 391 ha forested site at Bottle Rock
Peninsula, Queen Charlotte Sound, Marlborough
(41°06’30” S, 174°14’06” E) dedicated to field trialling the
remove and protect system, and its component prototype
parts. Remove and protect is well suited to peninsulas as
they are easier to defend, with only one major exposed
front (with the sea ‘protecting’ the remainder). Interception
efforts can then be concentrated within a relatively small
zone to protect a much larger area.
Bottle Rock Peninsula was selected as it offered the
ideal initial size for rats and possums, and was a favourable
shape (2 km narrow neck with a bulbous peninsula).
Importantly, this peninsula is not a site of high biodiversity
priority for DOC (unpublished data, 2015), therefore it is
able to be manipulated without risk to vulnerable native
species. However, it does enable a ‘real world’ assessment
of new or modified technologies. [NB: the majority of the
field trials carried out at Bottle Rock to date have excluded
stoats on account of their home range size, mobility, and
our current lack of sensitive detection devices rendering
robust stoat research impracticable.]
Evaluation of the performance of the remove and
defend model at Bottle Rock Peninsula uses a ‘systems
design’ approach (Cabrera, et al., 2008), assessing the
whole, as opposed to a reductionist approach which seeks
to understand the role of the individual elements to explain
the utility of the system. The goal is to prove the system
works, not just some parts of it, hence multiple tools need to
be tested simultaneously in the defence system. Individual
considerations are secondary and are investigated by
‘switching off’ components to specifically test their relative
impact on the system’s performance.
606
It is expected that an aerially applied tool will be
required for the initial removal of predators at large-scale
implementation sites. Some of the early work developing
techniques for island eradications investigated sodium
fluoroacetate (1080) as an option (McFadden & Towns,
1991; Moors, 1985). However, it was subsequently
discounted because of its acute toxicity and the perception
that some individuals of the target populations could detect
it in the bait and avoid it (McFadden & Towns, 1991).
There has been significant improvement since that work,
namely prefeeding to increase toxicant uptake (Nugent, et
al., 2011) and manufacturing quality control (Nugent, et al.,
2010; Nugent, et al., 2012). Extensive use in suppression
operations has refined aerial 1080 use, but those operations
still do not remove all target individuals (Elliott & Kemp,
2016).
ZIP sought to test whether dual aerial 1080 operations,
each using different bait (to overcome learnt aversion;
Ross, et al., 2000) and coupled with multiple prefeed
applications, could completely remove rats and possums.
Success was deemed to be functional extinction. The
thresholds for achieving functional extinction were set at
≤1 possum per 400 ha (OSPRI, 2014); and ≤1 rat per 100
ha (Innes, et al., 2011).
The trial was carried out on a 1,600 ha area (39°15’30”
S, 174°07’45” E) on the north-eastern slope of Mt Taranaki.
A 400 ha core, set back with a 1 km buffer to minimise
reinvasion compromising the results (Griffiths & Barron,
2016), was intensively monitored for surviving rats and
possums after treatment with toxin. The trial excluded
stoats due to the scale being insufficient to account for stoat
home range size and mobility (Murphy & Dowding, 1994;
Murphy & Dowding, 1995).
Prior to commencing the trial, monitoring (using
peanut-butter filled chew cards, self-manufactured using
corflute supplied by Pest Control Research and Pic’s
peanut butter – Picot Productions Ltd) was deployed three
times for between two and 10 nights using between 36
and 55 cards each time. The cards were placed every 50
metres on 2–3 randomly selected lines (of between 1.6 and
2 km in length) within the 400 ha core. The purpose of
this monitoring was not to measure a relative abundance,
but merely to confirm presence of target animals. 98% of
total cards deployed were chewed by rats, 6% of total cards
deployed were chewed by possums.
The first phase of baiting consisted of multiple prefeed
baiting of non-toxic RS5, 6 g, cinnamon-masked cereal
pellets (manufactured by Orillion, formerly Animal Control
Products) applied by helicopter-slung bait-spreading
bucket – at (on-ground application rates of) 4 kg/ha; 2 kg/
ha (20 days later); 1 kg/ha (21 days later); 1 kg/ha (47 days
later). Application of (on-ground rate) 4 kg/ha of RS5, 6
g, 0.15% 1080, cinnamon-masked cereal pellets (Orillion)
followed 21 days later. Bait was flown with a 50% swath
overlap, as per island eradication best practice (Broome,
et al., 2014), to ensure no gaps in bait coverage. Baiting
was intended to be completed in winter, when 1080 has
been shown to be most effective (Veltman & Pinder, 2001;
Gillies, et al., 2003); but adverse weather resulted in the
toxin being applied on 1 December 2016.
In an effort to detect survivors, 835 chew cards were
deployed on a 50 m × 100 m grid throughout the 400 ha
core four nights after the toxin application, and checked
every eight days, for a total of 42 days. In addition, 421
pre-weathered tracking tunnels installed on a 100 m × 100
m grid were baited 17 days into the detection period and
maintained live until the same 42-day period post-toxin
Bell, et al.: The remove and protect model
application had passed. Furthermore, 80 motion-activated
cameras (Little Acorn, LTL5200 and LTL5300) were
deployed in a 100 m × 100 m grid in the north-eastern
corner of the ZIP block for the final 17 nights of the
detection period to validate the performance of the other
detection devices.
Functional extinction of possums was considered to
be achieved, with only one possum detection (chew card)
recorded across 36,430 detection nights across all applicable
detection devices (chew cards and cameras). The same was
not achieved for rats, with 42 detections (two chew cards;
25 tracking tunnels; 15 camera detections) recorded over
46,755 detection nights.
In light of the numbers of surviving rats, we attempted
to individually test them for any learnt bait aversion (rather
than undertake the second phase of toxic aerial baiting).
Research by Morgan (2004) suggested that cereal pellets
cannot overcome aversion if created by cereal pellets in the
first place; however, that study did not include prefeeding.
Morgan, in the same work, states that ‘learnt food safety’
(i.e. learnt through prefeeding) is a very strong behaviour
once established. Ross et al. (2000) achieved 30% mortality
in captive 1080 bait-shy possums when ‘postfed’ with
cereal (compared with 0% of non-postfed possums). We
sought to determine whether it is possible, in the wild, to
overcome any bait aversion in the surviving rats through
prefeeding with the different bait, even if it is cereal.
The 1,600 ha trial area was prefeed-baited twice, using
non-toxic Wanganui #7, 6g, double orange-masked cereal
pellets (Orillion) from a helicopter-slung bait-spreading
bucket, seven days apart (58 and 65 days after the first
toxin application). McGregor live-capture traps were set
in areas of known detections and baited with a single
Wanganui #7 0.15% 1080 6g double orange-masked cereal
pellet (Orillion). Traps were baited in such a way that the
rat had to interfere with the pellet to trigger the trap. Traps
were in place for 270 trap nights across various detection
sites.
Thirteen rats were caught that were deemed to be
survivors based on the weight:age profile (Bentley & Taylor,
1965); animals that were very likely to have been present
when the initial toxin application was carried out. Of those,
six were found dead in the trap (following consumption of
a lethal dose of the bait), while an additional two were alive
but showed clear signs of toxicosis with bait consumed
(with death expected). The remaining five animals were
all alive and were subsequently euthanised. While those
rats found alive suggest some level of aversion, the trap
itself may have contributed to the aversion once triggered,
or alternatively they may have received a sub-lethal dose
and did not return to the bait. It is expected that some rats
did not encounter the live capture traps or chose to avoid
them (and the bait within).
If the second aerial toxic baiting had been carried
out, the total cost of the novel prescription (including all
prefeed and toxic baiting applications) is estimated at
approximately NZ$90/ha, excluding costs associated with
gaining regulatory approvals. There is potential for this
cost to reduce further with economies of scale and reduced
prefeed applications.
ZIP retested the hypothesis in a trial on the West Coast
of the South Island during the second half of 2017. After
the first phase of baiting (two prefeed applications, and
one toxin application using Wanganui #7 0.15% 1080 6g
double orange-masked cereal pellet (Orillion)), zero rats
and possums were detected over 83,410 detection nights
across 55 days post-toxin application (unpublished data).
The trial was deemed a success, and ended here.
Case study 2: Protect – the ‘virtual barrier’
The virtual barrier is a system that aims to exclude 99%
of rats, and 95% of possums that attempt to enter a protected
area. The virtual barrier being tested across the 2 km neck
at Bottle Rock Peninsula consists of multiple defence lines,
100 m apart, comprising kill (for rats) and live capture (for
possums) traps only, with no toxins currently deployed in
the system. Devices are placed at high intensity along each
defence line, one every 10 m, based on the assumption
that this spacing would ‘guarantee’ no animals could
breach the barrier without encountering a device, i.e. if the
target animal is on the ground it is never more than five
metres from a device as it passes through a line. Whether
they choose to interact with that device is another matter
entirely!
Possums
The most effective virtual barrier for possums tested
to date consisted of four lines of leg hold traps (PCR
#1, Pest Control Research) running across the peninsula
and a 400 m long line of leg hold traps running along the
central, prominent ridge through the barrier. The leg hold
traps are set in a custom-made platform raised 1.2 metres
above the ground (to avoid non-target captures of weka,
Gallirallus australis, a ground dwelling endemic rail). The
traps are visually lured with a plain white corflute card
(Connovation Ltd) nailed to the tree approximately 30 cm
above the platform. Each platform has a wooden ramp
attached, at 60° to the horizontal. In addition to preventing
weka access, alternating trials by ZIP have shown that
ramps improve trap effectiveness by 18% (95% C.I. [2.5%,
29%]) compared with non-ramped traps.
Traditionally, live-capture leg hold traps must be
physically inspected by the trapper every day in order to
comply with New Zealand animal welfare legislation. ZIP
has developed an automated, remote reporting system that
uses a magnetically switched trap transmitter to advise
that a trap has been sprung, via a 433 MHz ‘daisy chain’
and the Iridium satellite network. To date (May 2017),
the remote reporting system has been in service for more
than 580,000 trap nights and has remotely reported over
500 possum captures – there has not been a single false
negative in this time. In conjunction, the NZ Ministry for
Primary Industries has developed industry guidelines to
allow the automated reporting of live-capture traps, while
conforming to animal welfare standards as required by
law (MPI, 2016). This innovation has reduced the labour
cost of servicing the traps by 95%, with only sprung traps
needing to be checked by the trapper.
During the period from 26 November 2016 to 17 May
2017, the virtual barrier caught 127 possums, with at least
11 possums breaching the barrier; i.e. 8% ‘leakage’ (95%
confidence interval, [4%, 14%]). Leakage was determined
from the number of possums killed in the protected area
(beyond the barrier), using leg hold traps, set up as per
the barrier, but placed on a one per 50 ha density, divided
by the total number that attempted to breach the barrier
(number killed in the barrier plus number killed beyond it).
In addition, a detection network of 554 chew cards (selffilled as described in the removal case study), serviced
every three weeks, confirms the ongoing absence or
presence of possums in the protected area. On average,
approximately 18 possums/month attempted to cross the
2 km wide barrier, with 1.5 possums/month succeeding.
Improvements to the system have been identified, and
therefore future versions of the barrier are expected to
approach the target of ≤ 5% leakage.
607
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
Ship rats
The current virtual rat barrier at Bottle Rock consists of
six lines of ‘Tun200s’ (two DOC200 single action stainless
steel kill traps (CMI Springs), in custom built ‘run-through
tunnel’ wooden trap box). The wooden tunnels have a
72 mm diameter entrance hole and 265 mm long tunnel
leading to the kill plates from both ends, to avoid nontarget captures of weka which cannot fit inside the entrance
hole nor stretch out to reach the traps themselves (currently
<1 kill every 35,000 trap nights).
From 26 June to 26 October 2015, the virtual barrier
caught 160 ship rats, with at least nine rats breaching the
barrier; i.e. 5% leakage (95% confidence interval, [2.5%,
10%]). Leakage was estimated as described for possums
in the previous section (e.g. number of rats killed on 100
m x 60 m grid of single-set DOC150 kill traps (CMI
Springs) in ‘standard’ wooden boxes, placed throughout
the peninsula beyond the barrier), in conjunction with the
detection network of chew cards (as described above for
possums) confirming the absence or presence of rats in the
protected area. On average, approximately 40 ship rats/
month attempted to cross the 2 km wide barrier, with two
rats per month succeeding. All Tun200 traps were lured
with peanut butter (Goodnature Ltd) during this period.
We found no evidence to suggest that the effectiveness of
identically lured, multiple lines of Tun200 traps declined
with repeated presentation (effectiveness 40%, 95%
C.I [33%, 46%], for all Tun200 lines treated as samples
from the same population, that is irrespective of the line
placement).
A variety of alternative food lures have subsequently
been trialled including Nutella (Ferrero Australia Pty
Ltd), Colby cheese (Mainland Ltd), milk chocolate (J.H.
Whittaker and Sons Ltd), and peanut butter (Goodnature
Ltd, and Pic’s - Picot Productions Ltd). These lures
performed similarly and intercepted on average 36% (95%
C.I. [33%, 39%]) of rats, as measured by the percentage of
rats that breached each line.
Costs of the barrier
Including the cost of track cutting and installation, the
current capital cost of a multiple line, ship rat and possum
virtual barrier at Bottle Rock Peninsula is approximately
NZ$250/m (excl. GST). This cost is for a 20-year life, and
includes device replacement, remote reporting system, and
an automated lure dispenser (in development to further
reduce labour costs).
The annual operating cost is approximately NZ$20/m
(8% of capital cost).
Case study 3: Detection – a ‘minimal infrastructure’
system
Ship rats
Considerable effort has gone into understanding the
exploratory behaviour of invading rats in predator-free
spaces, with substantial individual variation identified in
the roaming behaviour (Russell, et al., 2005; Russell, et
al., 2008; Russell, et al., 2010; Innes, et al., 2011). Not
unexpectedly, the majority of this work has been focussed
on the individual, as current biosecurity detection systems
are tailored towards intensively targeting the individual
invader.
ZIP is conceptualising an alternative approach that
looks beyond the individual, and rather focusses on
the emergent population (if it happens). So long as the
incursion events are infrequent, if the invading rat is alone
and non-pregnant, then the scale of their individual impact
is expected to be small and impacts only begin to have
608
significance once a new population emerges (Norbury,
et al., 2015; Elliott & Kemp, 2016). This is the point of
intervention ZIP proposes to target.
Targeting the first generation (Generation One) of
a pregnant female provides up to 11 individuals, 10
juveniles plus mother (Innes, 2005) to trigger detection
devices, rather than the sole invader, greatly increasing
the chances of interaction. Furthermore, the anticipated
dispersal footprint of Generation One is likely to lend
itself to a minimal infrastructure network spacing (perhaps
one detection device every 20 ha, based on emerging data
from ZIP trials such as that below). This network could be
further tailored to be predominantly coastal and waterway
biased, to maximise the probability of encounter. In
addition, we estimate that we could have up to 100 days
to detect and remove the first generation of invaders,
before those juveniles reach sexual maturity and begin
breeding themselves (based on reproductive biology;
Innes, 2005). Conversely, this approach will require bigger
treatment areas to remove the entire emerging population.
The response could well be aerially based, rather than the
ground-based responses traditionally deployed for island
incursions.
A ZIP field trial is currently underway (during the
drafting of this paper) at the confluence of the Jackson
and Arawhata Rivers, South Westland (44°03’00” S,
168°43’32” E) whereby a mother ship rat and her offspring
have been released into an area of very low rat abundance
to observe their dispersal footprint. Early indications, based
on the distance between release point and subsequent trap
capture points, are that some individual offspring dispersed
at least 650 m from the natal den location by the time they
were 86 days old.
If the concept works, the capital cost of installing this
system today would be NZ$20/ha. The annual surveillance
cost would be NZ$4/ha (using an automated reporting
kill trap as the ‘sentinel’ detection device), with an annual
response cost of NZ$5/ha (assuming a leakage rate of
0.5%).
Possums
Possums, once isolated, roam over considerable ranges,
in the order of 50–100 ha (Sweetapple & Nugent, 2009;
OSPRI, 2014), presumably looking for other possums. If
possum incursions are infrequent, their slow breeding rates
(Cowan, 2005) and curiosity (Carey et al., 1997) suggest
that delayed detection and response may be all that is
necessary to prevent possum re-establishment.
ZIP is currently trialling a minimal ‘lethal detection’
network for possums at Bottle Rock Peninsula. Six leghold traps, deployed as in the virtual barrier (excluding
ramp) but spaced at approximately one per 50 ha, have been
established beyond the virtual barrier. In the 12 months
since its deployment, this network has prevented possum
reestablishment; with 17 possums caught to June 2017
(and no sustained detections on the ‘background’ chew
card network, as described in case study 2). The capital
cost of installing this system today would be NZ$10/ha,
with a current operating cost of detection and response of
approximately NZ$5/ha/annum.
Automated reporting system
To support these minimal infrastructure detection
networks, development of an automated system for near
real-time updates on the status of remove and protect
sites is continuing. ZIP has already developed the ability
to use daisy chain communication for short range data
transmission, e.g. trap lines in a barrier setting. However,
a landscape scale network will require a different
transmission technology – one that can transmit reliably
Bell, et al.: The remove and protect model
over large distances, in rugged or forested terrain (Jones,
et al., 2015). Recent advances in the international
telecommunications industry are seeing the emergence
of low powered, long range radio technology (LoRa).
A small number of sensitive receiving stations allows
the use of many battery-powered transmitters across a
landscape. LoRa, used in combination with satellite-based
communications, is likely to be the platform technology on
which to build an incursion notification system for these
remote networks.
CONCLUSION
The New Zealand Government has announced the goal
of being predator-free by 2050. Momentum is building
on this goal, with the Predator Free 2050 Ltd company
established with a board of directors to guide strategic
investment into projects of significance (Anon., 2016).
While New Zealand has an internationally enviable track
record in island eradications and developed the predator
fenced sanctuary approach, these methodologies cannot be
scaled on the mainland.
It is widely acknowledged that new technologies,
along with a shift in operating model and cost structure,
will be required to completely eradicate predators from the
mainland. Such a shift from the suppression paradigm could
utilise the remove and protect model, where peninsulas
are able to be converted into ‘islands’ for eradication
operations. Zero Invasive Predators (ZIP), a not-for-profit
research and development company founded in 2015, is
helping to develop the techniques required to enable this
model on the mainland.
Further trials are underway to use a novel prescription
of dual aerial 1080 operations to drive initial removal at a
cost of less than NZ$100/ha (with no more than two prefeed
applications per toxin application). In-forest capability
exists now to intercept over 95% of all rats and possums
using a virtual barrier at a capital cost of approximately
NZ$250/m and an annual operating cost of less than
NZ$40/m. The initial testing of a minimal infrastructure
detection system shows promising signs of success. Large
social strides are still required to make predator-free New
Zealand a reality, but the first tentative technical steps are
being taken now.
ACKNOWLEDGEMENTS
The authors, and wider ZIP team, would like to thank all
of our investors for their support (financial and otherwise).
Thank you to the DOC Sounds team for their support of
the work being undertaken at Bottle Rock Peninsula. The
DOC Taranaki team were monumental in helping with
the ‘1080 to Zero’ and ‘Generation One’ trials on the
Mounga – thank you. The Generation One rat release trial
was approved by the Lincoln University Animal Ethics
Committee and conducted under approval #2016-09.
Thanks to James Russell for guidance on the structure and
concept behind this paper. Al Bramley and Charlotte Bell
(and two anonymous reviewers) provided many useful
comments on earlier versions of this manuscript, which
greatly improved this paper.
REFERENCES
Anon. (2016). Predator Free 2050 Ltd. Board Appointed. Press release
30 November 2016. Wellington: Office of the Minister of Conservation.
Bentley, E.W. and Taylor, E.J. (1965). ‘Growth of laboratory-reared ship
rats (Rattus rattus L.)’. Annals of Applied Biology 55: 193–205.
Broome, K., Cox, A., Golding, C., Cromarty, P., Bell, P. and McClelland,
P. (2014). Rat Eradication Using Aerial Baiting: Current Agreed Best
Practice Used in New Zealand (Version 3.0). Wellington: Department
of Conservation.
Brown, K., Elliott, G., Innes, J. and Kemp, J. (2015). Ship Rat, Stoat and
Possum Control on Mainland New Zealand: An Overview of Techniques,
Successes and Challenges. Wellington: Department of Conservation.
Burns, B., Innes, J. and Day, T. (2012). ‘The Use and Potential of Pestproof Fencing for Ecosystem Restoration and Fauna Conservation
in New Zealand’. In: M. Somers and M. Hayward (eds.) Fencing for
Conservation: Restriction of Evolutionary Potential or a Riposte to
Threatening Processes? pp. 65–90. New York: Springer-Verlag.
Cabinet (2016). Accelerating Predator Free New Zealand. CAB-16MIN-0335. Wellington: Cabinet Economic Growth and Infrastructure
Committee.
Cabrera, D., Colosi, L. and Lobdell, C. (2008). ‘Systems thinking’.
Evaluation and Program Planning 31: 299–310.
Carey, P.W., O’Connor, C.E., McDonald, R.M. and Matthews, L.R.
(1997). ‘Comparison of the attractiveness of acoustic and visual stimuli
for brushtail possums’. New Zealand Journal of Zoology 24: 273–276.
Clapperton, B.K. and Day, T.D. (2001). ‘Cost-effectiveness of exclusion
fencing for stoat and other pest control compared with conventional
control’. DOC Science Internal Series 14: 5–19.
Clout, M.N. (2011). ‘Ecological and Economic Costs of Alien Vertebrates
in New Zealand’. In: D. Pimentel (ed.) Biological Invasions: Economic
and Environmental Costs of Alien Plant, Animal, and Microbe Species,
pp. 283–291. Boca Raton: CRC Press.
Coleman, J. and Caley, P. (2000). ‘Possums as a Reservoir of Bovine
TB’. In: T. L. Montague (ed.) The Brushtail Possum. Biology, Impact
and Management of an Introduced Marsupial, pp. 92–104. Lincoln:
Manaaki Whenua Press.
Cowan, P.E. (2005). ‘Brushtail Possum’. In: C.M. King (ed.) The
Handbook of New Zealand Mammals, pp. 56–80. Auckland: Oxford
University Press.
Curnow, M. and Kerr, G.N. (2017). Predator Free Banks Peninsula:
Scoping Analysis. LEaP Research Report No. 44. Lincoln: Lincoln
University.
Elliott, G. and Kemp, J. (2016). ‘Large-scale pest control in New Zealand
beech forests’. Ecological Management and Restoration 17: 200–209.
Epro Ltd. (2006). Code of Practice: Aerial and Hand Broadcast of Pestoff
Rodent Bait 20R for the Intended Eradication of Rodents from Specified
Areas of New Zealand, Wellington: New Zealand Food Safety Authority.
Gillies, C., Campbell, J., Marsh, N. and Gembitsky, M. (2003). ‘Seasonal
Differences in Bait Acceptance by Forest-dwelling Rats Following
Simulated Aerial 1080 Possum Control Operations in New Zealand:
Interim Results’. In: G. Singleton, L. Hinds, C. Krebs and D. Spratt
(eds.) Rats, Mice and People: Rodent Biology and Management.
ACIAR Monograph 96, pp. 343–345. Canberra: Australian Centre for
International Agricultural Research.
Griffiths, J.W. and Barron, M.C. (2016). ‘Spatiotemporal changes
in relative rat (Rattus rattus) abundance following large-scale pest
control’. New Zealand Journal of Ecology 40: 371–380.
Hegg, D., MacKenzie, D.I. and Jamieson, I.G. (2013). ‘Use of Bayesian
population viability analysis to assess multiple management decisions
in the recovery programme for the endangered takahe Porphyrio
hochstetteri’. Oryx 47: 144–152.
Hitchmough, R. (2013). ‘Summary of Changes to the Conservation Status
of Taxa in the 2008–11 New Zealand Threat Classification System
Listing Cycle’. New Zealand Threat Classification Series 1. Wellington:
Department of Conservation.
Howald, G., Donlan, C.J., Galvan, J.P., Russell, J.C., Parkes, J.,
Samaniego, A., Wang, Y., Veitch, D., Genovesi, P., Pascal, M., Saunders,
A. and Tershy, B. (2007). ‘Invasive rodent eradication on islands’.
Conservation Biology 21: 1258–1268.
Innes, J. (2005). ‘Ship Rat’. In: C. M. King (ed.) Handbook of New
Zealand Mammals, pp. 187–203. Auckland: Oxford University Press.
Innes, J., Watts, C., Fitzgerald, N.L., Thornburrow, D., Burns, B., MacKay,
J. and Speedy, C. (2011). ‘Behaviour of invader ship rats experimentally
released behind a pest-proof fence, Maungatautari, New Zealand’. In:
C.R. Veitch, M.N. Clout, and D.R. Towns (eds.) Island invasives:
eradication and management, pp. 437–440. Occassional Paper SSC no.
42. Gland, Switzerland: IUCN and Auckland, New Zealand: CBB.
Innes, J., Lee, W.G., Burns, B., Campbell-Hunt, C. and Watts, C. (2012).
‘Role of predator-proof fences in restoring New Zealand’s biodiversity:
a response to Scofield, et al. (2011)’. New Zealand Journal of Ecology
36: 232–238.
Ismar, S.M.H., Baird, K.A., Gaskin, C.P., Taylor, G.A., Tennyson, A.J.D.,
Rayner, M.J., Bettesworth, D., Fitzgerald, N., Landers, T.J. and Imber,
M.J. (2014). ‘A case of natural recovery after the removal of invasive
predators – community assemblage changes in the avifauna of Burgess
Island’. Notornis 61: 188–195.
609
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
Jones, C., Warburton, B., Carver, J. and Carver, D. (2015). ‘Potential
applications of wireless sensor networks for wildlife trapping and
monitoring programs’. Wildlife Society Bulletin 39: 341–348.
King, C.M., Veale, A., Patty, B. and Hayward, L. (2014). ‘Swimming
capabilities of stoats and the threat to inshore sanctuaries’. Biological
Invasions 16: 987–995.
McFadden, I. and Towns, D. (1991). Eradication Campaigns Against
Kiore (Rattus exulans) on Rurima Rocks and Korapuki, Northern New
Zealand. Science and Research Internal Report No 97. Wellington:
Department of Conservation.
Moors, P.J. (1985). ‘Eradication campaigns against Rattus norvegicus on
the Noises Islands, New Zealand, using brodifacoum and 1080’. ICBP
Technical Publication 3: 143–155.
Morgan, D.R. (2004). ‘Maximising the Effectiveness of Aerial 1080
Control of Possums (Trichosurus vulpecula)’. PhD thesis. Lincoln:
Lincoln University.
MPI (2016). Remote Monitoring of Live Capture Traps for Vertebrates:
Guidelines for Good Practice. Wellington: Ministry for Primary
Industries.
Murphy, E.C. and Dowding, J.E. (1994). ‘Range and diet of stoats
(Mustela erminea) in a New Zealand beech forest’. New Zealand
Journal of Ecology 18: 11–18.
Murphy, E.C. and Dowding, J.E. (1995). ‘Ecology of the stoat in
Nothofagus forest: home range, habitat use and diet at different stages
of the beech mast cycle’. New Zealand Journal of Ecology 19: 97–109.
Myers, N., Mittermeier, R.A., Mittermeier, C.G., da Fonseca, G.A.B.
and Kent, J. (2000). ‘Biodiversity hotspots for conservation priorities’.
Nature 403: 853–858.
Norbury, G., Hutcheson, A., Reardon, J. and Daigneault, A. (2014). ‘Pest
fencing or pest trapping: a bio-economic analysis of cost-effectiveness’.
Austral Ecology 39: 795–807.
Norbury, G.L., Pech, R.P., Byrom, A.E. and Innes, J. (2015). ‘Densityimpact functions for terrestrial vertebrate pests and indigenous biota:
guidelines for conservation managers’. Biological Conservation 191:
409–420.
Nugent, G., Morgan, D., Clayton, R. and Warburton, B. (2010).
‘Improving the efficacy of aerial poisoning of brushtail possums
(Trichosurus vulpecula) through reduced fragmentation of bait’.
International Journal of Pest Management 57: 51–59.
Nugent, G., Warburton, B., Thomson, C., Sweetapple, P. and Ruscoe,
W.C. (2011). ‘Effect of prefeeding, sowing rate and sowing pattern
on efficacy of aerial 1080 poisoning of small-mammal pests in New
Zealand’. Wildlife Research 38: 249–259.
Nugent, G., Twigg, L.E., Warburton, B., McGlinchy, A., Fisher, P.,
Gormley, A.M. and Parkes, J.P. (2012). ‘Why 0.02 %? A review of
the basis for current practice in aerial 1080 baiting of rabbits in New
Zealand’. Wildlife Research 39: 89–103.
OSPRI (2014). Our First Year|2013/2014 OSPRI Annual Report and
TBfree Research and Disease Control Report. Wellington: OSPRI.
Parkes, J., Fisher, P. and Forrester, G. (2011). ‘Diagnosing the cause of
failure to eradicate introduced rodents on islands: brodifacoum versus
diphacinone and method of bait delivery’. Conservation Evidence 8:
100–106.
Parkes, J.P., Nugent, G., Forsyth, D.M., Byrom, A.E., Pech, R.P.,
Warburton, B. and Choquenot, D. (2017a). ‘Past, present and two
potential futures for managing New Zealand’s mammalian pests’. New
Zealand Journal of Ecology 41: 151–161.
Parkes, J.P., Byrom, A. and Edge, K.-A. (2017b). ‘Eradicating mammals
on New Zealand island reserves: what is left to do?’ New Zealand
Journal of Ecology 41: 1–8.
Pascal, M., Lorvelec, O., Bretagnolle, V. and Culioli, J-M. (2008).
‘Improving the breeding success of a colonial seabird: a costbenefit comparison of the eradication and control of its rat predator’.
Endangered Species Research 4: 267–276.
Ross, J.G., Hickling, G.J., Morgan, D.R. and Eason, C.T. (2000). ‘The role
of non-toxic prefeed and postfeed in the development and maintenance
of 1080 bait shyness in captive brushtail possums’. Wildlife Research
27(1): 69–74.
Royal Society of New Zealand (2014). ‘Challenges for Pest Management
in New Zealand’. Emerging Issues March 2014.
Russell, J.C., Towns, D.R., Anderson, S.H. and Clout, M.N. (2005).
‘Intercepting the first rat ashore’. Nature 437: 1107.
Russell, J.C., Beaven, B.M., MacKay, J.W.B., Towns, D.R. and Clout,
M.N. (2008). ‘Testing island biosecurity systems for invasive rats’.
Wildlife Research 35: 215–221.
Russell, J.C., McMorland, A.J.C. and MacKay, J.W.B. (2010).
‘Exploratory behaviour of colonizing rats in novel environments’.
Animal Behaviour 79: 159–164.
610
Russell, J.C., Innes, J.G., Brown, P.H. and Byrom, A.E. (2015). ‘Predatorfree New Zealand: Conservation country’. BioScience 65: 520–525.
Scofield, R.P., Cullen, R. and Wang, M. (2011). ‘Are predator-proof
fences the answer to New Zealand’s terrestrial faunal biodiversity
crisis?’ New Zealand Journal of Ecology 35: 312–317.
Scofield, R.P. and Cullen, R. (2012). ‘Fenced sanctuaries need critical
evaluation: a reply to Innes, et al. (2012)’. New Zealand Journal of
Ecology 36: 239–242.
Sweetapple, P. and Nugent, G. (2009). ‘Possum demographics and
distribution after reduction to near-zero density’. New Zealand Journal
of Zoology 36: 461–471.
Towns, D.R. and Broome, K.G. (2003). ‘From small Maria to massive
Campbell: forty years of rat eradications from New Zealand islands’.
New Zealand Journal of Zoology 30: 377–398.
Towns, D.R., Borrelle, S.B., Thoreson, J., Buxton, R.T. and Evans, A.
(2016). ‘Mercury Islands and their role in understanding seabird island
restoration’. New Zealand Journal of Ecology 40: 235–249.
Veltman, C.J. and Pinder, D.N. (2001). ‘Brushtail possum mortality and
ambient temperatures following aerial poisoning using 1080’. Journal
of Wildlife Management 65: 476–481.
R. Griffiths, S. Cranwell, D. Derand, T. Ghestemme, D. Will, J. Zito, T. Hall, M. Pott and G. Coulston
Griffiths, R.; S. Cranwell, D. Derand, T. Ghestemme, D. Will, J. Zito, T. Hall, M. Pott and G. Coulston.
Multi island, multi invasive species eradication in French Polynesia demonstrates economies of scale
Multi island, multi invasive species eradication in French Polynesia
demonstrates economies of scale
1
R. Griffiths , S. Cranwell2, D. Derand2, T. Ghestemme3, D. Will4, J. Zito4, T. Hall4, M. Pott4 and G. Coulston5
2
1
Island Conservation, Warkworth, New Zealand. <rgriffiths@islandconservation.org>. BirdLife International, Pacific
3
4
Secretariat, Fiji. Société d’Ornithologie de Polynésie, Tahiti, French Polynesia. Island Conservation, Santa Cruz, CA,
USA. 5Good Wood Aotearoa Ltd, Whangarei, New Zealand.
Abstract Eradication of invasive vertebrates on islands has proven to be one of the most effective returns on investment
for biodiversity conservation. To recover populations of the critically endangered Polynesian ground dove (Gallicolumba
erythroptera), the endangered white-throated storm-petrel (Nesofregetta fuliginosa), the endangered Tuamotu sandpiper
(Prosobonia cancellata) as well as other native plant and animal species, a project was undertaken to eradicate five species
of invasive alien vertebrates: Pacific rat (Rattus exulans), ship rat (R. rattus), feral cat (Felis catus), rabbit (Oryctolagus
cuniculus) and goat (Capra hircus), on six islands spanning 320 km of open ocean in the Tuamotu and Gambier
Archipelagos of French Polynesia. Using a ship to deliver supplies and equipment, a helicopter for offloading and bait
application, and ground teams for follow up trapping and hunting, invasive vertebrates were successfully removed from
five of the six islands. Pacific rats survived at one site. The project was planned and executed by a partnership consisting
of international and local conservation NGO’s, working together with local communities. Combining the different
eradication operations into one expedition added complexity to project planning and implementation and increased the
risk of the operation failing on any one island but generated greater returns on investment allowing six islands to be
targeted at significantly less cost than if each island had been completed individually. An extensive and thorough planning
effort, effective relationships with local stakeholders and communities, a good operational strategy and a partnership of
stakeholders that each brought complementary capacities to the project contributed to its success.
Keywords: cat, conservation, goat, rabbit, rat, restoration, threatened species recovery
INTRODUCTION
The removal of alien species from islands, especially
invasive vertebrates, offers one of the best returns on
investment for the protection of indigenous biodiversity
(Donlan & Wilcox, 2008; Genovesi, 2011). There is
now a growing list of island species no longer regarded
as endangered because a key invasive species threat has
been lifted (Russell, et al., 2016). The San Nicolas island
night lizard (Xantusia riversiana) (Rice & Clark, 2016),
the Seychelles magpie robin (Copsychus sechellarum)
(Burt, et al., 2016) and the northern tuatara (Sphenodon
punctatus) (Towns, et al., 2016) are just three examples of
the many species whose threat status has been downgraded
to a more secure category as a consequence (IUCN, 2010).
Eradication projects can be expensive (Simberloff,
2002). The remote nature of many islands and the necessity
to target every individual within a population requires
extensive planning effort, meticulous execution (Cromarty,
et al., 2002) and resourcing that often exceeds the means of
a single organisation. For many island nations, eradication
projects are simply unaffordable and for some projects, the
cost may exceed the annual environmental expenditure of
an entire country.
French Polynesia is an overseas collectivity (political
unit) of the French Republic. It is composed of 118
geographically dispersed islands and atolls scattered over
an expanse of more than 5,030,000 km2 in the South Pacific
Ocean. Like many other tropical island archipelagos,
French Polynesia is biologically rich and its remoteness
has led its flora and fauna to be characterised by high levels
of endemism (Gillespie, et al., 2008; Meyer & Butaud,
2009). Sixty three percent of its plants and 72% of its birds
are found nowhere else (Gillespie, et al., 2008; Meyer &
Butaud, 2009). As witnessed elsewhere, French Polynesia
has been severely affected by habitat loss and invasive
species. Nineteen of its bird species have become extinct
since the 16th century and of the 25 surviving endemic
birds, 18 are listed as threatened and five as critically
endangered (Zarzoso-Lacoste, 2013).
Invasive vertebrates are widely considered the most
significant threat to French Polynesia’s avifauna (ZarzosoLacoste, 2013). Interventions, to remove invasive
vertebrates, could be made to improve security for many
species. However, investment within the collectivity for
the management of invasive alien species remains small
and a national invasive species strategy has not yet been
developed. The collectivity does not appear to have the
financial mechanisms to undertake vertebrate eradications
and outside financial support will be required if species
extinctions are to be avoided.
In 2015, four species of invasive alien vertebrates,
Pacific rat (Rattus exulans), ship rat (R. rattus), feral cat
(Felis catus) and rabbit (Oryctolagus cuniculus), were
successfully removed from five of six islands spanning
320 km of open ocean in the Tuamotu and Gambier
Archipelagos of French Polynesia. The project failed to
remove rats from one project site and completion of goat
(Capra hircus) eradication from another was delayed until
2017.
Here we describe the methods used to remove invasive
vertebrates from the project sites and the logistics associated
with the project. We explain how cost efficiencies were
gained by combining operations and targeting multiple
islands and define how the project partnership was
instrumental to the project’s success.
METHODS
Site description
Six islands in the south-east of French Polynesia were
targeted for the removal of invasive vertebrates (Fig. 1).
These included the two atolls of Vahanga and Tenania
(Tenarunga) that, together with Tenararo and Matureivavao,
make up the Acteon Island Group. Vahanga and Tenararo
are identified as a Key Biodiversity Area (Atherton, 2007)
and as an Important Bird Area (Raust & Sanford, 2007).
Tenararo is one of four islands in French Polynesia never
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 611–617. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
611
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
part of the year for copra harvesting and Temoe is regularly
visited by local fisherman. Table 1 summarises the general
characteristics of the six sites.
Project feasibility and planning
Planning for rat eradication on Vahanga began in
2006 after a previous attempt undertaken in 2000–2001
was confirmed as having failed (Pierce, et al., 2006).
Research was completed on the atoll to quantify the impact
of terrestrial crabs on rodent bait availability, assess bait
uptake by rats and quantify the effort required to hand
broadcast bait across Vahanga (Griffiths, et al., 2011).
Following this, an operational plan for rat eradication on
the island was prepared (Broome, et al., 2011), but lack of
funding delayed the project’s implementation. A feasibility
assessment for the removal of rats from Kamaka and
Makaroa, completed in 2008 (Faulquier, 2008), stipulated
the need for a helicopter due to the steep topography of
these sites.
Fig. 1 Location of the six project sites within French
Polynesia.
to have had invasive vertebrates and the atoll remains
a stronghold for the critically endangered Polynesian
ground dove and the endangered Tuamotu sandpiper as
well as other native species (Blanvillain, et al., 2002). The
operation also included Makaroa, Kamaka and Manui,
three of the higher elevation islands that form Mangareva
atoll complex and Temoe atoll, which lies to the south of
Mangareva. Makaroa, Kamaka and Manui, together with
the pest-free Motu Teiku, are classified as an IBA (Raust
& Sanford, 2007). The endangered white-throated storm
petrel breeds on both Manui and Motu Teiku.
All three of the atolls included in the operation were
planted with coconut (Cocos nucifera) and used historically
for copra production although only Tenania continues to
be used for this purpose. Consequently, although areas of
indigenous vegetation remain, C. nucifera dominates many
of the forested parts of the atolls. Makaroa, Kamaka and
Manui have also been extensively modified by burning,
and the introduction of herbivores such as goats and
rabbits. At the time of the project, little ground cover
existed on Makaroa and large areas of Kamaka and Manui
were covered in the invasive molasses grass (Melinus
minutiflora). Of the six targeted sites, only the islet of
Kamaka in the Gambier group is permanently inhabited.
The atoll of Tenania in the Acteon group is occupied for
Following several high profile rat eradication failures
on tropical islands, a global review of eradication methods
was undertaken in 2013 to increase success rates (Russell
& Holmes, 2015). New best practice guidelines were
published, recommending higher bait application rates and
longer periods of bait availability (Keitt, et al., 2015). The
new guidelines meant it would be extremely challenging
logistically to complete a ground-based operation for
Vahanga and the use of a helicopter was recommended.
Funding for rat eradication on Vahanga was eventually
obtained in 2014. However, due to the costs associated
with transporting a helicopter to the south-east corner of
French Polynesia and the relatively low cost of including
additional sites and invasive species, a decision was
made by project partners to target five additional, high
conservation value islands in the area. This decision was
facilitated by broadening the project partnership and
securing additional funding. An operational plan was
devised that prescribed the aerial application of rodent bait
containing brodifacoum to target rats followed by trapping
and hunting to target cats, rabbits and goats across the six
project sites (Derand, et al., 2015).
The target bait application rate for rat eradication was
derived using the methods described by Pott et al. (2015)
to interpret bait availability data collected by Griffiths et al.
(2011). The proposed application rate, coupled with reported
island sizes and areas derived from available satellite
imagery and a 15% contingency for lost or damaged bait,
were then used to estimate the total amount of bait required.
Immediately before the project’s implementation, higher
resolution satellite imagery acquired from the Millennium
Coral Reef Mapping Project (Andréfouët, et al., 2005)
Table 1 Characteristics of the six sites targeted for invasive vertebrate removal.
Island
Vahanga
Area
(ha)
380
Tenania
425
5
Kamaka
Makaroa
Manui
Temoe
58
22
8
431
166
136
54
5
612
Elevation Location
(m)
5
Acteon
Acteon
Gambier
Gambier
Gambier
Gambier
Native threatened species expected to benefit Targeted invasives
Pacific ground-dove, Tuamotu sandpiper, atoll
fruit-dove (Ptilinopus coralensis), Murphy’s
petrel (Pterodroma ultima), bristle-thighed
curlew (Numenius tahitiensis), green turtle
(Chelonia mydas)
Pacific Ground-dove, Tuamotu sandpiper,
bristle-thighed curlew, green turtle
Polynesian storm petrel, Murphy’s petrel
Polynesian storm petrel
Polynesian storm petrel, Murphy’s petrel
Murphy’s petrel
Rattus exulans
R. exulans, R. rattus,
Felis catus
R. exulans
R. exulans, Capra hircus
Oryctolagus cuniculus
R. exulans
Griffiths, et al.: Multi invasive species eradications, French Polynesia
and EVS-Islands digital earth imagery showed that initial
estimates of island areas had been overestimated, in some
cases by as much as 22%. Consequently, this made more
bait available for distribution at each site than had been
planned.
Project implementation
Staging
Ninety-two tonnes of rodent bait, 30,000 l of Jet
A1 helicopter fuel, three bait-spreading buckets, and
equipment and supplies necessary for the project were
shipped from the port of Papeete to the project sites by the
coastal freighter ‘Nuku Hau’. Rodent bait was transported
in 22.7 kg paper-walled sacks stacked inside Ox Boxes
(waxed cardboard pods) (hereafter referred to as pods) and
the fuel in 200 l drums. A single-engine Squirrel AS350 B2
supplied by Tahiti Helicopters was flown from Papeete by
‘island hopping’ between four intermediate islands (a total
distance of 1500 km) before converging with the Nuku
Hau at Vahanga to commence the offloading process.
Immediately prior to unloading, the island’s coastal
boundaries were flown to confirm the size of the area to
be treated and revalidate the amount of bait and fuel to be
unloaded at each site. All equipment and supplies were first
offloaded from the Nuku Hau to a small barge which was
then unloaded by helicopter in separate sling loads. Bait
and fuel, sufficient for each atoll, were staged on Vahanga,
Tenania, and Temoe with unloading taking between 4–6
hours for each atoll. Supplies for the three closely grouped
Gambier Islets were staged in less than four hours on
Kamaka. To minimise flying between island groups,
two bait spreading buckets were offloaded in the Acteon
Group, one for use on Vahanga and the other for Tenania.
One bucket and a range of spare parts were stationed in the
Gambier Islands for use on both Temoe and the Gambier
Islets.
Project team members, 24 in total, were also deployed
at this time. Team members were stationed on Vahanga
(6), Tenania (5), Temoe (3) and Kamaka (1). The project
manager, GIS analyst, baiting team (3), pilots (2) and
helicopter mechanic travelled by helicopter between the
islands to complete bait applications and one person,
stationed in Mangareva, provided logistical support. In
between bait applications, the project manager, GIS analyst
and members of the baiting team deployed to different
islands to provide support for monitoring, trapping and
hunting.
On Tenania, large piles of broken coconut husks
containing coconut flesh were found across the atoll.
These byproducts of the recent copra harvest represented a
signficant alternative food source for rats and a risk to the
project’s likelihood of success. To reduce risk, members of
the project team systematically burned piles of coconuts.
This laborious activity greatly reduced the amount of
coconut available to rats but did not eliminate it.
Bait application
After staging was complete, bait application took place
sequentially beginning on Temoe followed by the Gambier
Islets, Vahanga and finally Tenania. Each of the three atolls
took more than one day to complete due to the amount
of bait applied and the requirement to break the circular
atolls into multiple blocks. Dividing the operational area
into blocks maximised the length of flight lines that could
be flown thereby simplifying the operation for the pilot.
Adjacent baiting swaths were overlapped by 50% to reduce
the risk of gaps in coverage (e.g. Fig. 2). In addition to
parallel flight lines across each island’s interior, a swath
with a deflector bucket (which spreads bait in one direction
Fig. 2 First bait application completed on Vahanga.
only) was completed along the edge of both coastal
and lagoon vegetation. Additional bait was applied by
helicopter over areas considered to be higher risk, such as
areas of human habitation or sites known to support the
highest crab densities. At the same time as bait was applied
by helicopter, rodent bait was placed in small dishes within
all buildings still in use and scattered by hand underneath
buildings and inside all derelict or abandoned structures.
Following an 18 day interval, a second application
of bait was completed at the project sites in the same
sequence. The length of the interval was dictated by a
desire to ensure that all individuals (including juveniles)
within the targeted rat populations were exposed to
bait, as discussed in Keitt, et al. (2015), and also by the
resource limitations of the project partnership. Operational
specifications for the second bait application were the
same, except for the exclusion from bait application of
barren storm-washed coral habitat across all three atolls.
Bait availability monitoring and anecdotal observations
suggested negligible disappearance of bait from these
atolls and thus no advantage in re-treating these areas.
This action was also seen as a means of reducing risk to
non-target species such as Tuamotu sandpiper (Prosobonia
cancellata). Operational areas treated were thus smaller in
the second application for Vahanga, Tenania and Temoe
(Table 2). Dates of bait application and the application
rates achieved are provided in Table 2. No significant
delays because of sustained rainfall or excessive winds
were encountered.
Loading of bait spreading buckets was undertaken from
platforms constructed from an 18 mm thick plywood sheet
set atop two pods. A second plywood sheet was placed on
the ground in front of these pods to ensure a level footing
for the spreader bucket. The helicopter was fitted with a
VHF radio for ground to air communications with the bait
loading team.
The pod and pallet containment system withstood
crushing (some pods were stacked up to seven high in the
hold of the Nuku Hau), being dipped in saltwater (as they
were airlifted onto the islands), tropical temperatures and
periods of heavy rain. Water was found inside the internal
plastic bag (used to protect sacks of bait) in just four pods
and of these pods only the bags at the bottom of the pod
were affected. Only four of the 4,065 bags (<0.1%) of
bait shipped were considered unfit for application. Water
ingress into pods was primarily a result of damage incurred
to the cardboard during shipping and unloading, coupled
613
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
Average bucket
sow rate (kg/ha)*
15,111.9
16,433.5
31,545.4
429.1
341.5
429.1
24.4
29.9
26.8
June 10
June 28
Gambier
Gambier
1st
2nd
3,797.8
2,986.9
6,784.7
88.2
86.6
88.2
32.3
21.1
26.8
June 12–13
July 3
Vahanga
Vahanga
1st
2nd
11,715.5
14,272.8
25,988.3
382.6
333.3
382.6
21.8
27.7
24.5
June 14
July 4–5
Tenania
Tenania
1st
2nd
13,479.3
14,315.3
27,794.6
419.6
394
419.6
24.3
23.5
23.9
Average ground
application rate
(kg/ha)+
Island area
treated (ha)
1st
2nd
Bait spread
(T/hr)
Bait used (kg)
Temoe
Temoe
Operational
hours
Application
June 8–9
June 26–27
Date
Island
Table 2 Bait application summary.
11
9
20
1.37
1.83
1.57
35.2
48.1
73.5
1.27
1.09
1.18
43.1
34.5
76.9
7.5
6.5
14
1.56
2.20
1.86
30.6
42.8
67.9
9
8
17
1.50
1.79
1.64
32.1
36.3
66.2
3
2.75
5.75
*Average rate at which bait was spread from the bucket (bait used/TracMap recorded area). +Average rate at which bait was
available on the ground (bait used/island area treated).
with water pooling on the lid of the pod. Intact pods showed
no sign of water ingress despite water pooling.
Rabbits
Based on the results of other projects (e.g. Griffiths, et
al., 2014), most rabbits were expected to consume rodent
bait and succumb to poisoning on Manui. This proved to
be the case, with just four survivors found and one of these
appeared to be close to death at the time it was shot. Two
staff began follow-up work targeting surviving rabbits,
nine days after the first application of bait, to eliminate
survivors before the team departed French Polynesia.
Fresh rabbit sign was found in three discrete locations
during the period of follow up searching. In each case
the discovery of fresh sign led to the location of freshly
dead or surviving rabbits within the same area. One adult
female was found during the day on 19 June and shot. This
individual superficially appeared to be in good condition
but was presumed to be in the last stages of anticoagulant
All accessible areas of the island offering apparently
suitable habitat were searched for sign and surviving
rabbits during the day and at night, using powerful head
lamps (see Fig. 3). Some inaccessible parts of the island
such as cliff faces were searched with spotlights at night,
but comprehensive searching of these areas was not
possible. To manage search effort and spatial data, the
island was divided into zones. Generally, the same zone
was searched during the day and then again at night. Areas
where fresh sign was found, or a live rabbit sighted were
searched more intensively. Search effort was logged using
handheld GPS and a map used to identify areas that had not
yet been visited. Waypoints were recorded for any fresh
sign found and live rabbits sighted.
Two trail cameras were established on the island
from 26 April 2015 and seven added from 10 June. Three
cameras were kept at the same locations throughout the
operation while the remaining four were moved to locations
where fresh sign was found or where rabbit presence was
suspected. Old rabbit sign (faeces and chewed vegetation)
was found in most parts of the island except within the
molasses grass sward and the coastal littoral zone. Ten
freshly dead rabbits (presumed poisoned) were found.
Carcasses were found in three discrete locations and were
generally associated with areas of concentrated old sign.
614
Fig. 3 Search effort and location of surviving rabbits on
Manui 10 June to 3 July 2015.
Griffiths, et al.: Multi invasive species eradications, French Polynesia
poisoning as it did not move when approached and,
although not evident in the gut, the lower intestine was full
of blood.
Two young rabbits (a male and female) were found
together while spotlighting and shot.. These individuals
appeared healthy and a necropsy indicated no evidence
of bait ingestion or anticoagulant poisoning. The last
surviving rabbit was shot 100 m further north also after
finding fresh sign. This individual, an adult female, was in
excellent condition and showed no sign of bait ingestion
or anticoagulant poisoning. Following removal of this
individual, no further fresh sign or images on cameras were
found during five more days of search effort.
Cats
Cat trapping on Tenania began on 6 June prior to the 1st
bait application and continued until 4 July when the team
departed. A total of 564 trap nights (sum of the number of
active traps for each trap night) were achieved. Trapping
was conducted primarily with leg hold traps, No. 2 Bridger
padded, and No. 1.5 Oneida/Victor unpadded leg hold traps
set in a combination of cubby, trail and bucket sets. Traps
were baited with either canned or fresh fish, lured with a
commercial lure or left un-baited in the case of some trail
sets. Every 2–3 days, traps that had sprung were triggered
and reset. Trap locations were changed as necessary when
sign (tracks, scat, and visual observation) was encountered
in the field and camera data collected.
Traps were raised (i.e. positioned on top of sand filled
buckets) or left unlured to minimise crab interference.
However, traps were often triggered by what remaining sign
indicated was crabs, mostly Coenobita perlatus. Spotlight
surveys were undertaken but only one cat detected using
this method and this method was not pursued. A total of
10 remote trail cameras were installed on 5 June and data
collected daily, in most cases, to inform trap placement.
Monitoring with cameras continued until 4 July. Camera
data were also used to verify the identity of captured cats.
Ten distinct individual cats were detected with trail cameras
and, of these, nine were caught: five female and four male.
All were mature adults and one female was pregnant with
three foetuses at the time of capture. Feral cat captures were
made exclusively with leg-hold traps; three were caught in
cubby sets, five in trail sets and one in a bucket set. The last
cat captured displayed signs of internal haemorrhaging,
likely due to secondary exposure to brodifacoum. The
last feral cat detected by trail camera on 29 June was not
captured, despite concentrated trapping in the vicinity of
detection, and is assumed to have succumbed to secondary
poisoning. An individual with distinct black and white fur
patterns, seen during the first spotlight survey, was also
never observed again despite follow-up surveys.
All cats captured appeared to be in excellent body
condition. When the captured cats’ stomach contents were
examined, the only prey remains observed were rodents.
Interestingly, the stomachs of two cats captured contained
coconut flesh. Rat remains encapsulating the observed
coconut within one individual, indicated rats to be the
source; however, the other had its stomach completely full
of coconut.
Goats
Despite local reports to the contrary, eight goats were
still present on Makaroa in 2015 at the time of the project’s
implementation. One of these (a young female) was shot,
but further hunting effort was abandoned due to insufficient
capacity, a lack of suitable firearms and the remaining goats
being extremely wary due to having been hunted recently.
Two experienced hunters returned in 2017, each with a
.308 calibre rifle and thermal imaging equipment. Eighteen
goats were removed during the first four days and no more
were seen in the subsequent six days of intensive search
effort (Table 3). It is unknown whether goats ate rodent
bait, but its application to remove rats had no apparent
impact on the population, and the presence of goats did not
impact the success of the rat eradication.
Non-target species mitigation
The proposed application of rodent bait posed a
potential risk to non-target native species such as the
Polynesian ground dove and the Tuamotu sandpiper.
Tuamotu sandpiper were considered at high risk based
on observations made on Tahanea Atoll during a rat
eradication in 2011 (Pott, et al., 2014). Concerns were also
held for Polynesian ground dove although other projects
had targeted rats in the presence of conspecifics without
apparent losses (Griffiths, 2014). Both species were
recorded in very low numbers at just one of the project
sites (Vahanga) and, because of their conservation status,
mitigation was undertaken.
Prior to bait application, efforts were made to catch
all Polynesian ground dove on Vahanga and translocate
them to Tenararo. Of the five to six birds observed, two
were captured and transferred. The others evaded capture
and were monitored over the course of the project’s
implementation, along with two individuals sighted on
Tenania. Transferred birds had two of the outermost
primaries of each wing removed to lessen the chances of
them flying back to Vahanga.
Efforts were also made to capture and transfer Tuamotu
sandpiper. More birds were found on Vahanga than had
been anticipated and five of the six birds present were
caught. One escaped, but four were translocated to Tenararo
with outermost primaries plucked on both wings (1–3 per
wing, depending on bird condition) to prevent their return
to Vahanga.
Table 3 Monitoring completed to confirm eradication success at the six project sites.
Island
Invasive species
Corrected
trap nights
Vahanga
Tenania (Tenarunga)
Temoe
Kamaka
Makaroa
Manui
Pacific rat
Pacific rat, ship rat, cat
Pacific rat
Pacific rat
Pacific rat, goat
Rabbit
345
213
455
612
210
Monitoring effort
Sign
Spotlighting searches
(hrs)
(hrs)
16
17
25
8
20
112
112
128
188
232
Outcome
Trail
cameras
(hrs)
0
420
0
0
440
1,230
Successful
Successful
Successful
Failed
Successful
Successful
615
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
These interventions were partially effective for both
species with 90% of captured birds resighted in 2017
(R. Pierce pers. comm.). Ground dove that remained on
Vahanga and Tenania were resighted throughout the period
of implementation, suggesting any risks to this species
were low. Sightings of an uncaptured Tuamotu sandpiper
displaying symptoms of poisoning were made on Vahanga.
This individual was not seen again highlighting the
vulnerability of this species.
ERADICATION SUCCESS
Trapping, spotlight searches and searches for sign of
invasive vertebrate presence, conducted in April and May
2017 nearly two years after the project was implemented,
confirmed the project was successful at removing invasive
vertebrates at five of the six sites. No rats were found on
Vahanga, Tenania, Temoe or Makaroa. No cats were found
on Tenania or rabbits on Manui and goats were finally
removed from Makaroa. The monitoring effort expended
for each site to confirm eradication success is provided
in Table 3. Despite Kamaka being inhabited, rats were
not detected until monitoring was instigated nearly 12
months after the project was implemented. Rats are now
widespread on the island. Analysis of DNA confirmed that
some rats survived the operation.
DISCUSSION
Implementing the project described in this paper was
challenging due to the remote nature of the islands, the
number of sites, the range of invasive species targeted, and
the lack of infrastructure and resources available within
French Polynesia. Overcoming these challenges required
an extensive and thorough planning effort. An added
benefit of the time taken for project planning was the clear
identification of roles and responsibilities for each project
partner. Each of the project partners provided capabilities
that could not readily have been supplied by the other
partners.
OUTCOMES
An operational strategy, informed by a contemporary
review of rat eradications on tropical islands (Keitt, et
al., 2015) contributed to project success although, as
noted, rats survived on Kamaka despite the application
of best practice guidelines. Reasons why rats survived on
Kamaka are unknown but an investigation to determine
causal factors is currently underway. The project also
benefited from generally favourable weather through the
implementation phase. In hindsight, sufficient time, effort
and resources were put in place to ensure successful cat and
rabbit removal from Tenania and Manui. However, more
time spent on each of these islands would have increased
the level of confidence held by departing teams that
surviving individuals had been removed. Local reports that
goats were no longer present on Makaroa proved incorrect
and eradication of this species had to be postponed.
In removing invasive species from five islands, the
project increased the total number of islands free of invasive
vertebrates within French Polynesia from four to nine and
created an additional 1,426 ha of secure habitat, effectively
tripling the area available for Polynesian ground dove and
Tuamotu sandpiper recovery. Early signs of recovery were
observed in 2017 with more individuals of Polynesian
ground dove seen on both Vahanga and Tenania in 2017
and Tuamotu sandpiper recorded on Tenania for the first
time. Recovery of native vegetation was observed on both
Manui and Makaroa. Longer term monitoring is required
to confirm trends.
The cost efficiencies gained in this project through
removing invasive species from multiple islands are
evident. Completing each of the islands as a standalone
project would have increased the total cost of removing
invasive species from the six sites by a factor of three.
Resources for conservation are scarce and similar
approaches will need to be considered for many projects
to make them economically viable. The proposed removal
of rats and cats from five uninhabited islands in the
Marquesas archipelago is one such example. The high
costs of shipping and helicopters would rule out doing any
one of the islands as a standalone project.
With the removal of rats, the risk of rodent-borne
leptospirosis has been eliminated from Tenania and the
quality and quantity of copra produced appears to have
increased, although the increase in income generated for
the local community has yet to be quantified. Local skills
to undertake future eradication projects were developed
and support from policy-makers, funders and the public for
future rodent eradications on other atolls/islands generated.
Interventions to mitigate the impacts of the operation
to non-target species were largely effective (Pierce, et
al., 2015) and the level of mortality sustained will be
outweighed by the anticipated benefits to populations
following the removal of rats from Vahanga. Although
it is too early to measure the full impact of this
conservation intervention, Polynesian ground dove and
Tuamotu sandpiper should increase in abundance on both
Vahanga and Tenania, eventually forming self-sustaining
populations. The number of populations of Polynesian
ground dove will increase from three to five and for
PROJECT COST
The operational cost of the project was estimated based
on expenditure records kept by project partners. The total
cost of the project, from when concerted planning began in
2014 to completion of the operation in 2015, was €1.4M
with the largest costs being the helicopter, shipping, rodent
bait and personnel. The cost efficiency of the project gained
by targeted all six islands was assessed by comparing the
total cost of the project with estimates completed separately
for eradicating invasive mammals independently at each
site (Table 4). Costs such as helicopter, shipping and
staff travel would all have added significantly to cost if
each island had been completed as a standalone project.
Postponement of goat eradication on Makaroa increased
costs for this component of the project but only by a
relatively small margin as the cost of goat eradication was
small (<€20,000).
Table 4 Projected standalone project costs and
the actual costs incurred for removing invasive
vertebrates from the six project sites.
Island
Vahanga
Tenania
Temoe
Kamaka
Makaroa
Manui
Total
Projected cost
€1.1M
€1.1M
€1.1M
€0.4M
€0.5M
€0.4M
€4.6M
Actual cost1
€0.3M
€0.3M
€0.3M
€0.15M
€0.2M
€0.15M
€1.4M
Costs such as flying the helicopter from Tahiti were divided
equally between project sites.
1
616
Griffiths, et al.: Multi invasive species eradications, French Polynesia
Tuamotu sandpiper from six to eight. Polynesian storm
petrel (Nesofregetta fuliginosa) along with other sea birds
are expected to recolonise Makaroa increasing the number
of breeding sites for this species from at least six to seven.
Translocations of these species and others are also now
possible.
Completion of the project provided greater security
from extinction for a number of plant and animal species
but most importantly for bird species listed as critically
endangered or endangered by the IUCN (IUCN, 2010),
Polynesian ground dove, white-throated storm petrel and
Tuamotu sandpiper. The project also delivered socioeconomic benefits to local communities through increased
production from a coconut plantation on Tenania and
greater resilience for harvested seabird populations on
Temoe. In doing so, the project provides a precedent for
further action within French Polynesia to protect endemic
biodiversity and livelihoods.
ACKNOWLEDGEMENTS
This project received financial support from the
European Union ENTRP, the David and Lucile Packard
Foundation, the British Birdwatching Fair, the Mohamed
bin Zayed Species Conservation Fund, and National
Geographic Society. Sponsorships were received from Bell
Laboratories and the T-Gear Trust of Canada. The project
also received in-kind support from the New Zealand
Department of Conservation, Pacific Invasives Initiative
and the Government of French Polynesia. We are indebted
to these funders and organisations for making this project
possible and to the many individuals who worked on or
supported the project.
REFERENCES
Andréfouët, S., Chauvin, C., Spraggins, S., Torrespuzilla, D. and
Kranenbourg, C. (2005). Atlas des Recifs Coralliens de Polynesie
Francaise. Nouvelle-Caledonie: Centre IRD de Noumea.
Atherton, J. (2007). Polynesia-Micronesia Biodiversity Hotspot. Critical
Ecosystem Profile. Arlington, Virginia, USA: Ecosystem Partnership
Fund, Conservation International–Melanesia Center for Biodiversity
Conservation.
Blanvillain, C., Florent, C. and Thenot, V. (2002). ‘Land birds of Tuamotu
Archipelago, Polynesia: Relative abundance and changes during the
20th century with particular reference to the critically endangered
Polynesian ground-dove (Gallicolumba erythroptera)’. Biological
Conservation 103(2): 139–149.
Broome, K., Boudjelas, S. and Gouni, A. (2011). Feasibility Study for
the Eradication of Pacific Rats from Vahanga Atoll, French Polynesia.
Unpublished Report, Tahiti, French Polynesia: Societe d’Ornithologie
de Polynesie, Manu.
Burt, A.J., Gane, J., Olivier, I., Calabrese, L., De Groene, A., Liebrick,
T., Marx, D. and Shah, N. (2016). ‘The history, status and trends of
the endangered Seychelles magpie-robin Copsychus sechellarum’. Bird
Conservation International 26(4): 505–523.
Cromarty, P.L., Broome, K.G., Cox, A., Empson, R.A., Hutchinson,
W.M. and McFadden, I. (2002). ‘Eradication planning for invasive
alien animal species on islands – the approach developed by the New
Zealand Department of Conservation’. In: C.R. Veitch and M.N. Clout
(eds.) Turning the tide: the eradication of invasive species. pp. 85–91.
Occasional Paper SSC no. 28. IUCN SSC Invasive Species Specialist
Group. IUCN, Gland, Switzerland and Cambridge, UK.
Derand, G.-D., Griffiths, R., Ghestemme, T. and Cranwell, S. (2015).
Operational Plan to Eradicate Rats, Cats, Rabbits, Goats, and Control
Lantana on Three Islands and Three Atolls In French Polynesia.
Unpublished Report. Suva, Fiji: BirdLife International, Pacific
Secretariat.
Donlan, C.J. and Wilcox, C. (2008). ‘Integrating invasive mammal
eradications and biodiversity offsets for fisheries bycatch: Conservation
opportunities and challenges for seabirds and sea turtles’. Biological
Invasions 10(7): 1053–1060.
Faulquier, L. (2008). Feasibility Study for the Eradication of Rats from
Islands of the Gambier Archipelago (French Polynesia). Suva, Fiji:
BirdLife International, Pacific Secretariat.
Genovesi, P. (2011). ‘Are we turning the tide? Eradications in times of
crisis: how the global community is responding to biological invasions’.
In: C.R. Veitch, M.N. Clout and D.R. Towns (eds.) Island invasives:
eradication and management, pp. 5–8. Occassional Paper SSC no. 42.
Gland, Switzerland: IUCN and Auckland, New Zealand: CBB.
Gillespie, R.G., Claridge, E.M. and Goodacre, S.L. (2008). ‘Biogeography
of the fauna of French Polynesia: Diversification within and between
a series of hot spot archipelagos’. Philosophical Transactions of the
Royal Society B: Biological Sciences 363(1508): 3335–3346.
Griffiths, R. (2014). Feasibility and Risk Assessment for the Removal of
Invasive Species from Vahanga, Tenarunga, Temoe, Kamaka, Manui
and Makaroa. Unpublished Report. Santa Cruz, CA, USA: Island
Conservation.
Griffiths, R., Miller, A. and Climo, G. (2011). ‘Addressing the impact
of land crabs on rodent eradications on islands’. Pacific Conservation
Biology 17(4): 347–353.
Griffiths, R., Buchanan, F., Broome, K., Neilsen, J., Brown, D. and
Weakley, M. (2014). ‘Successful eradication of invasive vertebrates on
Rangitoto and Motutapu Islands, New Zealand’. Biological Invasions
16(11): 1–15.
IUCN. (2010). ‘IUCN Red List of Threatened Species. Version 2010.1’
<www.iucnredlist.org>. Accessed 19 May 2010.
Keitt, B., Griffiths, R., Boudjelas, S., Broome, K., Cranwell, S., Millett,
J., Pitt, W. and Samaniego-Herrera, A. (2015). ‘Best practice guidelines
for rat eradication on tropical islands’. Biological Conservation 185:
17–26.
Meyer, J.-Y. and Butaud, J.-F. (2009). ‘The impacts of rats on the
endangered native flora of French Polynesia (Pacific Islands): Drivers of
plant extinction or coup de grâce species?’. Biological Invasions 11(7):
1569–1585.
Pierce, R.J., Boudjelas, S., Broome, K., Cox, A., Denny, C., Gouni, A.
and Raust, P. (2006). Ecological Restoration of Vahanga Atoll, Acteon
Group, Tuamotu Archipelago – Draft Operational Plan. Tahiti, French
Polynesia: Société d’Ornithologie de Polynesie.
Pierce, R., Blainvillain, C. and Burle, M.-H. (2015). Bird Research and
Monitoring Associated with SOP Manu Pest Eradications in Acteon–
Gambier Archipelagos, June 2015. Unpublished Report. Suva, Fiji:
BirdLife International, Pacific Secretariat.
Pott, M., Griffiths, R., Burle, M.-H. and Wegmann, A. (2014). ‘Protecting
the Tuamotu Sandpiper (Prosobonia cancellata) One Island at a Time’.
In: R.M. Timm and J.M. O’Brien (eds.) Proceedings of the 26th
Vertebrate Pest Conference, pp. 166–171. Davis, CA: University of
California.
Pott, M., Wegmann, A., Griffiths, R., Samaniego-Herrera, A., Cuthbert,
R., Brooke, M., Pitt, W., Berentsen, A., Holmes, N., Howald, G. and
Russell, J.C. (2015). ‘Stacking the deck: Assessing bait availability
before rodent eradications to inform bait application rates’. Biological
Conservation 185: 27–35.
Raust, P. and Sanford, G. (2007). Les Zones Importantes pour la
Conservation des Oiseaux en Polynésie Française. Papeete, Tahiti,
French Polynesia: Société d’Ornithologie de Polynésie.
Rice, S.E. and Clark, R.W. (2016). ‘Factors affecting intra-island genetic
connectivity and diversity of an abundant and widespread island
endemic, the San Clemente Island night lizard’. bioRχiv: 060038.
Russell, J.C., Cole, N.C., Zuël, N. and Rocamora, G. (2016). ‘Introduced
mammals on western Indian Ocean islands’. Global Ecology and
Conservation 6: 132–144.
Russell, J.C. and Holmes, N.D. (2015). ‘Tropical island conservation: Rat
eradication for species recovery’. Biological Conservation 185: 1–7.
Simberloff, D. (2002). ‘Today Tiritiri Matangi, tomorrow the world. are
we aiming too low in invasives control?’. In: C.R. Veitch and M.N.
Clout (eds.) Turning the tide: the eradication of invasive species, pp.
4–12. Occasional Paper SSC no. 28. IUCN SSC Invasive Species
Specialist Group. IUCN, Gland, Switzerland and Cambridge, UK.
Towns, D., Miller, K., Nelson, N. and Chapple, D. (2016). ‘Can
translocations to islands reduce extinction risk for reptiles? Case studies
from New Zealand’. Biological Conservation 204: 120–127.
Zarzoso-Lacoste, D. (2013). ‘Vers une Meilleure Compréhension des
Interactions Trophiques Directes et Indirectes entre Prédateurs Invasifs
et Espèces Natives au Sein des Écosystèmes Insulaires’. PhD thesis.
Marseille, France: Aix-Marseille University.
617
T. Harvey-Samuel, K.J. Campbell, M. Edgington and L. Alphey
Harvey-Samuel, T.; K.J. Campbell, M. Edgington and L. Alphey. Trialling gene drives to control invasive species: what, where and how?
Trialling gene drives to control invasive species: what, where and how?
T. Harvey-Samuel1, K.J. Campbell2, M. Edgington1 and L. Alphey1
1
The Pirbright Institute, Ash Road, Woking, GU24 0NF, UK. <tim.harvey-samuel@pirbright.ac.uk>. 2Island
Conservation, 2100 Delaware Ave, Santa Cruz, CA, 95060, USA.
Abstract The control of invasive species would be enhanced through the addition of novel, more effective and sustainable
pest management methods. One control option yet to be trialled in the field is to deploy transgene-based ‘Gene Drives’:
technologies which force the inheritance of a genetic construct through the gene pool of a wild population, suppressing
it or replacing it with a less harmful form. There is considerable interest in applying gene drives to currently intractable
invasives across a broad taxonomic range. However, not all species will make efficient or safe targets for these technologies.
Additionally, the safety and efficacy of these systems will vary according to where they are deployed, the specific
molecular design chosen, and how these factors interact with the ecology of the target pest. Given the transformative but
also controversial nature of gene drives, it is imperative that their first field trials are able to successfully demonstrate
that they can be used safely and efficiently. Here, we discuss how to maximise the probability of this outcome through
considering three important questions: What types of invasive species should we use to trial gene drives? Where should
we be trialling them? and How should these trials be conducted? In particular, we focus on the ecological, genetic and
geographic features of small, isolated islands which make them ideal locations for these initial trials. A case study of an
island invasive that is deemed highly appropriate for gene drive intervention, and for which gene drive development is
currently underway (Mus musculus), is used to further explore these concepts.
Keywords: biodiversity conservation, CRISPR, Culex quinquefasciatus, gene drive, island invasive, Mus musculus,
population eradication, restoration
INTRODUCTION
Molecular advancements have made feasible a new
range of Genetic Pest Management (GPM) strategies –
the transgene-based gene drives (Sinkins & Gould, 2006).
These technologies aim to introduce DNA sequences
(the gene drive transgene) into the genome of a wild pest
population through the release of genetically engineered
individuals which go on to mate with conspecifics in the
field. Once introduced, the inheritance of the gene drive is
forced – driven – through the target population gene pool
along with its control phenotype. This driving effect can
be achieved, for example, by biasing inheritance of the
transgene above normal mendelian levels, or through placing
an evolutionary advantage on inheritance of the transgene
at the population level. Proposed control phenotypes aim
either to reduce/eradicate a pest population – “population
suppression” strategies – or to leave a population intact
but modify it so that it is less harmful (e.g. by spreading a
transgene which makes a mosquito population less able to
transmit a particular disease) – “population replacement”
strategies (Alphey, 2014). Within population suppression,
current proposals aim to spread either a sex ratio bias
(usually in favour of males) or a genetic load, e.g. female
sterility (Deredec, et al., 2008).
Theoretically, gene drives could be engineered that are
capable of spreading to every member of an interbreeding
population from one or several relatively small initial
releases (Deredec, et al., 2008). This autonomous
nature is appealing for invasive species control, where
programmes often extend into remote/inaccessible areas
and less than total eradication may be viewed as failure.
Indeed, there is increasing interest in applying gene drives
to currently intractable invasive species that threaten
biodiversity (Alphey, 2002; Gould, 2008; Esvelt, et al.,
2014; Simberloff, 2014; Thresher, et al., 2014; Campbell,
et al., 2015; NASEM, 2016; Harvey-Samuel, et al., 2017;
Piaggio, et al., 2017). However, two primary concerns
arise from their proposed use. Firstly, that a gene drive
transgene could unintentionally spread beyond a target
geographic area (e.g. from an invasive population into the
native range of the invader) or into a non-target species
through hybridisation/horizontal-gene transfer – here
collectively termed ‘transgene escape’. Secondly, that
their persistence, once released, could cause unintended
ecological effects that are difficult to reverse (Sutherland,
et al., 2014; Webber, et al., 2015; NASEM, 2016).
Previous field testing of gene drives is limited
to non-transgenic population replacement utilising
artificial infections of Aedes aegypti mosquitoes with
the intracellular bacterium Wolbachia (Hoffmann, et al.,
2011; Schmidt, et al., 2017). Wolbachia technologies are
considered non-transgenic as they do not, deliberately,
involve the introduction of DNA sequences into the target
pest genome. Proposed application of transgene-based
gene drives to invasive species differs from Wolbachia
in that the systems available are, potentially, significantly
more powerful and flexible and their taxonomic scope
is broader, encompassing groups as divergent as plants,
mammals, fish and molluscs, in addition to insects
(Gould, 2008; Hodgins, et al., 2009; Thresher, et al., 2014;
Campbell, et al., 2015; Sytsma, et al., 2015; Webber, et
al., 2015). The first open-field trials of transgene-based
gene drive technologies will thus represent a precedentsetting milestone. As recommended by the USA National
Academy of Sciences (NASEM, 2016) , these trials will
seek to examine whether the efficacy (e.g. its ability to
invade a target population and induce a desired control
phenotype therein) and safety (e.g. our ability to constrain
its spread to the target population using molecular or
experimental designs) of a gene drive system conform
with theoretical expectations, themselves informed by
preliminary laboratory experiments and mathematical
modelling (Benedict, et al., 2008; Brown, et al., 2014).
As such, open-field trials can be considered extensions of
initial highly biocontained laboratory experiments where
artificial biocontainment (Akbari, et al., 2015) is ‘relaxed’
because aspects of efficacy and safety have previously been
demonstrated. Both these aspects – efficacy and safety –
are important in order to convince a potentially sceptical
public that they may have confidence in the wider use of
these technologies.
Here we summarise the primary considerations
involved in conducting the precedent-setting open-field
trials of transgene-based gene drives (henceforth ‘gene
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
618
up to meet the challenge, pp. 618–627. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Harvey-Samuel, et al.: Trialling gene drives
drives’) in invasive species through posing three questions
(1) What types of invasives are appropriate targets for
these trials? (2) Where should the trials of these systems
be located? (3) How should these trials be conducted?
These questions are considered with the aim of exploring
how these technologies could be trialled against invasive
species as efficaciously as possible, whilst minimising the
risk of transgene escape. In order to increase the value of
this discussion these points are addressed in a general,
rather than taxon-specific manner. Additionally, we
explore their implications for a specific invader currently
being targeted for control using gene drives – the house
mouse, Mus musculus (See case study: GBIRd and Table
1). We bring this forward with the purpose of encouraging
dialogue and improving criteria for such trials.
WHAT CHARACTERISTICS ARE IMPORTANT
WHEN CHOOSING A TARGET ORGANISM?
General characteristics of a gene drive target
Minimum requirements for gene drive development are
that the target pest is sexually reproductive, is amenable to
laboratory rearing/germ-line transgenesis and is genetically
well characterised.
As barriers to gene-flow within a population will
decrease the efficiency of a gene drive’s spread (see: The
importance of dispersal), target species should preferably be
obligately sexually reproductive (Alphey, et al., 2010) and
incapable of self-fertilisation, which may simultaneously
reduce the potential for gene drive resistance evolution
(Bull, 2016). As such, it is unlikely these systems will
be broadly applicable to invasive plants, of which many
propagate vegetatively or through self-fertilisation (Kolar
& Lodge, 2001; Rambuda & Johnson, 2004). Regarding
transgenesis, the ease with which the germ-line cells can
be manipulated will influence the speed that new transgene
designs can be tested. Insect transgenesis has predominantly
been through microinjection of pre-blastoderm embryos
which requires that the fertilised egg is accessible.
Transformation of species which are viviparous (e.g.
the tsetse fly) or whose embryos are laid in inaccessible
protective structures (e.g. pods or cases) may prove more
challenging (Bourtzis, et al., 2016). Finally, as gene drives
require the expression of various genetic components in
highly temporal or spatially explicit patterns, often to target
precise genomic loci, a good knowledge of the genetics of a
target, e.g. a high-quality genome/transcriptome sequence
and an understanding of the molecular-genetic basis of sex
determination, is imperative.
Desirable characteristics are not absolutely necessary
for gene drive development but, in practice, species whose
biology diverged significantly from these characteristics
would be deemed as inappropriate targets for these
technologies.
Chief amongst desirable characteristics is a short
generation time. This will minimise the time taken for
strain development, and for these vertically transmitted
systems to spread through and control a target population.
Similarly, species with complex mating systems (e.g. the
synchronised and ephemeral mating events of termites or
ants) or where subsets of the population can remain dormant
and inaccessible (e.g. long-term seed banks) effectively
Table 1 Idealised ecological selection criteria proposed as an initial filter for potential trial islands for potential gene drive
constructed mice trials within Australia, New Zealand, USA. Additional steps will be required prior to any potential field
trial, including engagement with stakeholders (e.g. land managers, local communities) and regulators to determine final
approval.
Criteria
1 Island is biosecure
Desktop assessment indicates:
a. Closed to public or infrequent/controlled visitation
b. Remote enough (>1 km from other land masses) to
avoid unassisted immigration or emigration
2. No significant challenges exist to treatment using
traditional methods to eradicate mice, e.g.:
a. Uninhabited (besides research station or similar)
b. No livestock
c. No native rodents
d. No non-target species of concern
e. Regulatory environment allows the use of
brodifacoum bait products and no rodenticide
resistance alleles present
f. Island size <300 ha
g. Single land manager
3. M. musculus are the only rodent present or could be
introduced.
4. Reasonably economical and feasible to visit the island
year-round.
Rationale
● Mice typically invade remote islands through human
mediated transport, not through swimming (Russell
& Clout, 2005).
● M. musculus are known to have swum up to 500 m
between land masses (Harris, et al., 2012).
● Closed population required for proof-of-concept
● After desktop assessment. If the island passes
other filters and is tentatively selected, conduct a
biosecurity risk assessment. Island biosecurity plans
for individual islands or island groups should be
developed and implemented if island is selected
(Fritts, 2007; Russell, et al., 2008; AAS, 2017)
● Provides a means to terminate experiments (i.e. exit
strategy) using traditional methods without known
complicating factors.
● Mouse behaviour is known to change significantly in
the presence of rats (Harper & Cabrera, 2010).
● There may be man-made or other islands that are
suitable that don’t currently have M. musculus
present.
● Some islands are cost prohibitive to visit.
● Seasonal conditions may impact safe access to the
island.
619
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
extend the generation time and may limit transgene
introgression into or through a wild population (Alphey,
et al., 2010). Furthermore, it is critical that there is a good
knowledge of the ecology (e.g. mating systems, population
dynamics and community interactions) of the target and in
the case of vectors, the ecology and epidemiology of the
pathogen and disease. The importance of this knowledge
when developing a GPM strategy – from choosing the most
appropriate/effective system, to predicting the impact of a
strategy on a target population and community – cannot
be overstated (Yakob, et al., 2008; Bax & Thresher, 2009;
Yakob & Bonsall, 2009; Bonsall, et al., 2010; Thresher,
et al., 2013; Piaggio, et al., 2017). Finally, it is desirable
that the target is the dominant and ideally, sole, cause of
an impact. As these strategies are vertically transmitted,
they are extremely species-specific, making scenarios
where there are multiple contributors to an impact (e.g.
the spread of avian pox in Hawaii, where there are both
mechanical and vector-based disease transmission routes)
less appropriate.
The importance of dispersal
Gene-flow between populations
In limiting transgene escape into non-target areas, two
important and interacting considerations are the level of
gene-flow between a target and non-target population and
the invasion threshold of the gene drive deployed (Figs 1
and 2) (Marshall & Hay, 2012). The invasion threshold is
the theoretical frequency a gene drive transgene must be
present at in a population before it will begin to spread.
Highly invasive gene drives spread from very low invasion
thresholds (e.g. the introduction of a few individuals into
a target population) while less invasive systems may
require significant levels of introduction before they begin
to spread (high invasion threshold). Transgene escape
may be considered an issue if gene-flow occurs at a
frequency which makes it probable that a gene drive will
exceed its invasion threshold in a non-target population
within the time-frame of a trial (Akbari, et al., 2013). An
‘acceptable’ level of gene-flow between target and nontarget populations will therefore be significantly higher for
less invasive gene drives (Fig. 2). As the choice of gene
drive may be constrained by the desired outcome (less
invasive systems are generally more suited to replacement
rather than suppression), it may not always be possible to
choose less invasive designs to prevent transgene escape
in species which are capable of long-distance gene-flow. A
more flexible option is to trial gene drives in species which
show limited ability to disperse and where human-mediated
dispersal pathways can be managed. As previously noted
(NASEM, 2016), important considerations here are the
distance, frequency and life-stage of dispersal. Generally,
species which disperse as juveniles/adults will show lower
rates of gene-flow between populations than those which
disperse as fertilised embryos (seeds or spores) or gametes
(e.g. wind-borne pollen) (NASEM, 2016). Furthermore,
dispersal via gametes may be more likely to result in
interspecific hybrids, potentially increasing the risk of
transgene escape into non-target species (NASEM, 2016).
Consideration of these dispersal issues may make terrestrial
animals more attractive targets than plants or marine
species. As social interactions can strongly influence adult/
juvenile dispersal events, it is important to consider how the
predicted outcome of a particular gene drive may interact
with these species-specific behavioural cues. For example,
mate-limitation or increased inbreeding at low population
densities or highly skewed sex ratios (both expected
outcomes of proposed suppression gene drive designs)
could in some species/scenarios result in increased levels
of dispersal (Clobert, et al., 2012; Matthysen, 2012) and
potentially also transgene escape.
Prior knowledge of the dispersal behaviour of an
invasive population is therefore a prerequisite to safely
deploying a gene drive. Fortunately, for many important
invaders details of their dispersal mechanisms, invasion
rates and levels of gene-flow within their invaded range
already exist – in addition to other useful details such as
the observed variance in their population size. Potential
target species and populations could be short-listed based
on the existence of this historical information, which could
then be used to inform models predicting the potential for
transgene escape during the expected time-frame of a trial.
Gene-flow within a population
Fig. 1 Gene drives may be classified by their level of
invasiveness, which is defined as the frequency they
must reach in a target population before they begin to
spread (the invasion threshold).Relatively non-invasive
gene drives such as underdominance-based systems
(Reeves, et al., 2014) (solid lines) require a high minimum
allele frequency (dashed line) to be exceeded before
they will begin to spread (50% of the population in this
simulation). This differs from highly invasive (also known
as “global”) gene drives such as homing-based systems
(Deredec, et al., 2008; Unckless, et al., 2015) (dotted)
that will theoretically spread throughout a population
even from a very low initial allele frequency, at least in
the absence of resistant alleles.
620
Reaction-diffusion models have shown that dispersal
rates will affect the speed that a gene drive travels through
a target population (Beaghton, et al., 2016). Under more
realistic scenarios, barriers to gene-flow within a population
may have a more qualitative effect on whether a gene drive
will spread or persist (North, et al., 2013). This concern
could be reduced by avoiding targets whose populations
show strong local spatial structuring, e.g. those which
engage in high levels of sib-sib mating (Hamilton, 1967).
However, even less extreme levels of spatial structuring
resulting from limited life-time dispersal can significantly
affect the ability of a gene drive to spread through and
collapse a target population (Huang, et al., 2011; Eckhoff,
et al., 2017). In particular, species whose population
dynamics are significantly affected by seasonality may
provide more fragmented landscapes for a gene drive
to attempt to traverse. Models comparing gene drive
dynamics in spatially explicit and homogenous mosquito
populations suggest that increased structuring of a target
population decreases the parameter space under which the
target population is successfully eradicated (Eckhoff, et al.,
2017). In these models, sub-populations became explicit
annually in response to lowered population densities during
the dry season. If sub-populations became explicit prior to
arrival of the spreading transgene, these areas could act
as a source for wild-type reinvasion into areas where the
Harvey-Samuel, et al.: Trialling gene drives
drive had eradicated the pest the previous season. Although
limited within-population dispersal can be overcome
through increasing the ‘patchiness’ or number of transgenic
releases (Huang, et al., 2011; Eckhoff, et al., 2017), this
tactic partially negates the primary advantage of employing
gene drives. In choosing a target it is thus critical to have
evaluated whether, given their population spatial-structure
and the gene drive chosen, the release effort required to
efficiently eradicate or replace a population is low enough
to justify intervention with this technology.
Relatedness to important pests
Development and trialling of gene drives against
invasives will proceed most efficiently if target species
impact multiple values (e.g. human or animal health,
agriculture, conservation). If these ‘dual-target’ species can
be identified then the financial burden of developing gene
drive strategies could be shared amongst different funding
agencies, efficient designs/components shared between
different researchers and the benefits of, and motivation
for gene drive deployment shared amongst varied
stakeholders. If a target invasive did not impact multiple
values, gene drive development would still benefit if they
were closely related to species in which GPM technology
had previously been investigated, due to the transferability
of many underlying molecular designs and components
(Harvey-Samuel, et al., 2017). Examples of dual-target
species are the mosquito (Culex quinquefasciatus) – a
vector for multiple human diseases (Eldridge, 2005) and
an invasive vector of avian malaria in Hawaii (LaPointe, et
al., 2012) – and rodents including the house mouse (Mus
musculus) and rats (e.g. Rattus exulans, R. norvegicus and
R. rattus) which collectively are serious economic pests
of agriculture (Aplin, et al., 2003; Pimentel, et al., 2005),
impact infrastructure, are hosts for human, domestic animal
and wildlife disease (Banks & Hughes, 2012), and amongst
the most damaging invasives of island ecosystems (Angel,
et al., 2009; Harper & Bunbury, 2015). Encouragingly,
germ-line transgenesis and genome sequences already exist
for C. quinquefasciatus (Allen, et al., 2001; Arensburger,
et al., 2010), M. musculus (Waterston, et al., 2002; Ivics,
et al., 2014) and R. norvegicus (Gibbs, et al., 2004; Ivics,
et al., 2014). Moreover, all these species are invasive in
isolated, uninhabited areas where there are no closely
related species: desirable characteristics for a gene drive
trial location (see next section).
WHERE SHOULD TRIALS BE CONDUCTED?
In order to maximise containment and efficacy, small,
isolated islands are ideal locations for the first trials of gene
drives (WHO/TDR, 2014).
Advantages of island locations to trial safety
Limiting intraspecific transgene escape
Gene-flow from an invasive population to conspecifics
in its native range will decrease with increasing interpopulation distance, the ecological inhospitality of the
intervening area and the size of the invasive ‘source’
population. Locating trials on small, isolated islands
can therefore act as an ecological containment strategy
(WHO/TDR, 2014; NASEM, 2016), reducing the risk of
intraspecific transgene escape. The effectiveness of this
containment will depend on the proximity of a trial island
to the native range of an invader, its natural and human-
Fig. 2 The invasiveness of gene drive systems affects their containment ability once deployed in the field. Relatively noninvasive systems require large initial introductions before the gene drive will begin to spread and therefore migration
alone is unlikely to exceed their invasion threshold. Highly invasive gene drives, on the other hand, can spread from only
a few initial colonists and are predicted to spread through all linked populations. This is illustrated above using a three
deme population genetics mathematical models. In each case we assume that a target (bottom) island and a nearby
neighbour (middle) exchange 2% of their respective populations by migration in each generation while the nearby
neighbour and a more remote (top) island exchange just 1%. It is assumed that no direct migration occurs between
the target and remote islands due to the distance between them. Resulting transgene frequencies for each island at
various times after the transgenic release are represented diagrammatically by a series of 25 mice (each representing
a transgene frequency of 4%). White and shaded mice respectively represent wild-type and underdominance/homing
drive transgenic allele frequencies, rounded to the nearest 4% (i.e. to the nearest whole mouse). Panel (a) shows results
for a frequency dependent, single locus haploinsufficient underdominance-based system (Reeves, et al., 2014). This is
a relatively non-invasive system with a high invasion threshold of 50% (See Fig. 1, solid lines). Here it is assumed that
wild-type and transgene homozygotes suffer no fitness cost while 50% of heterozygous offspring are non-viable. For
an initial transgene frequency of 55% it can be seen that the system spreads throughout the target population but does
not reach significant levels in the neighbouring populations. Panel (b) shows results from a homing-based gene drive
(Deredec, et al., 2008; Unckless, et al., 2015) which imparts no fitness cost on individuals and converts heterozygotes
to homozygotes with 100% efficiency, introduced with an initial transgene frequency of 0.1%. The population genetics
of this gene drive are shown in Fig. 1, (dotted line). Even this low initial frequency allows this highly invasive gene drive
to spread throughout the target, and in time, the neighbouring populations also.
621
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
mediated dispersal ability and the invasiveness of the gene
drive being trialled. A set of case-studies illustrating the
interplay between these factors is the open-field releases of
artificial Wolbachia infections aimed at local replacement
of A. aegypti mosquito populations in Australia. After
deliberate establishment in relatively isolated trial A.
aegypti populations (Hoffmann, et al., 2011), it was
found that long-distance dispersal was taking Wolbachia
infected mosquitos into areas beyond the trial site (up to
1.86 km away) but that migration rates were insufficient
over this distance to overcome the relatively high invasion
threshold of the Wolbachia system (>30%) which
remained largely contained to the trial site (Hoffmann, et
al., 2014). Conversely, in subsequent releases where the
trial site formed part of a larger, continuous A. aegypti
population, Wolbachia was capable of spreading, albeit
slowly, to high frequency beyond release sites and into the
wild target population (Schmidt, et al., 2017). Gene drives
with lower invasion thresholds than Wolbachia will require
significantly greater isolation and/or molecular safeguard
designs to limit transgenes to target populations/areas
(discussed in the How section). This concept is illustrated
for transgene-based gene drives in Figs 1 and 2.
In the context of island trial locations, the potential for a
gene drive to cover large geographic distances, potentially
back to mainland populations, through ‘island-hopping’
should not be overlooked (Bellemain & Ricklefs, 2008).
For suppression drive designs, this island-hopping would
require the existence of viable populations extending back
to a native range and for the drive to escape each invaded
‘stepping-stone’ population before that population was
itself eliminated by the drive. However, for replacement
drives these aspects would not be a pre-requisite.
Limiting interspecific transgene escape
Transgene escape between species could take place
either through horizontal gene transfer (HGT – acquisition
of genetic material from an organism other than a direct
ancestor) or introgression following hybridisation. Signals
of HGT in metazoans can be seen by sequence comparisons
between species (e.g. Crisp, et al., 2015). However, even
the most frequent of these HGT events are rare, seen in
nature on timescales of millions of years (e.g. Ortiz, et al.,
2015). Therefore, as discussed generally for mosquitoes
(Besansky, 2015) and specifically for homing-drives (Burt,
2003), HGT of a gene drive is held to be unlikely to occur
at a frequency which will make it a realistic concern.
Regular gene-flow between native and invasive
species through introgressive hybridisation, however,
is well documented (Mooney & Cleland, 2001). Here,
island locations provide both benefits and disadvantages
in terms of limiting transgene escape. A benefit is that,
given a frequency of fertile hybridisation events, stochastic
elimination of an escaped transgene prior to its spread in a
non-target species is more likely in small, island populations
than at continental scales. However, hybridisation between
closely related invasive and native species may be higher
in insular compared to continental communities (Rhymer
& Simberloff, 1996), potentially allowing transgenes to
introgress into native populations at increased rates on
islands. The potential genetic homogeneity of an island
invasive population and simplicity of island communities
(reducing the number of hybridising congeners) may prove
advantageous in designing sequence-specific molecular
safeguards to limit this risk.
622
Advantages of island locations to trial efficacy
Geographic isolation
Trials of gene drives will seek to achieve a series of predefined scientific endpoints (Brown, et al., 2014; NASEM,
2016). These will include evidence that the transgene is
able to spread efficiently in the wild population, as well
as endpoints specific to individual designs (e.g. reduced
population density or reduced number of fully-competent
vectors for suppression and replacement strategies,
respectively). As immigration of wild-type individuals
into a target population effectively dilutes the frequency
of the transgene, unanticipated immigration will cause
drive rates to be estimated inaccurately; this has been a
frequently-observed problem in trials of sterile insects for
population suppression (Klassen & Curtis, 2005) and is
assumed to have prevented fixation of artificial Wolbachia
infected mosquitoes in open-field trials (Hoffmann, et al.,
2014). A sufficiently isolated island trial site will reduce
this concern through minimising wild-type immigration.
What constitutes ‘sufficient’ geographic isolation could
be considered in conjunction with estimating outward
gene-flow from a proposed trial island, acknowledging
that migration rates between populations may not be
symmetrical (Kawecki, 2004) and may only occur during
infrequent events (e.g. El Niño, hurricanes).
Small population size
For equivalent release numbers/resources, introductions
can be made at a higher population allele frequency on small
islands than at larger, continental scales. This is primarily
advantageous in testing gene drives with high invasion
thresholds. However, even for more invasive systems, test
releases would likely take place at frequencies well above
the estimated minimum to protect against stochastic loss of
the transgene in initial generations. Increased introduction
rates will also allow the transgene to reach fixation (or a
stable internal equilibrium) more rapidly (Deredec, et al.,
2008). Moreover, for population-suppression strategies,
smaller target populations may mean that densitydependent processes such as Allee effects (Tobin, et al.,
2011) and environmental stochasticity (Eckhoff, et al.,
2017) can be leveraged to more rapidly drive populations
to extinction.
Genetically distinct
Small, insular populations arising from recent single
invasion events are likely to be relatively genetically
homogenous (Dlugosch & Parker, 2008). Assuming that
heritable resistance to gene drives is possible (Bull, 2015),
but founder individuals did not carry resistance alleles,
this would provide target populations initially entirely
susceptible to a released drive. Furthermore, given a
constant mutation rate, such a gene drive resistance allele
is less likely to arise in smaller, isolated populations within
the time-frame of a trial. Conversely, however, if founder
individuals did display pre-existing resistance it may occur
at high frequencies. Target island populations should be
screened prior to a trial for the presence of pre-existing
resistance mutations; a relatively simple task for sequencespecific homing-drives, but potentially less straightforward
for other technologies.
Which islands?
Islands that are small and sufficiently isolated to provide
effective ecological containment could provide ideal
locations for trialling gene drives. However, there are a
Harvey-Samuel, et al.: Trialling gene drives
number of biological, geographic and social criteria which
will, in general, make a location more or less suitable for
trialling GPM strategies (Benedict, et al., 2008; Lavery, et
al., 2008; Brown, et al., 2014) and which can be extended
to identify particularly promising examples within this
group.
Biological criteria
If sufficiently isolated, invasive populations will be
allopatric from conspecifics in their native range but
sympatric with native congenic populations with which
they might hybridise. If it occurs at an appreciable
frequency, interspecific gene-flow may therefore be
considered the more likely of the two risks when trialling
gene drives in these locations. The most effective solution
would be to avoid locations where there are closely related
native species. For example, targeting invasive rodents
on off-shore islands in New Zealand (which has no native
terrestrial mammals) would pose low/no risk of transgene
escape into native species, whereas deployment of the
same technology in areas with diverse endemic rodent
fauna such as south-east Asian archipelagos (Amori, et
al., 2008) would likely require extensive pre-trial risk
assessment. A further point to consider is that hybridisation
events may be unidirectional with regards to sex (Rhymer
& Simberloff, 1996). Molecular designs such as Y-drive
which are transmitted exclusively through the paternal
line would not be introgressed into a native population
if hybrids formed via crosses between native males and
invasive females.
If sufficient safety measures are taken, gene drives are
expected to act in an extremely species-specific manner
and are thus highly suitable for deployment in ecologically
sensitive locations. However, a precautionary approach
would suggest that precedent-setting trials be conducted
in locations devoid of endangered/threatened flora or
fauna (Brown, et al., 2014). This is particularly relevant if
broader spectrum conventional control methods are used to
terminate the trial at a pre-defined endpoint (see Table 1).
Geographic criteria
Barriers to gene-flow will decrease the efficiency
of a released gene drive. Islands with relatively simple
geographies and a resulting homogenous invasive
population, for example low-lying oceanic islands,
will therefore be most amenable to initial trials of these
technologies. Where multiple islands occur in close
proximity, these areas could be used to test assumptions
on the spread of a drive technology within/between
populations depending on the dispersal of the target
(e.g. coral atolls/archipelagos for short/longer distance
dispersal, respectively).
Social criteria
Challenges associated with invasive species control in
inhabited areas are well-documented (Oppel, et al., 2011;
Glen, et al., 2013). The novel and controversial nature of
gene drives means that these challenges are likely to be
exacerbated during their first trials. Levels of regulatory/
engagement costs, risk assessment and societal objection
are all likely to be more favourable if initial trials take
place in uninhabited areas which are not of great cultural
value. At least as importantly, restricting traffic off an
island during a trial will substantially reduce the likelihood
of transgene escape via intraspecific gene-flow. Employing
modified biosecurity measures currently employed during
conventional eradication efforts (Russell, et al., 2008), this
would be far more feasible for uninhabited areas.
Previous experience in choosing sites for self-limiting
GPM mosquito trials suggest that two social criteria
critical for site identification are the existence of a credible
regulatory structure and an enthusiastic local participant
(e.g. academic researcher or wildlife management agency)
with expertise regarding the invasive being targeted
(Brown, et al., 2014). The regulatory framework in
operation is relevant at multiple stages during planning and
implementing a gene drive trial, from granting importation
permits for gene drive organisms (Brown, et al., 2014) to
determining appropriate risk assessment (NASEM, 2016)
and public engagement (Lavery, et al., 2008) activities and
experimental design/biosecurity during and after a trial
(Benedict, et al., 2008). A robust and defensible regulatory
framework allows public confidence in approved trials and
reduces the likelihood of a trial being halted prematurely
due to previously unvoiced concerns (Brown, et al., 2014).
As regulation of gene drive trials is expected to take place
on a case-by-case basis (Oye, et al., 2014) a local participant
with knowledge of the regional ecological, social,
economic, political and cultural context of deployment is
invaluable. Additionally, due to the relative complexity and
large scale (both temporal and spatial) expected of a gene
drive trial, access to experienced research teams provided
by a local collaborator would likely be necessary.
How should trials be conducted?
Practical guidance on how to conduct field-trials of selflimiting GPM mosquitoes (e.g. aspects of experimental
design, safety and efficiency endpoints) is available
(Benedict, et al., 2008; Brown, et al., 2014) and has been
extended to the case of gene drives (WHO/TDR, 2014;
NASEM, 2016). We will not replicate this discussion, but
instead focus on how molecular designs can be utilised to
increase the safety of a gene drive trial.
Proactive approaches
Proactive designs aim to limit the probability of
transgene escape in the first instance. ‘Precision’ CRISPRCas9 gene drives (Esvelt, et al., 2014), which have been
demonstrated in yeast (DiCarlo, et al., 2015) target the
Cas9 endonuclease to cut a fixed DNA sequence in the
genome unique to the specific target population, with the
gene drive transgene then copied across into the cut site.
The occurrence of such unique targeting sites is more
probable in isolated populations derived recently from
small numbers of initial founders and therefore may be
particularly useful against island invasives. Alternatively,
a ‘daisy-chain’ drive design could be employed (Noble, et
al., 2016). Here a CRISPR-Cas9 gene drive is divided into
a linear series of sub-components where each component
will only drive in the presence of the component directly
beneath it in the series. Critically, the basal component
in a daisy-chain cannot drive and will be subject to loss
over time through purifying selection. These components
are then integrated at independent loci in a release strain
meaning that the system is constrained spatially and
taxonomically (multiple, sequential, components must
escape an island population in the same individual or be
combined again through interbreeding in order to continue
driving) and temporally (selection will erode each basal
component of the daisy-chain in turn until it is flushed from
the population). Daisy-chain drives are currently being
investigated for the island invasives C. quinquefasciatus
and M. musculus, however analysis so far is theoretical,
with – to our knowledge – no prototype strains reported in
any metazoan.
623
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
An alternative proactive approach is to place inherent
fitness costs on a gene drive such that it will persist for
a time in a target population, potentially suppressing it,
but not increase in frequency. Proposed examples include
utilising a gene drive to spread a dominant female-lethal
transgene, as proposed for mosquitoes (RIDL-with-drive)
(Thomas, et al., 2000), and the endogenous t-haplotype
meiotic-drive system to spread the male-determining Sry
gene in mice. Although these systems utilise independent
technologies and gene targets, their effects are the same:
the transgene doubles in frequency each generation but half
those individuals inheriting it (females) are non-viable. If
transgenic individuals suffer from reduced fitness, or the
drive is less than 100% efficient at biasing its inheritance
– both of which are likely in the field – these systems will
decrease in frequency over time once deployed (Backus &
Gross, 2016), reducing the risk of transgene escape from a
trial site but also their efficiency as suppression systems.
Responsive approaches
Responsive designs are complete or partial genetic
systems, likely themselves gene drives, designed to be
deployed in the event of an escaped drive in order to curtail
its spread and potentially remove it or its phenotypic
effects from a non-target population. These can include
for example a ‘reversal-drive’ designed to target, spread
into and disrupt the DNA sequence of an escaped drive,
or the ‘immunisation-drive’ designed to spread into a nontarget population and recode the wild-type target locus,
making it unrecognisable to an escaped drive (Esvelt, et
al., 2014). These designs can be combined into a single
‘immunising-reversal’ drive and be made less invasive
through using the daisy-chain architecture. A more complex
‘restoration-drive’ design integrates a relatively noninvasive underdominance system (Figs 1 and 2) into this
daisy-chain ‘immunising-reversal’ drive to theoretically
allow the entire system to be flushed from the non-target
population once the escaped drive has been halted (Min,
et al., 2017).
Although reversal drives have been demonstrated in
lab yeast colonies (DiCarlo, et al., 2015) and a non-driving
equivalent in Drosophila (Wu, et al., 2016) it is unclear
how effective these and other responsive approaches would
be in the field. There is also concern that, in the event of
an escaped gene drive, there may be considerable pressure
against rectifying the situation through the release of another
gene drive. A more realistic, but not mutually exclusive,
approach would be to integrate a high level of conventional
control methods at all potential transgene escape points
(e.g. connected docking areas/airstrips) during and for a
period after a trial. It is clear that responsive approaches
should not be relied upon as critical containment methods
during a gene drive trial.
Case study: Genetic Biocontrol of Invasive Rodents
(GBIRd)
The Genetic Biocontrol of Invasive Rodents (GBIRd)
programme aims to develop multiple gene drive systems
in mice (Mus musculus) for simultaneous evaluation of
their safety and efficacy using biosafety standards beyond
those required by existing law, while carefully assessing
the social, cultural and policy acceptability of such an
approach (Campbell et al., 2019). The programme’s
first stage culminates in the potential submission of an
application to a regulatory agency for release of gene
drive constructed mice with a spatial control mechanism
on a small, biosecure island to test eradication of the wild,
invasive mouse population (Campbell et al., 2019). This
step-wise approach follows recommendations from USA
and Australian National Academies of Sciences (NASEM,
624
2016; AAS, 2017). Ecological criteria for selecting an
appropriate trial island for this application have been
proposed (Campbell et al., 2019; Table 1). However,
these criteria are just an initial filter and additional steps
will be required prior to any potential field trial, including
engagement with stakeholders (e.g. land managers, local
communities) and regulators to determine final approval
(Campbell et al., 2019).
Mus musculus are non-native in countries within
the GBIRd partnership (Australia, New Zealand, USA).
Mice are not consumed as a food item by people in these
countries; negatively impact native species, stored foods,
crops, and infrastructure and can carry zoonotic diseases
that impact the health of people and their livestock
(Stenseth, et al., 2003; Meerburg, et al., 2009; Capizzi, et
al., 2014), likely increasing socio-political acceptability.
Further, these countries have (or are expected to have)
appropriate regulatory capacity and systems established to
evaluate a GBIRd proposal, if one is submitted (Campbell
et al., 2019). Idealised island selection criteria for potential
trials within these countries are provided in Table 1.
CONCLUSION
Gene drives hold enormous potential for application
against invasive species and there is increasing interest
in adapting them to this purpose. As a transformative
but controversial set of technologies, it is important that
the first instances of their use in the field are successful,
both in terms of efficacy and safety. As discussed, the
likelihood of a successful trial can be increased by making
appropriate decisions at multiple stages of a gene drive’s
development and deployment. Making these decisions
requires input from a broad range of scientific disciplines
(Gould, 2008; Piaggio, et al., 2017) involving, for example,
conservationists identifying potential targets, ecologists
advising on the biological appropriateness of these
targets and efficiencies of different gene drive strategies,
molecular biologists advising on the feasibility of building
proposed designs, mathematical modellers devising the
most efficient means of deploying these systems and,
finally, managers who will ultimately advise on the logistic
feasibility of deployment. Although described in a linear
series, in practice this will require informed dialogue
between all these parties from the outset – there is no point
in developing a system that performs well in computer
models or in the lab if it is ultimately deemed impractical
to deploy in the field. With proof-of-principle suppression
(Hammond, et al., 2016) and replacement (Gantz, et al.,
2015) drives functional in anopheline mosquitoes, it is
critical that these conversations begin now to ensure these
technologies are applied as safely, efficiently and rapidly as
possible to the control of invasive species.
ACKNOWLEDGEMENTS
THS and LA are supported by European Union H2020
Grant nEUROSTRESSPEP (634361). LA is supported by
core funding from the UK Biotechnology and Biological
Sciences Research Council (BBSRC) to the Pirbright
Institute (BBS/E/I/00001892). ME is funded on a
Wellcome Trust Investigator Award (110117/Z/15/Z) made
to LA. KC is supported by Island Conservation.
REFERENCES
AAS. (2017). Synthetic gene drives in Australia: implications of emerging
technologies. Australian Academy of Science. <www.science.org.au/
gene-drives>. Accessed 17 November 2017.
Akbari, O.S., Matzen, K.D., Marshall, J.M., Huang, H.X., Ward, C.M. and
Hay, B.A. (2013). ‘A synthetic gene drive system for local, reversible
modification and suppression of insect populations’. Current Biology
23(8): 671–677.
Harvey-Samuel, et al.: Trialling gene drives
Akbari, O.S., Bellen, H.J., Bier, E., Bullock, S.L., Burt, A., Church, G.M.,
Cook, K.R., Duchek, P., Edwards, O.R., Esvelt, K.M., Gantz, V.M.,
Golic, K.G., Gratz, S.J., Harrison, M.M., Hayes, K.R., James, A.A.,
Kaufman, T.C., Knoblich, J., Malik, H.S., Matthews, K.A., O’ConnorGiles, K.M., Parks, A.L., Perrimon, N., Port, F., Russell, S., Ueda, R.
and Wildonger, J. (2015). ‘Safeguarding gene drive experiments in the
laboratory’. Science 349(6251): 927–929.
Allen, M.L., O’Brochta, D.A., Atkinson, P.W. and Levesque, C.S. (2001).
‘Stable, germ-line transformation of Culex quinquefasciatus (Diptera :
Culicidae)’. Journal of Medical Entomology 38(5): 701–710.
Alphey, L. (2002). ‘Re-engineering the sterile insect technique’. Insect
Biochemistry and Molecular Biology 32(10): 1243–1247.
Alphey, L., Benedict, M., Bellini, R., Clark, G.G., Dame, D.A., Service,
M.W. and Dobson, S.L. (2010). ‘Sterile-insect methods for control of
mosquito-borne diseases: An analysis’. Vector-Borne and Zoonotic
Diseases 10(3): 295–311.
Alphey, L. (2014). ‘Genetic control of mosquitoes’. Annual Review of
Entomology 59: 205–224.
Amori, G., Gippoliti, S. and Helgen, K.M. (2008). ‘Diversity, distribution,
and conservation of endemic island rodents’. Quaternary International
182: 6–15.
Angel, A., Wanless, R.M. and Cooper, J. (2009). ‘Review of impacts of
the introduced house mouse on islands in the Southern Ocean: are mice
equivalent to rats?’. Biological Invasions 11(7): 1743–1754.
Aplin, K.P., Chesser, T. and ten Haven, J. (2003). ‘Evolutionary Biology
of the Genus Rattus: Profile of an Archetypal Rodent Pest’. In: G.R.
Singleton, L.A. Hinds, C.J. Krebs and D.M. Spratt (eds.) Rats, Mice
and People: Rodent Biology and Management, pp. 487–498. Canberra:
Australian Centre for International Agricultural Research.
Bourtzis, K., Lees, R.S., Hendrichs, J. and Vreysen, M.J.B. (2016). ‘More
than one rabbit out of the hat: Radiation, transgenic and symbiont-based
approaches for sustainable management of mosquito and tsetse fly
populations’. Acta Tropica 157: 115–130.
Brown, D.M., Alphey, L.S., McKemey, A., Beech, C. and James, A.A.
(2014). ‘Criteria for identifying and evaluating candidate sites for openfield trials of genetically engineered mosquitoes’. Vector-Borne and
Zoonotic Diseases 14(4): 291–299.
Bull, J.J. (2015). ‘Evolutionary decay and the prospects for long-term
disease intervention using engineered insect vectors’. Evolution,
Medicine and Public Health 2015(1): 152–166.
Bull, J.J. (2016). ‘OUP: Lethal gene drive selects inbreeding’. Evolution,
Medicine and Public Health 2017(1): 1–16.
Burt, A. (2003). ‘Site-specific selfish genes as tools for the control and
genetic engineering of natural populations’. Proceedings of the Royal
Society B-Biological Sciences 270(1518): 921–928.
Campbell, K.J., Beek, J., Eason, C.T., Glen, A.S., Godwin, J., Gould, F.,
Holmes, N.D., Howald, G.R., Madden, F.M., Ponder, J.B., Threadgill,
D.W., Wegmann, A.S. and Baxter, G.S. (2015). ‘The next generation
of rodent eradications: Innovative technologies and tools to improve
species specificity and increase their feasibility on islands’. Biological
Conservation 185: 47–58.
Campbell, K.J. Saah, J.R., Brown, P.R., Godwin, J., Gould, F., Howald,
G.R., Piaggio, A., Thomas, P., Tompkins, D.M., Threadgill, D.,
Delborne, J., Kanavy, D.M., Kuikin, T., Packard, H., Serr, M. and Shiels,
A. (2019). ‘A potential new tool for the toolbox: assessing gene drives
for eradicating invasive rodent populations’. In: C.R. Veitch, M.N.
Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) Island invasives:
scaling up to meet the challenge, pp. 6–14. Occasional Paper SSC no.
62. Gland, Switzerland: IUCN.
Arensburger, P., Megy, K., Waterhouse, R.M., Abrudan, J., Amedeo, P.,
Antelo, B., Bartholomay, L., Bidwell, S., Caler, E., Camara, F., Campbell,
C.L., Campbell, K.S., Casola, C., Castro, M.T., Chandramouliswaran,
I., Chapman, S.B., Christley, S., Costas, J., Eisenstadt, E., Feschotte,
C., Fraser-Liggett, C., Guigo, R., Haas, B., Hammond, M., Hansson,
B.S., Hemingway, J., Hill, S.R., Howarth, C., Ignell, R., Kennedy, R.C.,
Kodira, C.D., Lobo, N.F., Mao, C.H., Mayhew, G., Michel, K., Mori, A.,
Liu, N.N., Naveira, H., Nene, V., Nguyen, N., Pearson, M.D., Pritham,
E.J., Puiu, D., Qi, Y.M., Ranson, H., Ribeiro, J.M.C., Roberston, H.M.,
Severson, D.W., Shumway, M., Stanke, M., Strausberg, R.L., Sun,
C., Sutton, G., Tu, Z.J., Tubio, J.M.C., Unger, M.F., Vanlandingham,
D.L., Vilella, A.J., White, O., White, J.R., Wondji, C.S., Wortman, J.,
Zdobnov, E.M., Birren, B., Christensen, B.M., Collins, F.H., Cornel,
A., Dimopoulos, G., Hannick, L.I., Higgs, S., Lanzaro, G.C., Lawson,
D., Lee, N.H., Muskavitch, M.A.T., Raikhel, A.S. and Atkinson, P.W.
(2010). ‘Sequencing of Culex quinquefasciatus establishes a platform
for mosquito comparative genomics’. Science 330(6000): 86–88.
Deredec, A., Burt, A. and Godfray, H.C. (2008). ‘The population genetics
of using homing endonuclease genes in vector and pest management’.
Genetics 179(4): 2013–2026.
Backus, G.A. and Gross, K. (2016). ‘Genetic engineering to eradicate
invasive mice on islands: modelling the efficiency and ecological
impacts’. Ecosphere 7(12): 01589. 10.1002/ecs2.1589.
DiCarlo, J.E., Chavez, A., Dietz, S.L., Esvelt, K.M. and Church, G.M.
(2015). ‘Safeguarding CRISPR-Cas9 gene drives in yeast’. Nature
Biotechnology 33(12): 1250–1255.
Banks, P.B. and Hughes, N.K. (2012). ‘A review of the evidence for
potential impacts of black rats (Rattus rattus) on wildlife and humans in
Australia’. Wildlife Research 39(1): 78–88.
Dlugosch, K.M. and Parker, I.M. (2008). ‘Founding events in species
invasions: Genetic variation, adaptive evolution, and the role of multiple
introductions’. Molecular Ecology 17(1): 431–449.
Bax, N.J. and Thresher, R.E. (2009). ‘Ecological, behavioral, and
genetic factors influencing the recombinant control of invasive pests’.
Ecological Applications 19(4): 873–888.
Eckhoff, P.A., Wenger, E.A., Godfray, H.C.J. and Burt, A. (2017). ‘Impact
of mosquito gene drive on malaria elimination in a computational
model with explicit spatial and temporal dynamics’. Proceedings of the
National Academy of Sciences 114(2): E255–E264.
Beaghton, A., Beaghton, P.J. and Burt, A. (2016). ‘Gene drive through
a landscape: Reaction-diffusion models of population suppression and
elimination by a sex ratio distorter’. TheoreticalPpopulation Biology
108: 51–69.
Bellemain, E. and Ricklefs, R.E. (2008). ‘Are islands the end of the
colonization road?’. Trends in Ecology and Evolution 23(8): 461–468.
Benedict, M., D’Abbs, P., Dobson, S., Gottlieb, M., Harrington, L.,
Higgs, S., James, A., James, S., Knols, B., Lavery, J., O’Neill, S., Scott,
T., Takken, W., Toure, Y. and Containe, C.W.G.G. (2008). ‘Guidance for
contained field trials of vector mosquitoes engineered to contain a gene
drive system: Recommendations of a scientific working group’. VectorBorne and Zoonotic Diseases 8(2): 127–166.
Besansky, N. (2015). Ecological & Evolutionary Conditions for Geneflow in Mosquitoes. Science, Ethics, and Governance Considerations
for Gene Drive Research. National Academies of Science, Engineering
and Medicine. <http://nas-sites.org/gene-drives/2015/10/03/secondpublic-meeting/>. Accessed 17 November 2017.
Capizzi, D., Bertolino, S. and Mortelliti, A. (2014). ‘Rating the rat: Global
patterns and research priorities in impacts and management of rodent
pests’. Mammal Review 44(2): 148–162.
Clobert, J., Massot, M. and Le Galliard, J. (2012). ‘Multi-determinism
in Natal Dispersal: The Common Lizard as a Model System’. In: J.
Clobert, M. Baguette, T.G. Benton and J.M. Bullock (eds.) Dispersal
Ecology and Evolution, pp. 29–40. Oxford, UK: Oxford University
Press.
Crisp, A., Boschetti, C., Perry, M., Tunnacliffe, A. and Micklem, G. (2015).
‘Expression of multiple horizontally acquired genes is a hallmark of
both vertebrate and invertebrate genomes’. Genome Biology 16: https://
doi.org/10.1186/s13059-015-0607-3.
Eldridge, B.F. (2005). ‘Mosquitoes, the Culicidae’. In: W.C. Marquardt
and B.C. Kondratieff (eds.) Biology of Disease Vectors, pp. 95–111.
London, UK: Elsevier Academic Press.
Esvelt, K.M., Smidler, A.L., Catteruccia, F. and Church, G.M. (2014).
‘Concerning RNA-guided gene drives for the alteration of wild
populations’. eLife: doi:10.7554/eLife.03401.
Fritts, E.I. (2007).’Wildlife and People at Risk: A Plan to Keep Rats Out of
Alaska. Juneau, AK: Alaska Department of Fish and Game.
Gantz, V.M., Jasinskiene, N., Tatarenkova, O., Fazekas, A., Macias,
V.M., Bier, E. and James, A.A. (2015). ‘Highly efficient Cas9-mediated
gene drive for population modification of the malaria vector mosquito
Anopheles stephensi’. Proceedings of the National Academy of Sciences
112(49): E6736-E6743.
Bonsall, M.B., Yakob, L., Alphey, N. and Alphey, L. (2010). ‘Transgenic
control of vectors: The effects of interspecific interactions’. Israel
Journal of Ecology and Evolution 56(3–4): 353–370.
625
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
Gibbs, R.A., Weinstock, G.M., Metzker, M.L., Muzny, D.M., Sodergren,
E.J., Scherer, S., Scott, G., Steffen, D., Worley, K.C., Burch, P.E.,
Okwuonu, G., Hines, S., Lewis, L., DeRamo, C., Delgado, O.,
Dugan-Rocha, S., Miner, G., Morgan, M., Hawes, A., Gill, R., Holt,
R.A., Adams, M.D., Amanatides, P.G., Baden-Tillson, H., Barnstead,
M., Chin, S., Evans, C.A., Ferriera, S., Fosler, C., Glodek, A., Gu,
Z.P., Jennings, D., Kraft, C.L., Nguyen, T., Pfannkoch, C.M., Sitter,
C., Sutton, G.G., Venter, J.C., Woodage, T., Smith, D., Lee, H.M.,
Gustafson, E., Cahill, P., Kana, A., Doucette-Stamm, L., Weinstock, K.,
Fechtel, K., Weiss, R.B., Dunn, D.M., Green, E.D., Blakesley, R.W.,
Bouffard, G.G., de Jong, J., Osoegawa, K., Zhu, B.L., Marra, M.,
Schein, J., Bosdet, I., Fjell, C., Jones, S., Krzywinski, M., Mathewson,
C., Siddiqui, A., Wye, N., McPherson, J., Zhao, S.Y., Fraser, C.M.,
Shetty, J., Shatsman, S., Geer, K., Chen, Y.X., Abramzon, S., Nierman,
W.C., Gibbs, R.A., Weinstock, G.M., Havlak, P.H., Chen, R., Durbin,
K.J., Egan, A., Ren, Y.R., Song, X.Z., Li, B.S., Liu, Y., Qin, X.,
Cawley, S., Weinstock, G.M., Worley, K.C., Cooney, A.J., Gibbs, R.A.,
D’Souza, L.M., Martin, K., Wu, J.Q., Gonzalez-Garay, M.L., Jackson,
A.R., Kalafus, K.J., McLeod, M.P., Milosavljevic, A., Virk, D., Volkov,
A., Wheeler, D.A., Zhang, Z.D., Bailey, J.A., Eichler, E.E., Tuzun, E.,
Birney, E., Mongin, E., Ureta-Vidal, A., Woodwark, C., Zdobnov, E.,
Bork, P., Suyama, M., Torrents, D., Alexandersson, M., Trask, B.J.,
Young, J.M., Smith, D., Huang, H., Fechtel, K., Wang, H.J., Xing,
H.M., Weinstock, K., Daniels, S., Gietzen, D., Schmidt, J., Stevens,
K., Vitt, U., Wingrove, J., Camara, F., Schmidt, J., Stevens, K., Vitt,
U., Wingrove, J., Camara, F., Alba, M.M., Abril, J.F., Guigo, R., Smit,
A., Dubchak, I., Rubin, E.M., Couronne, O., Poliakov, A., Hubner, N.,
Ganten, D., Goesele, C., Hummel, O., Kreitler, T., Lee, Y.A., Monti, J.,
Schulz, H., Zimdahl, H., Himmelbauer, H., Lehrach, H., Jacob, H.J.,
Bromberg, S., Gullings-Handley, J., Jensen-Seaman, M.I., Kwitek, A.E.,
Lazar, J., Pasko, D., Tonellato, P.J., Twigger, S., Ponting, P., Duarte,
J.M., Rice, S., Goodstadt, L., Beatson, S.A., Emes, R.D., Winter, E.E.,
Webber, C., Brandt, P., Nyakatura, G., Adetobi, M., Chiaromonte, F.,
Elnitski, L., Eswara, P., Hardison, R.C., Hou, M.M., Kolbe, D., Makova,
K., Miller, W., Nekrutenko, A., Riemer, C., Schwartz, S., Taylor, J.,
Yang, S., Zhang, Y., Lindpaintner, K., Andrews, T.D., Caccamo, M.,
Clamp, M., Clarke, L., Curwen, V., Durbin, R., Eyras, E., Searle, S.M.,
Cooper, G.M., Batzoglou, S., Brudno, M., Sidow, A., Stone, E.A.,
Venter, J.C., Payseur, B.A., Bourque, G., Lopez-Otin, C., Puente, X.S.,
Chakrabarti, K., Chatterji, S., Dewey, C., Pachter, L., Bray, N., Yap,
V.B., Caspi, A., Tesler, G., Pevzner, P.A., Haussler, D., Roskin, K.M.,
Baertsch, R., Clawson, H., Furey, T.S., Hinrichs, A.S., Karolchik, D.,
Kent, W.J., Rosenbloom, K.R., Trumbower, H., Weirauch, M., Cooper,
D.N., Stenson, P.D., Ma, B., Brent, M., Arumugam, M., Shteynberg, D.,
Copley, R.R., Taylor, M.S., Riethman, H., Mudunuri, U., Peterson, J.,
Guyer, M., Felsenfeld, A., Old, S., Mockrin, S., Collins, F. and Cons,
R.G.S.P. (2004). ‘Genome sequence of the brown Norway rat yields
insights into mammalian evolution’. Nature 428(6982): 493–521.
Glen, A.S., Atkinson, R., Campbell, K.J., Hagen, E., Holmes, N.D.,
Keitt, B.S., Parkes, J.P., Saunders, A., Sawyer, J. and Torres, H. (2013).
‘Eradicating multiple invasive species on inhabited islands: The next big
step in island restoration?’. Biological Invasions 15(12): 2589–2603.
Gould, F. (2008). ‘Broadening the application of evolutionarily based
genetic pest management’. Evolution 62(2): 500–510.
Hamilton, W.D. (1967). ‘Extraordinary sex ratios’. Science 156(3774):
477–488.
Hammond, A., Galizi, R., Kyrou, K., Simoni, A., Siniscalchi, C.,
Katsanos, D., Gribble, M., Baker, D., Marois, E., Russell, S., Burt, A.,
Windbichler, N., Crisanti, A. and Nolan, T. (2016). ‘A CRISPR-Cas9
gene drive system targeting female reproduction in the malaria mosquito
vector Anopheles gambiae’. Nature biotechnology 34(1): 78–83.
Harper, G. and Cabrera, L. (2010). ‘Response of mice (Mus musculus)
to the removal of black rats (Rattus rattus) in arid forest on Santa Cruz
Island, Galápagos’. Biological Invasions 12(6): 1449–1452.
Harper, G. and Bunbury, N. (2015). ‘Invasive rats on tropical islands:
Their population biology and impacts on native species’. Global
Ecology and Conservation 3: 607–627.
Harris, D., Gregory, S.D., Bull, L. and Courchamp, F. (2012). ‘Island
prioritization for invasive rodent eradications with an emphasis on
reinvasion risk’. Biological Invasions 14(6): 1251–1263.
Harvey-Samuel, T., Ant, T. and Alphey, L. (2017). ‘Towards the genetic
control of invasive species’. Biological Invasions 19(6): 1683–1703.
Hodgins, K.A., Rieseberg, L. and Otto, S.P. (2009). ‘Genetic control of
invasive plants species using selfish genetic elements’. Evolutionary
Applications 2(4): 555–569.
Hoffmann, A.A., Montgomery, B.L., Popovici, J., Iturbe-Ormaetxe, I.,
Johnson, P.H., Muzzi, F., Greenfield, M., Durkan, M., Leong, Y.S.,
Dong, Y., Cook, H., Axford, J., Callahan, A.G., Kenny, N., Omodei, C.,
McGraw, E.A., Ryan, P.A., Ritchie, S.A., Turelli, M. and O’Neill, S.L.
(2011). ‘Successful establishment of Wolbachia in Aedes populations to
suppress dengue transmission’. Nature 476(7361): 454–U107.
626
Hoffmann, A.A., Iturbe-Ormaetxe, I., Callahan, A.G., Phillips, B.,
Billington, K., Axford, J.K., Montgomery, B., Turley, A.P. and O’Neill,
S.L. (2014). ‘Stability of the wMel Wolbachia infection following
invasion into Aedes aegypti populations’. PLOS Neglected Tropical
Diseases 8(9): e3115.
Huang, Y.X., Lloyd, A.L., Legros, M. and Gould, F. (2011). ‘Gene drive
into insect populations with age and spatial structure: A theoretical
assessment’. Evolutionary Applications 4(3): 415–428.
Ivics, Z., Mates, L., Yau, T.Y., Landa, V., Zidek, V., Bashir, S., Hoffmann,
O.I., Hiripi, L., Garrels, W., Kues, W.A., Bosze, Z., Geurts, A., Pravenec,
M., Rulicke, T. and Izsvak, Z. (2014). ‘Germline transgenesis in rodents
by pronuclear microinjection of Sleeping Beauty transposons’. Nature
Protocols 9(4): 773–793.
Kawecki, T. (2004). ‘Ecological and Evolutionary Consequences of
Source-Sink Population Dynamics’. In: Island Hanski and O.E.
Gaggiotti (eds.) Ecology, Genetics and Evolution of Metapopulations,
pp. 387–414. Boston: Elsevier.
Klassen, W. and Curtis, C.F. (2005). ‘History of the Sterile Insect
Technique’. In: V.A. Dyck, J. Hendrichs and A.S. Robinson (eds.) Sterile
Insect Technique: Principles and Practice in Area-Wide Integrated Pest
Management, pp. 3–36. Dordrect, Netherlands: Springer.
Kolar, C.S. and Lodge, D.M. (2001). ‘Progress in invasion biology:
Predicting invaders’. Trends in Ecology and Evolution 16(4): 199–204.
LaPointe, D.A., Atkinson, C.T. and Samuel, M.D. (2012). ‘Ecology
and conservation biology of avian malaria’. Annals of the New York
Academy of Sciences 1249: 211–226.
Lavery, J.V., Harrington, L.C. and Scott, T.W. (2008). ‘Ethical, social, and
cultural considerations for site selection for research with genetically
modified mosquitoes’. American Journal of Tropical Medicine and
Hygiene 79(3): 312–318.
Marshall, J.M. and Hay, B.A. (2012). ‘Confinement of gene drive systems
to local populations: A comparative analysis’. Journal of Theoretical
Biology 294: 153–171.
Matthysen, E. (2012). ‘Multicausality of Dispersal: A Review’. In: J.
Clobert, M. Baguette, T.G. Benton and J.M. Bullock (eds.) Dispersal
Ecology and Evolution, pp. 3–18. Oxford, UK: Oxford University Press.
Meerburg, B.G., Singleton, G.R. and Kijlstra, A. (2009). ‘Rodentborne diseases and their risks for public health’. Critical Reviews in
Microbiology 35(3): 221–270.
Min, J., Noble, C., Najjar, D. and Esvelt, K. (2017). ‘Daisy quorum drives
for the genetic restoration of wild populations’. bioRχiv http://dx.doi.
org/101101/115618.
Mooney, H.A. and Cleland, E.E. (2001). ‘The evolutionary impact of
invasive species’. Proceedings of the National Academy of Sciences
98(10): 5446–5451.
NASEM. (2016). Gene Drives on the Horizon: Advancing Science,
Navigating Uncertainty, and Aligning Research with Public Values,
Washington (DC): The National Academies Press.
Noble, C., Min, J., Olejarz, J., Buchthal, J., Chavez, A., Smidler, A.,
DeBenedictis, E., Church, G., Nowak, M. and Esvelt, K. (2016). ‘Daisychain gene drives for the alteration of local populations’:. bioRχiv:
https://doi.org/10.1101/057307.
North, A., Burt, A. and Godfray, H.C.J. (2013). ‘Modelling the spatial
spread of a homing endonuclease gene in a mosquito population’.
Journal of Applied Ecology 50(5): 1216–1225.
Oppel, S., Beaven, B.M., Bolton, M., Vickery, J. and Bodey, T.W. (2011).
‘Eradication of invasive mammals on islands inhabited by humans and
domestic animals’. Conservation Biology 25(2): 232–240.
Ortiz, M.F., Wallau, G.L., Graichen, D.A.S. and Loreto, E.L.S. (2015).
‘An evaluation of the ecological relationship between Drosophila
species and their parasitoid wasps as an opportunity for horizontal
transposon transfer’. Molecular Genetics and Genomics 290(1): 67–78.
Oye, K.A., Esvelt, K., Appleton, E., Catteruccia, F., Church, G., Kuiken,
T., Lightfoot, S.B.Y., McNamara, J., Smidler, A. and Collins, J.P.
(2014). ‘Regulating gene drives’. Science 345(6197): 626–628.
Piaggio, A.J., Segelbacher, G., Seddon, P.J., Alphey, L., Bennett, E.L.,
Carlson, R.H., Friedman, R.M., Kanavy, D., Phelan, R., Redford, K.H.,
Rosales, M., Slobodian, L. and Wheeler, K. (2017). ‘Is it time for
synthetic biodiversity conservation?’. Trends in Ecology and Evolution
32(2): 97–107.
Pimentel, D., Zuniga, R. and Morrison, D. (2005). ‘Update on the
environmental and economic costs associated with alien-invasive
species in the United States’. Ecological Economics 52(3): 273–288.
Rambuda, T.D. and Johnson, S.D. (2004). ‘Breeding systems of invasive
alien plants in South Africa: Does Baker’s rule apply?’. Diversity and
Distributions 10(5–6): 409–416.
Harvey-Samuel, et al.: Trialling gene drives
Reeves, R.G., Bryk, J., Altrock, P.M., Denton, J.A. and Reed, F.A. (2014).
‘First steps towards underdominant genetic transformation of insect
populations’. PLOS ONE 9(5): e97557.
Rhymer, J.M. and Simberloff, D. (1996). ‘Extinction by hybridization and
introgression’. Annual Review of Ecology and Systematics 27: 83–109.
Russell, J. and Clout, M. (2005). ‘Rodent Incursions on New Zealand
islands’. In: J.Parkes, M.Statham and G Edwards,. (Comps.)
Proceedings of the 13th Australasian Vertebrate Pest Conference, pp.
324–330. Lincoln, NZ: Landcare Research,.
Russell, J.C., Towns, D.R. and Clout, M. (2008). ‘Review of Rat Invasion
Biology: Implications for Island Biosecurity’. Science for Conservation
286, Wellington, NZ: Department of Conservation.
Schmidt, T.L., Barton, N.H., Rasic, G., Turley, A.P., Montgomery, B.L.,
Iturbe-Ormaetxe, I., Cook, P.E., Ryan, P.A., Ritchie, S.A., Hoffmann,
A.A., O’Neill, S.L. and Turelli, M. (2017). ‘Local introduction and
heterogeneous spatial spread of dengue-suppressing Wolbachia through
an urban population of Aedes aegypti’. PLOS Biology 15(5): e2001894.
Simberloff, D. (2014). ‘Biological invasions: What’s worth fighting and
what can be won?’. Ecological Engineering 65: 112–121.
Sinkins, S.P. and Gould, F. (2006). ‘Gene drive systems for insect disease
vectors’. Nature Reviews Genetics 7(6): 427–435.
Stenseth, N.C., Leirs, H., Skonhoft, A., Davis, S.A., Pech, R.P.,
Andreassen, H.P., Singleton, G.R., Lima, M., Machang’u, R.S. and
Makundi, R.H. (2003). ‘Mice, rats, and people: The bio-economics of
agricultural rodent pests’. Frontiers in Ecology and the Environment
1(7): 367–375.
Sutherland, W.J., Aveling, R., Brooks, T.M., Clout, M., Dicks, L.V.,
Fellman, L., Fleishman, E., Gibbons, D.W., Keim, B., Lickorish,
F., Monk, K.A., Mortimer, D., Peck, L.S., Pretty, J., Rockstrom, J.,
Rodriguez, J.P., Smith, R.K., Spalding, M.D., Tonneijck, F.H. and
Watkinson, A.R. (2014). ‘A horizon scan of global conservation issues
for 2014’. Trends in Ecology and Evolution 29(1): 15–22.
Sytsma, M.D., Phillips, S. and Counihan, T.D. (2015). ‘Dreissenid
Mussel Research Priorities Workshop’. Center for Lakes and Resevoirs
Publications and Presentations 49: 11.
Thomas, D.D., Donnelly, C.A., Wood, R.J. and Alphey, L.S. (2000).
‘Insect population control using a dominant, repressible, lethal genetic
system’. Science 287(5462): 2474–2476.
Thresher, R.E., Canning, M. and Bax, N.J. (2013). ‘Demographic effects
on the use of genetic options for the control of mosquitofish, Gambusia
holbrooki’. Ecological Applications 23(4): 801–814.
Thresher, R.E., Hayes, K., Bax, N.J., Teem, J., Benfey, T.J. and Gould,
F. (2014). ‘Genetic control of invasive fish: Technological options and
its role in integrated pest management’. Biological Invasions 16(6):
1201–1216.
Tobin, P.C., Berec, L. and Liebhold, A.M. (2011). ‘Exploiting Allee effects
for managing biological invasions’. Ecology Letters 14(6): 615–624.
Unckless, R.L., Messer, P.W., Connallon, T. and Clark, A.G. (2015).
‘Modeling the manipulation of natural populations by the mutagenic
chain reaction’. Genetics 201(2): 425–431.
Waterston, R.H., Lindblad-Toh, K., Birney, E., Rogers, J., Abril, J.F.,
Agarwal, P., Agarwala, R., Ainscough, R., Alexandersson, M., An, P.,
Antonarakis, S.E., Attwood, J., Baertsch, R., Bailey, J., Barlow, K.,
Beck, S., Berry, E., Birren, B., Bloom, T., Bork, P., Botcherby, M.,
Bray, N., Brent, M.R., Brown, D.G., Brown, S.D., Bult, C., Burton,
J., Butler, J., Campbell, R.D., Carninci, P., Cawley, S., Chiaromonte,
F., Chinwalla, A.T., Church, D.M., Clamp, M., Clee, C., Collins, F.S.,
Cook, L.L., Copley, R.R., Coulson, A., Couronne, O., Cuff, J., Curwen,
V., Cutts, T., Daly, M., David, R., Davies, J., Delehaunty, K.D., Deri, J.,
Dermitzakis, E.T., Dewey, C., Dickens, N.J., Diekhans, M., Dodge, S.,
Dubchak, I., Dunn, D.M., Eddy, S.R., Elnitski, L., Emes, R.D., Eswara,
P., Eyras, E., Felsenfeld, A., Fewell, G.A., Flicek, P., Foley, K., Frankel,
W.N., Fulton, L.A., Fulton, R.S., Furey, T.S., Gage, D., Gibbs, R.A.,
Glusman, G., Gnerre, S., Goldman, N., Goodstadt, L., Grafham, D.,
Graves, T.A., Green, E.D., Gregory, S., Guigo, R., Guyer, M., Hardison,
R.C., Haussler, D., Hayashizaki, Y., Hillier, L.W., Hinrichs, A., Hlavina,
W., Holzer, T., Hsu, F., Hua, A., Hubbard, T., Hunt, A., Jackson, I.,
Jaffe, D.B., Johnson, L.S., Jones, M., Jones, T.A., Joy, A., Kamal, M.,
Karlsson, E.K., Karolchik, D., Kasprzyk, A., Kawai, J., Keibler, E.,
Kells, C., Kent, W.J., Kirby, A., Kolbe, D.L., Korf, I., Kucherlapati,
R.S., Kulbokas, E.J., Kulp, D., Landers, T., Leger, J.P., Leonard, S.,
Letunic, I., Levine, R., Li, J., Li, M., Lloyd, C., Lucas, S., Ma, B.,
Maglott, D.R., Mardis, E.R., Matthews, L., Mauceli, E., Mayer, J.H.,
McCarthy, M., McCombie, W.R., McLaren, S., McLay, K., McPherson,
J.D., Meldrim, J., Meredith, B., Mesirov, J.P., Miller, W., Miner, T.L.,
Mongin, E., Montgomery, K.T., Morgan, M., Mott, R., Mullikin, J.C.,
Muzny, D.M., Nash, W.E., Nelson, J.O., Nhan, M.N., Nicol, R., Ning,
Z., Nusbaum, C., O’Connor, M.J., Okazaki, Y., Oliver, K., Larty, E.O.,
Pachter, L., Parra, G., Pepin, K.H., Peterson, J., Pevzner, P., Plumb, R.,
Pohl, C.S., Poliakov, A., Ponce, T.C., Ponting, C.P., Potter, S., Quail,
M., Reymond, A., Roe, B.A., Roskin, K.M., Rubin, E.M., Rust, A.G.,
Santos, R., Sapojnikov, V., Schultz, B., Schultz, J., Schwartz, M.S.,
Schwartz, S., Scott, C., Seaman, S., Searle, S., Sharpe, T., Sheridan, A.,
Shownkeen, R., Sims, S., Singer, J.B., Slater, G., Smit, A., Smith, D.R.,
Spencer, B., Stabenau, A., Strange-Thomann, N.S., Sugnet, C., Suyama,
M., Tesler, G., Thompson, J., Torrents, D., Trevaskis, E., Tromp, J.,
Ucla, C., Vidal, A.U., Vinson, J.P., von Niederhausern, A.C., Wade,
C.M., Wall, M., Weber, R.J., Weiss, R.B., Wendl, M.C., West, A.P.,
Wetterstrand, K., Wheeler, R., Whelan, S., Wierzbowski, J., Willey, D.,
Williams, S., Wilson, R.K., Winter, E., Worley, K.C., Wyman, D., Yang,
S., Yang, S.P., Zdobnov, E.M., Zody, M.C., Lander, E.S. and Consor,
M.G.S. (2002). ‘Initial sequencing and comparative analysis of the
mouse genome’. Nature 420(6915): 520–562.
Webber, B.L., Raghu, S. and Edwards, O.R. (2015). ‘Opinion: Is CRISPRbased gene drive a biocontrol silver bullet or global conservation
threat?’. Proceedings of the National Academy of Sciences 112(34):
10565–10567.
WHO/TDR (2014). The Guidance Framework for Testing Genetically
Modified Mosquitoes, Geneva, Switzerland: World Health Organisation.
Wu, B., Luo, L. and Gao, X.J. (2016). ‘Cas9-triggered chain ablation of
cas9 as a gene drive brake’. Nature Biotechnology 34: 137–138.
Yakob, L., Alphey, L. and Bonsall, M.B. (2008). ‘Aedes aegypti
control: The concomitant role of competition, space and transgenic
technologies’. Journal of Applied Ecology 45(4): 1258–1265.
Yakob, L. and Bonsall, M.B. (2009). ‘Importance of space and
competition in optimizing genetic control strategies’. Journal of
Economic Entomology 102(1): 50–57.
627
N.D. Holmes, B.S. Keitt, D.R. Spatz, D.J. Will, S. Hein, J.C. Russell, P. Genovesi, P.E. Cowan and B.R. Tershy
Holmes, N.D.; B.S. Keitt, D.R. Spatz, D.J. Will, S. Hein, J.C. Russell, P. Genovesi, P.E. Cowan and B.R. Tershy.
Tracking invasive species eradications on islands at a global scale
Tracking invasive species eradications on islands at a global scale
N.D. Holmes1, B.S. Keitt1, D.R. Spatz1,2, D.J. Will1, S. Hein1, J.C. Russell3, P. Genovesi4, P.E. Cowan5 and B.R. Tershy2
Island Conservation, 2100 Delaware Avenue Suite 1, Santa Cruz California USA 95060. <nick.holmes@
islandconservation.org>. 2Coastal Conservation Action Laboratory, University of California at Santa Cruz, USA.
3
School of Biological Sciences, University of Auckland, Private Bag 92019, Auckland 1142, New Zealand. 4ISPRA and
Chair IUCN Invasive Species Specialist Group, Via V. Brancati 48, 00144, Rome, Italy. 5Manaaki Whenua Landcare
Research, PO Box 69040, Lincoln 7640, New Zealand.
1
Abstract Indicators for tracking conservation efforts at a global scale are rare but important tools for understanding
trends and measuring progress towards global conservation targets. Eradication of invasive species from islands is an
increasingly used conservation intervention in countries and territories around the world. With a goal of collating these
efforts, the Database of Islands and Invasive Species Eradications (DIISE) holds records of the location, target species,
year and outcome of invasive mammal and bird eradications on islands from around the world. The database is publicly
available in Spanish and English, at <diise.islandconservation.org>, and represents a partnership between the University
of California at Santa Cruz, University of Auckland, IUCN Invasive Species Specialist Group, Landcare Research and
Island Conservation. The database holds records for more than 1,200 eradication attempts. This database will continue
to be added to and evolve as new opportunities for its application arise; thus, we expect these numbers to change over
time as new events are added and knowledge about existing events improves. Updating the DIISE relies on contributions
from experts and reporting from island restoration activities. Here we present database history, parameter definitions and
database considerations. We also highlight additional studies the underlying data have contributed to, including evaluating
the native species benefit from invasive mammal eradications on islands, and global indices to track progress towards the
Convention of Biological Diversity Aichi target 9 (Invasive Alien Species), that explicitly requires an increased effort of
eradication of priority invasive species.
Keywords: database, eradication, global, invasive, islands, mammal
INTRODUCTION
Eradication of invasive species from islands is an
increasingly used conservation intervention in countries
and territories around the world. Indicators for tracking
conservation efforts at a global scale are rare but important
tools for understanding trends, and measuring progress
towards global conservation targets (McGeoch, et al.,
2010). The number of eradications of invasive species
on islands is one response indicator that contributes to
measuring such progress. The number of eradications of
invasive species on islands is a particularly good metric
as these events tend to take place over discrete periods
of time, occur in clearly defined spatial areas, and have a
clear measure of success or failure (Niemeijer & de Groot,
2008).
With a goal of collating these efforts, the Database of
Islands and Invasive Species Eradications (DIISE) holds
records of, at a minimum, the location, target species, year
and outcome of invasive mammal and bird eradications on
islands around the world. Data within the database focus
on terrestrial vertebrate species, primarily mammals and
birds. Fish eradications are not included, nor are plant
or invertebrate eradications (but see Tobin, et al., 2014;
Hoffmann, et al., 2016). As of 2016, the database holds
records for more than 1,200 eradication attempts. The
database is publicly available in Spanish and English,
at <diise.islandconservation.org>, and represents an
ongoing partnership between the University of California
at Santa Cruz, University of Auckland, IUCN Invasive
Species Specialist Group, Landcare Research and Island
Conservation.
Here we present database history, parameter definitions
and database considerations. During 2017, a major update
to the data is underway with a goal of using the 2017
Island Invasives Conference as a venue to engage island
restoration practitioners to help improve the dataset.
MATERIALS AND METHODS
Database history
The first synthesis of the database (then known as the
Global Islands Invasive Vertebrate Eradication Database)
was published in the proceedings of the Island Invasives:
eradication and management conference in Auckland
in 2010 (Keitt, et al., 2011). Data for this synthesis were
gathered from published, grey and unpublished literature,
with the majority of data from reviews of eradications for
rodents (Howald, et al., 2007), goats (Campbell & Donlan,
2005) and cats (Nogales, et al., 2004; Campbell, et al.,
2011). Following the conference, the database was shared
with all of the attendees of the conference (240 topic
experts from 20 countries) with the goals of checking facts
and adding missing eradication events. Attendees were
encouraged to share the database with their networks to
help achieve these goals.
In 2013–2014 an update of the database was undertaken
using additional review papers on invasive mice (MacKay,
et al., 2007) and small Indian mongoose (Barun, et al.,
2011), the two Island Invasives conference proceedings
(Veitch & Clout, 2002; Veitch, et al., 2011), summaries
of eradication on inhabited islands (Oppel, et al., 2011;
Glen, et al., 2013) and regional summaries for New
Zealand (Clout & Russell, 2006), Europe and overseas
territories (Genovesi, 2005; Genovesi & Carnevali, 2011),
USA and territories (Witmer & Fuller, 2011), Galapagos
(Carrion, et al., 2011; Harper & Carrion, 2011; Phillips,
et al., 2012), California Channel and north-western Baja
California Islands (McChesney & Tershy, 1998), Mexico
(Aguirre-Muñoz, et al., 2008; Aguirre-Muñoz, et al.,
2011), Hawaii and Central Pacific (Hess & Jacobi, 2011),
France and overseas territories (Lorvelec & Pascal, 2005)
and Seychelles (Beaver & Mougal, 2009).
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
628
up to meet the challenge, pp. 628–632. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Holmes, et al.: Tracking invasive species eradications
Other resources reviewed include, but were not limited
to, IUCN SSC Invasive Species Specialist Group Invasives
listserv, Pacific Seabird Group listserv, Pacific Invasives
Initiative listserv; new sites including Agreement for the
Conservation of Albatrosses and Petrels <http://acap.aq/
news>, Seychelles Island Foundation newsletter <http://
www.sif.sc/index.php?langue=eng&rub=19>;
industry
sources including the Australian Invasive Animals
Cooperative Research Centre <https://www.pestsmart.
org.au/tag/invasive-animals-cooperative-research-centre/>
and
<https://invasives.com.au/about/our-legacy/>,
Mediterranean Small Islands Initiative <http://initiativepim.org/> and the Web of Science for the key words
“island” and “eradication”. Further, we were fortunate
to benefit from communications with practitioners who
maintain regularly updated databases for territories
including the Falklands / Islas Malvinas (Falkland Islands
Rat Eradication Register, S. Poncet pers. comm.), France
and overseas territories (O. Lorvelec pers. comm.),
Seychelles (G. Rocamura pers. comm.), and worldwide (J.
Parkes pers. comm.). This effort also included an evaluation
period where entries were cross-checked with experts, and
review of emails sent to directly to database managers.
During 2017, a third update began, including review of
regional assessments including Italy (Capizzi, et al., 2016),
Australia (Gregory, et al., 2014), California (California
Department of Fish and Wildlife (CDFW, 2015) and the
Indian Ocean (Russell, et al., 2016). Additional listservs
and new sites reviewed include the NZ Department of
Conservation
media
<http://www.doc.govt.nz/news/
media-releases/>, South Pacific Regional Environment
Program media <http://www.sprep.org/news>, Pacific
Invasives Learning Network soundbites <http://www.
sprep.org/piln/soundbites-documents>, Battler resource
base <https://piln.sprep.org/>, and BirdLife news <http://
www.birdlife.org/news>. The keyword ‘eradication’ was
used to search these sites, plus the word ‘deratisation’ for
French language sites. This review is expected to continue
through 2017 including an expert review to validate new
or changed entries.
Parameter definitions
Keitt et al (2011) describe the general methods used to
populate the DIISE for the first synthesis. Each eradication
event is an attempt to eradicate an invasive vertebrate
population from an island. Where multiple invasive species
are eradicated from an island these are considered separate
eradication events, even if using the same technique. Each
eradication event has a unique identification number and
can generally be identified by the combination of the key
parameters of species removed + island + eradication end
date + eradication status. Citations for each eradication
event are recorded.
For the 2013–2014 update, the parameter definitions
were expanded to also include data quality, primarily to
classify how eradication events were verified for inclusion
the database. We assessed the quality of data available for
all eradication attempts within the database using criteria
in Table 1. We encourage other users of DIISE data to use
data classified as good or satisfactory data quality event
only. We retain events classified as poor data quality in
the online database in the hope others can help us further
qualify or remove these events.
Each eradication event was linked to an island.
Each island was given a unique ID based on the World
Conservation Monitoring Centre (WCMC) Global Islands
Database (GID) (Depraetere, 2007), a spatial dataset with
180,000 unique island locations of the world. Eradications
on different islands were recorded as separate events,
regardless of whether it was in the same archipelago or
treated concurrently (e.g. Montebello islands in Western
Australia). For coral atolls, if the project targeted
individual motu these were treated as separate events and
linked to individual motu accordingly. However, projects
that occurred at the atoll scale were treated as one event.
For islands that were not in the GID we allocated our
own ID number and metadata. Locations were verified
in Google Earth and corrected if necessary. Island names
are standardised to the common proper noun within the
larger country/territory, excluding frequently used words
for ‘island’ (e.g. islets, rocks, etc.). Country or territory
was based on International Standards Organization (ISO)
3166-1 alpha-2 codes. In 2016, the DIISE island locations
were migrated to the GID2, a higher resolution product
by WCMC that holds approximately 460,000 islands.
Each polygon used for the DIISE was validated for island
location and size against Google Earth and other satellite
imagery.
Each invasive species has a unique ID code, and the
common name, scientific name, family, trophic level
(omnivore, herbivore, carnivore), and nominate type
[amphibian, flying bird; non-flying bird; rodents (Mus);
rodents (Rattus); cat; dogs or foxes, mongooses or weasels,
rabbits or hares, reptiles (excluding snakes), snakes,
ungulates, or other mammals] were recorded. Invasive
species populations were either classified as feral, semiferal, domestic, or a combination, with semi-feral defined
as having some human care but not restricted in movement
(e.g. fences).
We also sought to classify the eradication type,
based upon the extent of the established invasive species
population on the island and thus the scale of the operation
necessary to achieve eradication. The aim of the database
is to only include events where the goal was complete
removal of an invasive species population from the island,
and not removal from only part of an island such as fenced
Table 1 Data quality definitions
Data quality
Good
Satisfactory
Poor
Unknown
Data quality definition
We can verify the attempt; we have a copy of the primary reference (e.g. from a report, or peer
reviewed publication) that details the effort, typically allowing us to populate almost all fields
An expert practitioner has verified the event and/or we have limited information about an eradication
but what we do have has come from a verifiable source (e.g. email from a reputable practitioner or
cited in a review paper), and we can typically identify all of the following attributes: the island, end
year (if applicable), invasive species type, eradication status, and primary eradication method
We cannot verify the attempt (conflicting information nor unverifiable resource) and/or we lack
evidence for at least one of the following parameters: island, end year (if applicable), invasive animal
type, eradication status, or primary eradication method
The data quality has not yet been assessed for this event
629
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
areas (however, note we retain events where fences are
used as a tool to achieve eradication at an island scale). We
delineate whether the operation required treatment of the
entire island, or only part of the island (restricted range), to
achieve eradication of the invasive species population at the
island scale. We also delineate between incursion responses
and restricted range, whereby incursions represent
operations to remove a recently arrived population prior
to their spread across the island. If an incursion response
fails, it is assumed a new eradication operation would be
necessary. Although some incursion responses are recorded
in the database, there is not a deliberate attempt to record
every incursion response for each island because these may
reflect a minor or ongoing management activity that may
go unrecorded in the sources accessed. A classification of
unknown is also used if it is unclear what the eradication
type was, and this is also typically used where the cause
of the extirpation of the invasive species population is
unknown.
The timing of the eradication operation is typically
based on the end date for the operation and is reported
in years only. We considered eradication end date to be
the year that major eradication operations ceased. This
typically coincided with the end of hunting / trapping for
ungulates and predators or the end of toxicant application
(or other methods) for rodent projects. We note that
monitoring required to determine if an operation was
successful often occurs in years after the operation ending.
The primary and secondary method of the eradication is
collected, including disease, hunting, trapping, toxicant,
other, or unknown. Where toxicant was used we sought to
identify the baiting method, including aerial broadcast, bait
station or bait piles, hand broadcast, unknown, or other,
plus the toxicant compound used.
Eradication status is based on definitions in Table
2. When an eradication event is declared successful, the
target invasive was removed from the entire island. We
considered failures to be operational failures, i.e. the project
did not successfully remove the entire invasive population.
We considered reinvasion as separate to operational failure
and recorded this separately. Reinvasion was defined as
a previously successfully removed population becoming
re-established back on the island. In the case of rodent
eradications, reinvasion may also represent misdiagnosed
failure (Russell, et al., 2010) but can be assessed through
techniques such as genetic analyses, distance to potential
source populations and the time elapsed between the
eradication operation and subsequent rodent detections.
When experts or source material indicated uncertainty
about whether an invasive rodent population remains due
to an operational failure or a reinvasion back onto the
island, we assumed operational failure and classified data
quality for the event as ‘poor’.
DISCUSSION
Collating the location, method, outcome and target
animal for invasive vertebrate eradications on islands
offers a unique opportunity to contribute to global
indicators for conservation. Collating these data over time
offers insight into the response of a state-pressure-response
model (Niemeijer & de Groot, 2008). The DIISE dataset
holds many characteristics identified as necessary for
effective threat (i.e. pressure in the state-pressure-response
model) databases at a global scale, including: being freely
available, spatially explicit, inclusion of a measure of
expert validation, and is updated in a reasonable timeframe
(Joppa, et al., 2016). The DIISE can contribute towards
measuring progress of Aichi Target 9 of the global
Convention on Biological Diversity, whereby signatory
parties (nations) are committed to controlling or eradicating
priority invasive alien species by 2020 (Convention on
Biological Diversity, 2011), and is being used for the
Biodiversity Indicator Partnership accordingly <https://
www.bipindicators.net/indicators/trends-in-invasive-alienspecies-vertebrate-eradications>.
The collation of more than 1,000 different eradication
events inevitably encounters challenges. Reconciling
the area (ha) and location (latitude and longitude) of
small islands targeted for invasive species eradications
against global data layers, has presented challenges to
maintaining accuracy. In general, relying on one dataset
(the GID) provides consistency, and seeking to validate
those locations with satellite or other imagery should
improve rigour. For rodent eradications, there is the risk
that some projects classified as successful but reinvaded
were in fact misdiagnosed operational failures. The time
Table 2 Eradication status definitions.
Eradication status
Successful
Failed
To be confirmed
In progress
Planned
Incomplete
Trial or research only
Unknown
Unknown pre-status
630
Definition
The operation to eradicate the invasive was successful and confirmed
The eradication operation was completed (there is an end date) yet it failed to remove
the entire invasive population. Operational failure (as opposed to reinvasion). For rodent
eradications, if there was uncertainty about why the invasive population remained (failure
versus reinvasion), we assumed operational failure and classified data quality as ‘poor’
The eradication effort is complete, but the operation has yet to be "confirmed" as successful
or failed. This stage is typical for rodent eradication operations, with confirmation
monitoring occurring 1–2 years after the eradication operation has ended
Eradication operation is currently in progress at time of reporting
Eradication is being planned for the island at time of reporting. End year will be unknown
accordingly
An eradication was started, but not followed through to completion
The eradication was undertaken for trial or research purposes and the goal was to gain new
knowledge, not eradicate invasive species
Information does not allow allocation into one of the other mutually exclusive categories and
an expert cannot do the same (e.g. unclear if an eradication took place or if the species "died
out" naturally). Selection of this category will often be aligned with poor data quality
Eradication was undertaken but the status of the invasive species was unclear beforehand.
Typically undertaken for precautionary measures for rodent eradications
Holmes, et al.: Tracking invasive species eradications
elapsed between the operation and invasion, and robust
genetic analyses can confirm this classification (Russell, et
al., 2010), but these may not be available on all projects,
particularly islands that are not visited regularly, or for
older projects where genetic tools were not available
(Holmes, et al., 2015). In general, data in the DIISE rely
on the eradication status provided by the practitioner.
Including successful but reinvaded in data summaries may
overestimate the success rate, but this can be mitigated by
excluding those events. Similarly, outcomes of multiple
adjacent islands that may function as a single eradication
unit may skew success rates if they are treated as separate
events. This can be accommodated for by selecting one
representative island in that unit (e.g. see Holmes, et al.,
2015).
Opportunities exist to improve and expand the schema
and content of the DIISE. The DIISE is currently organised
by island unit but currently does not link events based on
operation (islands treated concurrently) or eradication unit
(Abdelkrim, et al., 2005), whereby an invasive animal
population may move freely between adjacent islands
based on swimming or flying ability (‘natural’ reinvasion
risk – Harris, et al., 2012). Most (98%) of the target
animals in the DIISE are invasive mammals. A handful of
bird eradications are recorded although they may require
a different spatial organisational unit and consideration,
particularly where entire archipelagos are invaded, and
birds can move freely between islands. Some areas of the
world may be under-represented in the database, including
Small Island Developing States (Russell, et al., 2017)
where resources to report outcomes may be scarcer, and
the known lack of expert contacts in SE Asia, possibly
reflecting a language barrier. More deliberate attempts to
track these data may expand the dataset.
The DIISE dataset is freely available online, and
requests for datasets to answer specific questions are
responded to as best possible. There is a genuine resource
cost to maintaining this data accessibility and a more
significant investment required to undertake a major update.
Thus, ensuring financial investment is key to maintaining
this service. Despite the best of intentions, errors and
omissions may occur in the dataset and, depending on
the significance of the end goal users require the data
for, additional validation of events in the DIISE may be
warranted (e.g. Holmes, et al., 2015). A commonly soughtafter use is summary statistics, for which we encourage
those to check existing literature as they may already exist
from sufficiently recent summaries (e.g. Russell & Holmes,
2015). For those seeking novel statistics not reported
elsewhere, using only good or satisfactory data quality
events is encouraged, as is being conscious of eradication
type (whole island or restricted range). Events generating
failure rates for rodents may need to consider that some
reinvasion events may be misdiagnosed failures, and for
events targeting species that have agricultural or domestic
analogues (ungulates, dogs, cats), consideration may need
to be given to whether domestic or feral populations are
included. Using the data requires agreeing with a termsof-use and checking with database managers is strongly
encouraged to guide appropriate use of data.
Conservation databases provide a key role for informing
decision making and assessing trends (e.g. the IUCN Red
List) (Joppa, et al., 2016). At a project scale, data from the
DIISE regularly features within feasibility assessments,
by providing a comparison of proposed activities against
past efforts. Data from the DIISE dataset has been used
as a baseline to inform other conservation-based studies.
Holmes, et al. (2015) and Russell & Holmes (2015) used
the data to evaluate trends evident in why rodent eradication
failed at higher rates in the tropics although note that
predicting failure from operational covariates is not a
panacea. Russell, et al. (2017) evaluated trends in where
eradications occur, or may be under-reported, amongst
different countries of the world. Importantly, recent efforts
include Jones, et al. (2016) and Brooke, et al. (2017),
who used validated DIISE data to explore biodiversity
conservation outcomes, and seabird demographic response
to invasive mammal eradications, respectively. Jones, et
al. (2016) reported 596 populations of 236 native species
on 181 islands benefiting after eradications. These types
of studies are immensely valuable for measuring the true
‘effect’ (Kapos, et al., 2010) of eradication of invasive
species on islands as a management action.
ACKNOWLEDGEMENTS
The Database of Islands and Invasive Species
Eradications has benefitted greatly from many significant
contributions of many island restoration practitioners
around the world. Thank you kindly for the support of
keeping this database online and up to date. Numerous
volunteers have provided time to review data provided and
enter it into the database. The Center for Integrated Spatial
Research at the University of California of Santa Cruz
developed the online database portal and maintains the
website. Development of the DIISE was supported by the
David and Lucile Packard Foundation, and the National
Fish and Wildlife Foundation.
REFERENCES
Abdelkrim, J., Pascal, M., Calmet, C. and Samadi, S. (2005). ‘Importance
of assessing population genetic structure before eradication of invasive
species: Examples from insular Norway rat populations’. Conservation
Biology 19(5): 1509–1518.
Aguirre-Muñoz, A., Croll, D.A., Donlan, C.J., Henry, R.W., III,
Hermosillo, M.A., Howald, G.R., Keitt, B.S., Luna-Mendoza, L.,
Rodriguez-Malagón, M., Salas-Flores, L.M., Samaniego-Herrera, A.,
Sanchez-Pacheco, J.A., Sheppard, J., Tershy, B.R., Toro-Benito, J.,
Wolf, S. and Wood, B. (2008). ‘High-impact conservation: Invasive
mammal eradications from the islands of western Mexico’. Ambio
37(2): 101–107.
Aguirre-Muñoz, A., Samaniego-Herrera, A., Luna-Mendoza, L., OrtizAlcaraz, A., Rodriguez-Malagón, M., Méndez-Sánchez, F., FélixLizárraga, M., Hernández-Montoya, J.C., González-Gómez, R., TorresGarcia, F., Barredo-Barberena, J.M. and Loatofski-Robles, M. (2011).
‘Island restoration in Mexico: ecological outcomes after systematic
eradications of invasive mammals’. In: C.R. Veitch, M.N. Clout and
D.R. Towns (eds.) Island invasives: eradication and management, pp.
250–258. Occassional Paper SSC no. 42. Gland, Switzerland: IUCN
and Auckland, New Zealand: CBB.
Barun, A., Hanson, C.C., Campbell, K.J. and Simberloff, D. (2011). ‘A
review of small Indian mongoose management and eradications on
islands’. In: C.R. Veitch, M.N. Clout and D.R. Towns (eds.) Island
invasives: eradication and management, pp. 17–25. Occassional Paper
SSC no. 42. Gland, Switzerland: IUCN and Auckland, New Zealand:
CBB.
Beaver, K. and Mougal, J. (2009). Review of IAS control and eradication
programmes in Seychelles. With contributions from the Island
Conservation Society on control and eradication programmes carried
out under the ICS-FFEM Restoration of Island Ecosystems programme
– provided by Gerard Rocamora. Prepared by the Plant Conservation
Action
group.
<http://www.pcusey.sc/index.php/component/
remository/BS/Project-Technical-Reports/Outcome-3/TechReports3/
Review-of-IAS-Control-and-Eradication-Programmes-in-Seychelles2009/?Itemid=>
Brooke, M.d.L., Bonnaud, E., Dilley, B.J., Flint, E.N., Holmes, N.D.,
Jones, H.P., Provost, P., Rocamora, G., Ryan, P.G., Surman, C. and
Buxton, R.T. (2017). ‘Seabird population changes following mammal
eradications on islands’. Animal Conservation 21: 3–12.
CDFW (California Department of Fish and Wildlife) (2015). California
State Wildlife Action Plan, 2015 Update: A Conservation Legacy
for Californians. A.G. Gonzales and J. Hoshi (eds.). Prepared with
assistance from Ascent Environmental, Inc., Sacramento, California.
<https://www.wildlife.ca.gov/SWAP>.
Campbell, K. and Donlan, C.J. (2005). ‘Feral goat eradications on
islands’. Conservation Biology 19(5): 1362–1374.
631
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
Campbell, K.J., Harper, G., Hansen, C.C., Algar, D.S., Keitt, B.S. and
Robinson, S.A. (2011). ‘Review of feral cat eradications on islands’.
In: C.R. Veitch, M.N. Clout and D.R. Towns (eds.) Island invasives:
eradication and management. pp. 37–46. Occassional Paper SSC no.
42. Gland, Switzerland: IUCN and Auckland, New Zealand: CBB.
Capizzi, D., Baccetti, N. and Sposimo, P. (2016). ‘Fifteen Years of Rat
Eradication on Italian Islands’. In: F.M. Angelici (ed.) Problematic
Wildlife: A Cross-Disciplinary Approach. pp. 205–227. Cham: Springer
International Publishing.
Carrion, V., Donlan, C.J., Campbell, K.J., Lavoie, C. and Cruz, F. (2011).
‘Archipelago-wide island restoration in the Galápagos Islands: Reducing
costs of invasive mammal eradication programs and reinvasion risk’.
PLOS ONE 6(5): e18835.
Clout, M.N. and Russell, J.C. (2006). ‘The Eradication of Mammals from
New Zealand Islands ‘. In: F. Koike, M.N. Clout, M. Kawakichi, M. De
Poorter and K. Iwatsuki (eds.) Assessment and Control of Biological
Invasion Risks, pp. 127–141. Kyoto, Japan and Gland, Switzerland:
Shoukadoh Book Sellers and IUCN.
Convention on Biological Diversity (2011). Conference of the Parties
Decision X/2: Strategic Plan for Biodiversity 2011–2020. <www.cbd.
int/decision/cop/?id=12268>.
Depraetere, C. (2007). Global Islands Database (GID): A Technical Note
on a Global Dataset of Islands. In collaboration with Arthur Dahl,
February 2007, Global Islands Network.
Genovesi, P. (2005). ‘Eradications of invasive alien species in Europe: a
review’. Biological Invasions 7(1): 127–133.
Genovesi, P. and Carnevali, L. (2011). ‘Invasive alien species on
European islands: eradications and priorities for future work’. In: C.R.
Veitch, M.N. Clout and D.R. Towns (eds.) Island invasives: eradication
and management, pp. 56–62. Occassional Paper SSC no. 42. Gland,
Switzerland: IUCN and Auckland, New Zealand: CBB.
Glen, A., Atkinson, R., Campbell, K., Hagen, E., Holmes, N., Keitt, B.,
Parkes, J., Saunders, A., Sawyer, J. and Torres, H. (2013). ‘Eradicating
multiple invasive species on inhabited islands: the next big step in
island restoration?’. Biological Invasions 15(12): 2589–2603.
Gregory, S.D., Henderson, W., Smee, E. and Cassey, P. (2014).
Eradications of Vertebrate Pests in Australia: A Review and Guidelines
for Future Best Practice. PestSmart Toolkit publication, Canberra,
Australia: Invasive Animals Cooperative Research Centre.
Harper, G. and Carrion, V. (2011). ‘Introduced rodents in the Galápagos:
colonisation, removal and the future’. In: C.R. Veitch, M.N. Clout and
D.R. Towns (eds.) Island invasives: eradication and management, pp.
63–66. Occassional Paper SSC no. 42. Gland, Switzerland: IUCN and
Auckland, New Zealand: CBB.
Harris, D.B., Gregory, S.D., Bull, L.S. and Courchamp, F. (2012). ‘Island
prioritization for invasive rodent eradications with an emphasis on
reinvasion risk’. Biological Invasions 14(6): 1251–1263.
Hess, S.C. and Jacobi, J.D. (2011). ‘The history of mammal eradications in
Hawai’i and the United States Associated Islands of the Central Pacific’.
In: C.R. Veitch, M.N. Clout and D.R. Towns (eds.) Island invasives:
eradication and management, pp. 67–73. Occassional Paper SSC no.
42. Gland, Switzerland: IUCN and Auckland, New Zealand: CBB.
Hoffmann, B.D., Luque, G.M., Bellard, C., Holmes, N.D. and Donlan,
C.J. (2016). ‘Improving invasive ant eradication as a conservation tool:
A review’. Biological Conservation 198: 37–49.
Holmes, N.D., Griffiths, R., Pott, M., Alifano, A., Will, D., Wegmann, A.S.
and Russell, J.C. (2015). ‘Factors associated with rodent eradication
failure’. Biological Conservation 185: 8–16.
Howald, G., Donlan, C.J., Galvan, J.P., Russell, J.C., Parkes, J.,
Samaniego, A., Wang, Y., Veitch, D., Genovesi, P., Pascal, M., Saunders,
A. and Tershy, B. (2007). ‘Invasive rodent eradication on islands’.
Conservation Biology 21(5): 1258–1268.
Jones, H.P., Holmes, N.D., Butchart, S.H.M., Tershy, B.R., Kappes,
P.J., Corkery, I., Aguirre-Muñoz, A., Armstrong, D.P., Bonnaud, E.,
Burbidge, A.A., Campbell, K., Courchamp, F., Cowan, P.E., Cuthbert,
R.J., Ebbert, S., Genovesi, P., Howald, G.R., Keitt, B.S., Kress, S.W.,
Miskelly, C.M., Oppel, S., Poncet, S., Rauzon, M.J., Rocamora, G.,
Russell, J.C., Samaniego-Herrera, A., Seddon, P.J., Spatz, D.R., Towns,
D.R. and Croll, D.A. (2016). ‘Invasive mammal eradication on islands
results in substantial conservation gains’. Proceedings of the National
Academy of Sciences 113: 4033–4038.
Joppa, L.N., O’Connor, B., Visconti, P., Smith, C., Geldmann, J.,
Hoffmann, M., Watson, J.E.M., Butchart, S.H.M., Virah-Sawmy, M.,
Halpern, B.S., Ahmed, S.E., Balmford, A., Sutherland, W.J., Harfoot,
M., Hilton-Taylor, C., Foden, W., Minin, E.D., Pagad, S., Genovesi,
P., Hutton, J. and Burgess, N.D. (2016). ‘Filling in biodiversity threat
gaps’. Science 352(6284): 416–418.
632
Kapos, V., Manica, A., Aveling, R., Bubb, P., Carey, P., Entwistle,
A., Hopkins, J., Mulliken, T., Safford, R., Stattersfield, A., Walpole,
M.J. and Balmford, A. (2010). ‘Defining and Measuring Success in
Conservation’. In: N. Leader‐Williams, W.M. Adams and R.J. Smith
(eds.) Trade-Offs in Conservation. pp. 73–93. Oxford, UK: WileyBlackwell.
Keitt, B., Campbell, K., Saunders, A., Clout, M.N., Wang, Y., Heinz,
R., Newton, K. and Tershy, B. (2011). ‘The global islands invasive
vertebrate eradication database: a tool to improve and facilitate
restoration of island ecosystems’. In: C.R. Veitch, M.N. Clout and
D.R. Towns (eds.) Island invasives: eradication and management, pp.
74–77. Occassional Paper SSC no. 42. Gland, Switzerland: IUCN and
Auckland, New Zealand: CBB.
Lorvelec, O. and Pascal, M. (2005). ‘French attempts to eradicate
non-indigenous mammals and their consequences for native biota’.
Biological Invasions 7(1): 135–140.
MacKay, J.W.B., Russell, J.C. and Murphy, E.C. (2007). ‘Eradicating
House Mice from Islands: Successes, Failures and the Way Forward’.
In: G.W. Witmer, W.C. Pitt and K.A. Fagerstone (eds.) Managing
Vertebrate Invasive Species. Proceedings of an International
Symposium, pp. 294–304. Fort Collins, Colorado: USDA/APHIS/WS,
National Wildlife Research Center.
McChesney, G.J. and Tershy, B.R. (1998). ‘History and status of introduced
mammals and impacts to breeding seabirds on the California Channel
and Northwestern Baja California Islands’. Colonial Waterbirds 21:
335–347.
McGeoch, M.A., Butchart, S.H.M., Spear, D., Marais, E., Kleynhans, E.J.,
Symes, A., Chanson, J. and Hoffmann, M. (2010). ‘Global indicators of
biological invasion: species numbers, biodiversity impact and policy
responses’. Diversity and Distributions 16(1): 95–108.
Niemeijer, D. and de Groot, R.S. (2008). ‘A conceptual framework for
selecting environmental indicator sets’. Ecological Indicators 8(1):
14–25.
Nogales, M., Martin, A., Tershy, B.R., Donlan, C.J., Veitch, D., Puerta,
N., Wood, B. and Alonso, J. (2004). ‘A review of feral cat eradication
on islands’. Conservation Biology 18(2): 310–319.
Oppel, S., Beaven, B.M., Bolton, M., Vickery, J. and Bodey, T.W. (2011).
‘Eradication of invasive mammals on islands inhabited by humans and
domestic animals’. Conservation Biology 25(2): 232–240.
Phillips, R.B., Wiedenfeld, D.A. and Snell, H.L. (2012). ‘Current status of
alien vertebrates in the Galápagos Islands: invasion history, distribution,
and potential impacts’. Biological Invasions 14(2): 461–480.
Russell, J.C., Miller, S.D., Harper, G.A., MacInnes, H.E., Wylie, M.J.
and Fewster, R.M. (2010). ‘Survivors or reinvaders? Using genetic
assignment to identify invasive pests following eradication’. Biological
Invasions 12(6): 1747–1757.
Russell, J.C. and Holmes, N.D. (2015). ‘Tropical island conservation: Rat
eradication for species recovery’. Biological Conservation 185: 1–7.
Russell, J.C., Cole, N.C., Zuël, N. and Rocamora, G. (2016). ‘Introduced
mammals on Western Indian Ocean islands’. Global Ecology and
Conservation 6: 132–144.
Russell, J.C., Meyer, J.-Y., Holmes, N.D. and Pagad, S. (2017). ‘Invasive
alien species on islands: impacts, distribution, interactions and
management’. Environmental Conservation 44(4): 359–370.
Tobin, P.C., Kean, J.M., Suckling, D.M., McCullough, D.G., Herms,
D.A. and Stringer, L.D. (2014). ‘Determinants of successful arthropod
eradication programs’. Biological Invasions 16(2): 401–414.
Veitch, C.R., Clout, M.N. (eds.) (2002). Turning the tide: the eradication
of invasive species. Occasional Paper SSC no. 28. IUCN SSC Invasive
Species Specialist Group. IUCN, Gland, Switzerland and Cambridge,
UK.
Veitch, C.R., Clout, M.N. and Towns, D.R. (2011). (eds.) Island invasives:
eradication and management. Gland, Switzerland: IUCN and Auckland,
New Zealand: CBB.
Witmer, G.W. and Fuller, P.L. (2011). ‘Vertebrate species introductions in
the United States and its territories’. Current Zoology 57(5): 559–567.
B. Keitt, N. Holmes, E. Hagen, G. Howald and K. Poiani
Keitt, B.; N. Holmes, E. Hagen, G. Howald and K. Poiani. Going to scale: reviewing where we’ve been and where we need to go in invasive vertebrate eradications
Going to scale: reviewing where we’ve been and where we need to go
in invasive vertebrate eradications
B. Keitt1,2, N. Holmes1, E. Hagen1, G. Howald1 and K. Poiani1
Island Conservation, 2100 Delaware Ave Suite 1, Santa Cruz CA 950960. <bkeitt@abcbirds.org>. 2Current affiliation:
American Bird Conservancy 4249 Loudoun Ave, The Plains, VA 20198.
1
Abstract We are on the edge of the sixth mass extinction on Earth. Islands represent ca. 5% of the earth’s land area yet
are home to 61% of extinctions in the past 500 years, and currently support 39% of critically endangered species. Invasive
species are a leading cause of extinction and endangerment on islands. Invasive vertebrates, particularly mammals, are
among some of the most damaging invasive species on islands. Eradicating invasive mammals is an increasingly utilised
conservation tool. Nevertheless, conservation intervention needs greatly outstrip the island restoration community’s
capacity. There are thousands of islands where invasive vertebrates are driving species toward extinction. So, how can the
effort be matched to the scale of the problem? One approach is to improve outreach and communications to increase the
resources available for projects. There are great stories; but these need to be told compellingly and repeatedly. Increasing
social acceptance and support for invasive species eradications will reduce project costs associated with stakeholder
engagement. Broadening the funding base can be accomplished by building stronger cost benefit valuations as well
as engaging funders of climate change, marine conservation, human wellbeing, and food security. Furthermore, it is
important to build upon existing partnerships to create or grow coalitions that can access these resources as part of
broader, holistic efforts to address multiple conservation issues.
Keywords: communications, eradication, funding, invasive species, stakeholder engagement
INTRODUCTION
Multiple lines of evidence demonstrate that we are
facing a significant global extinction crisis through the
loss of biodiversity (Dirzo, 2003; Barnosky, et al., 2011).
At a global scale, the response to this crisis includes the
2011–2020 strategic plan for biodiversity, highlighted
in 20 targets to reduce pressures on the environment
and to curb biodiversity loss (CBD, 2011). Islands are a
logical place to focus conservation efforts because they
offer a disproportionately higher rate of biodiversity and
threatened species per unit area. Islands represent only ca.
5% of the earth’s land area yet support ca. 39% of critically
endangered species on the IUCN Red List (Tershy, et al.,
2015), and an endemic richness of plants and vertebrates
that is 8–9 times that on mainlands (Kier, et al., 2009).
Invasive alien species have been implicated as a leading
cause of extinctions and endangerment for native plants
and animals on islands (Tershy, et al., 2015). In particular,
invasive mammals pose a significant risk (Doherty, et
al., 2016). The development of tools and techniques
to completely remove invasive mammal populations
from islands has been a valuable intervention strategy
for island managers to overcome this threat (Veitch &
Clout, 2002; Veitch, et al., 2011). To date there have
been more than 1,200 vertebrate eradication attempts on
more than 700 islands with an 85% success rate (DIISE,
2014), and the pace and scale of eradications on islands is
increasing (Simberloff, et al., 2018). Following successful
eradications, demonstrable biodiversity conservation
gains have accrued. A recent literature review found
596 populations of 236 native insular species benefited
from 251 invasive mammal eradications on 181 islands
(Jones, et al., 2016). Benefits included resident population
recovery, recolonisation and unassisted colonisation,
plus the enabling of reintroductions and conservation
introductions. Similarly, Brooke, et al. (2017) investigated
population growth rates in seabirds following invasive
mammal eradications on islands and found a median
population growth rate of 1.119 based on 181 populations
of 69 seabird species.
NOTABLE ADVANCES AND INNOVATIONS
Several key innovations were critical to increasing
the rate at which eradications of invasive mammals on
islands have occurred. For rodents, New Zealand based
programmes that researched the effectiveness of bait station
approaches led to a series of successful implementations on
small islands (Howald, et al., 2007). The advancement of
aerial application techniques, including the use of satellite
navigation systems, enabled efforts on larger islands
and increased the number of islands treated, including
>11,000 ha Campbell Island (Towns & Broome, 2003).
These techniques have been exported internationally,
with Macquarie Island at >12,000 ha recently declared
successful, and implementation units recently treated
within the South Georgia eradication reaching almost
30,000 ha. Likewise, for invasive ungulates, the advent of
aerial hunting, extensive near real-time data management
combined with mapping technology to coordinate large
teams and different eradication methods, and the use of
Judas goats enabled similar increases in number and size
of islands treated (Campbell & Donlan, 2005) whereas
aerial application, toxicant development and remote trap
monitoring allowed continued increases in island size,
efficiency and efficacy to be obtained on cat eradications
(Campbell, et al., 2011), including the currently on-going
treatment of ca. 65,000 ha Dirk Harthog Island in Australia.
The cumulative impact of numerous existing and
on-going innovations is expected to increase the scope
and scale of eradications on islands. Models to confirm
eradication success (Ramsey, et al., 2009; Ramsey, et
al., 2011) provide significant opportunities to reduce
costs, particularly for large projects using hunting and/
or trapping techniques, by increasing the efficiency of
determining when a project is complete. Increased use of
these tools, and associated real time, digital data collection
and analysis tools, is recommended to increase efficacy,
reduce costs and provide more information to enable post
project review and analysis for future improvements (Will,
et al., 2015).
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 633–636. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
633
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
Efforts are ongoing to reduce reliance on second
generation anti-coagulants for rodent eradications,
whose efficacy comes with a trade-off of greater risk to
nontarget species (Howald, et al., 2007). These include
expanding the use of first generation anticoagulants that
pose less risk to non-target species (Poncet, et al., 2011),
investigating alternative compounds, such as Norbamide,
and investigating new bait recipes that could increase
efficacy, such as crab deterrents (Campbell, et al., 2015).
Self-resetting traps, a relatively new tool, that have been
deployed successfully for eradication on small islands in
Puerto Rico and New Zealand, present another alternative
on small islands where rodenticide use may not be possible
(Carter, et al., 2016). These self-resetting traps present
significant potential for biosecurity management and can
provide long term protection where reinvasion risk from
swimming rodents is high.
New strategies have been developed to overcome the
higher failure rate in rodent eradications in the tropics.
After a series of high profile rodent eradication failures on
tropical islands, a workshop of practitioners, The Tropical
Rodent Eradication Review, was convened to evaluate
reasons for these failures and develop recommendations
to increase success rates in the future. These guidelines
were published in 2015 (Keitt, et al., 2015) and several
projects implemented since have followed the spirit of
these guidelines. It remains to be seen whether efficacy
rates will increase as a result, though the second attempt
on Desecheo, which followed the guidelines, was declared
successful (Will, et al., 2019).
Another promising approach is genetic tools that can
lead to eradication of rodent populations (Campbell, et
al., 2015, Campbell et al., 2019). Genetically modifying
rodents to produce sterile offspring or only males and
using gene drives to push for near 100% inheritance of this
trait, could lead to eradication at large scales, including on
inhabited islands where eradication is not currently feasible.
This technology is in the early stages of development for
house mice and it is unlikely that it would be available
for field trials sooner than a decade from now; longer for
commensal rat species. However, there has been significant
concern raised about the safety and ethics of pursuing this
line of conservation, particularly around the potential for
a gene drive to run through an entire species and lead to
extinction (National Academies of Sciences, Engineering,
and Medicine, 2016). If this technology can be proven safe
and gain the appropriate social and political approvals, it
could have wide ranging impact on the conservation of
large inhabited islands while also providing significant
benefit to humans through reduced disease transmission
and reduction in agricultural loss.
LOOKING TO THE FUTURE
These efforts have made significant contributions
to global progress in protecting biodiversity. However,
there remains much to be done, and the conservation
need is high. Jones, et al (2016) predicted that 107 highly
threatened insular terrestrial vertebrates (229 populations)
have benefitted in some way from invasive mammal
eradications on islands, however this represents just 12%
of all 860 highly threatened terrestrial vertebrates occurring
on islands. The picture is brighter for seabirds, where 47%
of critically endangered and 74% of endangered species
were predicted to have benefitted from invasive mammal
eradications to date. Considering the future, McCreless, et
al. (2016) found that efforts to control or eradicate relevant
invasive species could prevent 41–75% of future predicted
extirpations of populations of threatened vertebrate
species. Almost half of these extirpations reflect species
with a single population (endemic) and thus extirpation is
the same as extinction.
634
The number of islands targeted for eradication are few
compared to the number of islands worldwide. Invasive
rodents are widespread, with estimates of 80% of the
world’s island groups being invaded (Atkinson, 1985).
Recent estimates suggest there are > 400,000 islands in
the world > 10 ha (UNEP-WCMC, 2013) yet only ca. 450
have been the focus of rodent eradications (DIISE, 2014).
Thus, the need to increase the scope and scale of efforts to
eradicate invasive vertebrates is known (Philips, 2010). A
considerable number of these invaded islands are outside
the boundaries of what is considered feasible for invasive
species eradications today, and innovative approaches
will need to be established to realise these opportunities
(Campbell, et al., 2015). These include use of some of the
innovations mentioned above as well as ones yet to be
envisioned. Two additional focal areas for development
include the social acceptability of these projects and
increased funding to implement projects.
CONSERVING SPECIES ON INHABITED
ISLANDS – UNDERSTANDING THE SOCIAL
CONTEXT
Due to the overlap of human settlements and biodiversity
there has been an increasing interest in eradication projects
on inhabited islands (Oppel, et al., 2011). Simberloff, et al.
(2018) reported 194 eradication attempts on 94 inhabited
islands, and a “sharp uptick” in numbers of attempts on
inhabited islands for all species except rodents in 1960 and
for rodents in 1990. Notable projects under consideration
include Lord Howe Island, Robinson Crusoe, Great
Barrier Island and Floreana Island. Glen, et al. (2013)
make the case that inhabited islands often support a suite
of invasive species and thus restoration efforts can require
multi-species eradications that must take into account the
ecological impacts of improper sequencing of removals
and potential negative consequences of allowing some
invaders to remain. Combining this challenge with that
of gaining social license to achieve eradication, inhabited
islands have been hailed as a next great challenge for
conservation (Glen, et al., 2013).
It is likely that most land managers attempting to
implement invasive vertebrate eradications on islands
would prefer to do so in the relatively accommodating
social environment of New Zealand, where signs in
tourism shops proudly report on their efforts to control
invasive species. Understanding the underlying reasons
for social acceptance, or lack thereof, for eradication
projects, is an important aspect of planning an appropriate
process to achieve stakeholder support and approval for
a project. As an example, the New Zealand conservation
movement arguably began with efforts to protect its
endemic birds, including the national bird, the kiwi
(genus Apteryx) (Stoltzenberg, 2011). Given that invasive
species currently are their greatest threat, it is natural
that control and eradication enjoy broad support within
the country. Contrast this with the United States, where
some suggest the environmental movement can be traced
back to Rachel Carson’s Silent Spring (1962), which
highlighted the imminent extinction of the US national
symbol, the bald eagle, (Haliaeetus leucocephalus) from
pesticide exposure. It is exactly these kinds of underlying
human conditions that can impact attitudes about invasive
species and the tools to control and eradicate them. Island
restoration projects have typically applied significant rigor
to the biological science necessary to understand invasive
mammal eradication projects. As projects face more
complex human dimensions, it will be necessary to apply
the same rigor to the social sciences in order to achieve the
necessary project support to proceed.
Keitt, et al.: Reviewing invasive vertebrate eradications
INCREASING FINANCIAL AND STAKEHOLDER
SUPPORT
values, programme goals, respective responsibilities and
definitions of success.
To expand funding opportunities for island restoration
projects it is important to expand project justifications
beyond biodiversity conservation to include human
health and livelihoods, and ecosystem services. This will
require new research to document and communicate the
non-biodiversity impacts of these projects. For example,
the Lord Howe Island rat eradication project underwent a
comprehensive Cost Benefit Analysis that demonstrated
there would be a benefit cost ratio of 17.0, i.e. 17 dollars
in benefits for every dollar spent on the project (Gillespie,
2016). A similar approach was completed for the Cabritos
Island donkey eradication in the Dominican Republic (Rijo,
2014). This analysis showed a benefit cost ratio of between
2.0 and 4.2 depending on the methods used to remove all
of the donkeys and resulting cost of the work. Additional
efforts to highlight the value of vertebrate eradications
on islands to humans, including human health (de Wit,
et al., 2017), ecosystem services (Peh, et al., 2014), and
agriculture will be key to securing the necessary support,
both financial and stakeholder, to meet the challenge.
Perhaps one of the greatest opportunities to grow
support for island conservation work in this increasingly
connected and wired planet is through communication and
outreach. Effective communication requires sharing the
right information with the right audience at the right time.
Story-telling is a communication approach that can make
difficult to understand ideas, such as the need to kill nonnative species to conserve native ones, more accessible.
Having island residents and stakeholders tell their stories
or presenting a project from the viewpoint of the native
species that will benefit, can resonate far more than
statistics and summaries of what has happened somewhere
else. The Goodman Center (<www.thegoodmancenter.
com>) is a resource that can help train how to develop and
tell compelling stories. For island restoration, the audiences
are varied – funders, stakeholders, island communities and
practitioners. This requires creating story arcs that reflect
the values of key decision-makers and involve rigorous and
defensible research to create story content. The recovery
associated with removing invasive species from islands
is often exceptional, providing compelling and dramatic
messages that can be shared to generate interest in projects.
Investing in the monitoring to document these stories is
often under-valued, yet the link to funding future projects
is clear. The platform for telling stories and reaching some
audiences is evolving quickly alongside technology, thus
these social media platforms require constant innovation
and novel approaches to reach audiences. Conversely,
many island communities communicate the same way they
did decades or even a century ago, with shared experiences
and face-to-face time as the key medium. Effective and
thoughtful planning of communications will continue to
evolve as necessary components of island restoration.
Making a strong link between island restoration and
marine conservation is important for maximising available
resources. Islands serve an important function in marine
ecosystems (Gove, et al., 2016), including providing key
breeding habitat for species that are dependent on marine
resources. Most seabirds, sea turtles and marine mammals
are dependent on islands to reproduce yet are key members
of marine ecosystems. Making this case to marine funders
and incorporating goals to protect and maintain populations
of top level native predators in the management plans of
these reserves is a good place to start.
Climate change is projected to have a significant impact
on islands and island species and there are significant
global financial resources available for addressing climate
change impacts. Tershy, et al. (2015) argue that some of
the same attributes that make island species vulnerable to
invasive species, primarily smaller ranges and population
sizes and less genetic diversity, also make them vulnerable
to climate change. For many island ecosystems, invasive
mammal eradications, in combination with other restoration
actions, can increase resilience to projected climate change
impacts, and provide refugia for species whose habitat is
projected to be lost. However, proposed island restorations
on low elevation islands should consider future sea level
rise projections (Courchamp, et al., 2014) and include this
in the project cost/benefit analysis.
Partnerships are not new to conservation, yet as island
restoration projects expand in size and scope, diverse
partnerships become more important to their success. Nongovernmental organisations and governments working
collaboratively together are becoming more commonplace.
For example, the United States Fish and Wildlife
Service has established a national level Memorandum
of Understanding with other government agencies and
US based NGOs to facilitate invasive species work and
move from a project focus to a more programmatic one.
Collaboration between NGOs internationally is also
becoming more commonplace in the implementation
of eradication projects. An example is the partnership
between Island Conservation and Birdlife International on
the multi-island, multi-species eradication in the Acteon
and Gambier archipelagos, a project led by the local
Tahitian NGO, and Birdlife Partner, SOP Manu. There are
opportunities to expand these types of governmental and
non-governmental partnerships, to enhance the capacity
for conservation actions worldwide. Much is written about
how to create successful partnerships, and common tenets
to these types of partnerships are working to clarify shared
CONCLUSION
There are few conservation approaches that can match
the return on investment of invasive mammal eradications
on islands. As the earth continues to lose biodiversity at a
rapid pace, with islands disproportionately affected, it is
urgent to increase the rate at which islands are restored.
Innovation has played a key role in past increases in
eradication efficacy and efficiency (Keitt, et al., 2011) and
new innovations are primed to do the same (Campbell,
et al., 2015). However, these innovations must expand
beyond the technical aspect of how to eradicate invasives
and include ways to increase funding and stakeholder
engagement and support. With greater buy in for island
restoration projects they will become easier to implement.
REFERENCES
Atkinson, I.A.E. (1985). ‘The Spread of Commensal Species of Rattus to
Oceanic Islands and their Effects on Island Avifaunas’. In: P.J. Moors (ed.)
Conservation of Island Birds, pp. 35–81. ICBP Technical Publication No.
3. Cambridge, UK: ICBP.
Barnosky, A.D., Matzke, N., Tomiya, S., Wogan, G.O.U., Swartz, B.,
Quental, T.B., Marshall, C., McGuire, J.L., Lindsey, E.L., Maguire, K.C.,
Mersey, B. and Ferrer, E.A. (2011). ‘Has the Earth’s sixth mass extinction
already arrived?’ Nature 471: 51–57.
Brooke, M.de L., Bonnaud, E., Dilley, B.J., Flint, B., Holmes, N.D.,
Jones, H.P., Provost, P., Rocamora, G., Ryan, P.G., Surman, C. and
Buxton, R.T. (2017). ‘Seabird population changes following mammal
eradications on islands’. Animal Conservation 10.1111/acv.12344.
Campbell, C. and Donlan, C.J. (2005). ‘Feral goat eradications on
islands’. Conservation Biology 19: 1362–1374.
Campbell, K.J.; Harper, G.; Algar, D.; Hanson, C.; Keitt, B.S. and
Robinson, S. 2011. ‘Review of feral cat eradications’. In: Veitch, C.R.;
Clout, M.N. and Towns, D.R. (eds.) Island invasives: eradication
and management, pp. 37–46. Occassional Paper SSC no. 42. Gland,
Switzerland: IUCN and Auckland, New Zealand: CBB.
635
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
Campbell, K.J., Beek, J., Eason, C.T., Glen, A.S., Godwin, J., Gould, F.,
Holmes, N.D., Howald, G.R., Madden, F.M., Ponder, J.P., Threadgill
D.W., Wegmann, A. and Baxter, G.S. (2015). ‘The next generation
of rodent eradications: Innovative technologies and tools to improve
species specificity and increase their feasibility on islands’. Biological
Conservation 185: 47–58.
Campbell, K.J., Saah, R.J., Brown, P.R., Godwin, J., Gould, F., Howald,
G.R., Piaggio, A., Thomas, P., Tompkins, D.M., Threadgill, D.,
Delborne, J., Kanavy, D.M., Kuikin, T., Leitschuh, C., Packard, H.,
Serr, M. and Shiels, A. (2019). ‘A potential new tool for the toolbox:
assessing gene drives for eradicating invasive rodent populations’. In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.)
Island invasives: scaling up to meet the challenge, pp. 6–14. Occasional
Paper SSC no. 62. Gland, Switzerland: IUCN.
Carson, R. (1962). Silent Spring. New York: Houghton Mifflin Company.
Carter, A., Barr, S., Bond, C., Paske, G., Peters, D. and van Dam, R.
(2016). ‘Controlling sympatric pest mammal populations in New
Zealand with self-resetting, toxicant-free traps: A promising tool for
invasive species management’. Biological Invasions 18: 1723–1736.
CBD. (Convention on Biological Diversity) (2011). ‘Conference of the
Parties Decision X/2: Strategic Plan for Biodiversity 2011–2020’.
<www.cbd.int/decision/cop/?id=12268>.
Courchamp, F., Hoffmann, B.D., Russell, J.C., Leclerc, C. and Bellard,
C. (2014). ‘Climate change, sea-level rise, and conservation: Keeping
island biodiversity afloat’. Trends in Ecology and Evolution 29: 127–
130.
de Wit, L.A., Croll, D.A., Tershy, B., Newton, K M., Spatz D.R., Holmes
N.D. and Kilpatrick, A.M. (2017). ‘Estimating burdens of neglected
tropical zoonotic diseases on islands with introduced mammals’. The
American Journal of Tropical Medicine and Hygiene 96: 749–757.
Doherty, T.S., Glen, A.S., Nimmo, D.G., Ritchie, E.G. and Dickman, C.R.
(2016). ‘Invasive predators and global biodiversity loss’. Proceedings of
the National Academy of Sciences 113: 11261–11265.
DIISE (2014). ‘The Database of Island Invasive Species Eradications,
Developed by Island Conservation, Coastal Conservation Action
Laboratory UCSC, IUCN SSC Invasive Species Specialist Group,
University of Auckland and Landcare Research New Zealand’. <http://
diise.islandconservation.org>. Accessed 14 September 2014.
Dirzo, R. and Raven, P.H. (2003). ‘Global state of biodiversity and loss’.
Annual Review of Environment and Resources 28: 137–167.
McCreless, E.E., Huff, D.D., Croll, D.A., Tershy, B.R., Spatz, D.R.,
Holmes, N.D., Butchart, S.H. M. and Wilcox, C. (2016). ‘Past and
estimated future impact of invasive alien mammals on insular threatened
vertebrate populations’. Nature Communications 7: 12488.
National Academies of Sciences, Engineering, and Medicine (2016). Gene
Drives on the Horizon: Advancing Science, Navigating Uncertainty, and
Aligning Research with Public Values. Washington, DC: The National
Academies Press. <https://doi.org/10.17226/23405>.
Oppel, S., Beaven, B.M., Bolton, M., Vickery, J. and Bodey, T.W. (2011).
‘Eradication of invasive mammals on islands inhabited by humans and
domestic animals’. Conservation Biology 25: 232–240.
Peh, K.H., Balmford, A., Birch, J., Brown, C., Butchart, S.M., Daley,
J., Dawson, J., Gray, G., Hughes, F.M.R., Mendes, S., Millett, J.,
Stattersfield, A.J., Thomas, D.H.L., Walpole, M. and Bradbury, R.
(2015). ‘Potential impact of invasive alien species on ecosystem
services provided by a tropical forested ecosystem: A case study from
Montserrat’. Biological Invasions 17: 461–475.
Phillips, R.A. (2010). ‘Eradications of invasive mammals from islands:
Why, where, how and what next?’ Emu 110: 1–7.
Poncet, S., Poncet, L., Poncet, D., Christie, D., Dockrill, C. and Brown,
D. (2011). ‘Introduced mammal eradications in the Falkland Islands and
South Georgia’. In: C.R. Veitch, M.N. Clout and D. Towns (eds.) Island
invasives: eradication and management, pp. 332–336. Occassional
Paper SSC no. 42. Gland, Switzerland: IUCN and Auckland, New
Zealand: CBB.
Ramsey, D.S.L., Parkes, J.P. and Morrison, S.A. (2009). ‘Quantifying
eradication success: The removal of feral pigs from Santa Cruz Island,
California’. Conservation Biology 23: 449–459.
Ramsey, D.S.L., Parkes, J.P., Will, D.J., Hanson, C.C. and Campbell, K.J.
(2011). ‘Quantifying the success of feral cat eradication, San Nicolas
Island, California’. New Zealand Journal of Ecology 35: 163–173.
Rijo, C. (2014). ‘Benefits to Biodiversity Outweigh Cost of Vertebrate
Eradication on Cabritos Island: Costs Benefits Analysis of Donkeys
Eradication’. In: A. Daigneault, P. Brown and N. Ramnanan (eds.)
Economic Impact of IAS in The Caribbean: Case Studies, pp. 30–37. Port
of Spain, Trinidad and Tobago: CABI.
Simberloff, D., Keitt, B., Will, D., Holmes, N., Pickett, Genovesi, P.
(2018). ‘Yes we can! Exciting progress and prospects for controlling
invasives on islands and beyond’. Western North American Naturalist
78: 942-958.
Gove, J.M., McManus, M.A., Neuheimer, A.B., Polovina, J.P., Drazen
, J.C., Smith, C.R., Merrifield, M.A., Friedlander, A.M., Ehses, J.S.,
Young, C.W., Dillon, A.K. and Williams, G.J. (2016). ‘Near-island
biological hotspots in barren ocean basins’. Nature Communications 7:
10581.
Stoltzenberg, W. (2011). Rat Island. New York: Bloomsbury USA.
Glen, A.S., Atkinson, R., Campbell, K.J., Hagen, E., Holmes, N.D.,
Keitt, B.S., Parkes, J.P., Saunders, A., Sawyer, J. and Torres, H. (2013).
‘Eradicating multiple invasive species on inhabited islands: The next
big step in island restoration’. Biological Invasions 15: 2589–2603.
Towns, D.R. and Broome, KG. (2003) ‘From small Maria to massive
Campbell: Forty years of rat eradications from New Zealand islands’.
New Zealand Journal of Zoology 30: 377–398.
Gillespie, R. (2016). ‘Economic Evaluation of the Lord Howe Island
Rodent Eradication Project’. Gillespie Economics, unpublished. Report
to the Lord Howe Island Board.
Howald, G., Donlan, C.J. Galvan, J.P. Russell, J.C. Parkes, J. Samaniego,
A. Wang Y., Veitch, D. Genovesi, P. Pascal, M. Saunders, A. and Tershy,
B. (2007). ‘Invasive rodent eradication on islands’. Conservation Biology
21: 1258–1268.
Jones, H.P., Holmes, N.D., Butchart, S.H., Tershy, B.R., Kappes, P.J.,
Corkery, I., Aguirre-Muñoz, A., Armstrong, D.P., Bonnaud, E., Burbidge,
A.A., Campbell, K., Courchamp, F., Cowan, P.E., Cuthbert, R.J.1.,
Ebbert, S., Genovesi, P., Howald, G.R., Keitt, B.S., Kress, S.W., Miskelly,
C.M., Oppel, S., Poncet, S., Rauzon, M.J., Rocamora, G., Russell, J.C.,
Samaniego-Herrera, A., Seddon, P.J., Spatz, D.R., Towns, D.R. and
Croll, D.A. (2016). ‘Invasive mammal eradication on islands results in
substantial conservation gains’. Proceedings of the National Academy
of Sciences 113: 4033–4038.
Keitt, B., Campbell, K., Saunders, A., Clout, M., Wang, Y., Heinz,
R., Newton, K. and Tershy, B. (2011). ‘The global islands invasive
vertebrate eradication database: a tool to improve and facilitate
restoration of island ecosystems’. In: C.R. Veitch, M.N. Clout and
D.R. Towns (eds.) Island invasives: eradication and management, pp.
74–77. Occassional Paper SSC no. 42. Gland, Switzerland: IUCN and
Auckland, New Zealand: CBB.
Keitt, B., Griffiths, R., Boudjelas, S., Broome, K., Cranwell, S., Millet, J.,
Pitt, W. and Samaniego-Herrera, A. (2015) ‘Best practice guidelines for
rat eradications on tropical islands’. Biological Conservation 185: 17–26.
Kier, G., Kreft, H., Lee, T.M., Jetz, W., Ibisch, P.L., Nowicki, C., Mutke,
J. and Barthlott, W. (2009). ‘A global assessment of endemism and
species richness across island and mainland regions’. Proceedings of
the National Academy of Sciences 106: 9322–9327.
636
Tershy, B.R., Shen, K.W., Newton, K.M., Holmes, N.D. and Croll D.A.
(2015). ‘The importance of islands for the protection of biological and
linguistic diversity’. BioScience 65: 592–597.
[UNEP-WCMC] ‘United Nations Environment Programme World
Conservation Monitoring Centre (2013). Global Islands Database
UNEP-WCMC 30 January, 2013’. <www.unep-wcmc.org>.
Veitch, C.R. and Clout, M.N. (2002). ‘Turning the tide of biological
invasion: the potential for eradicating invasive species’ In: C.R. Veitch
and M.N. Clout (eds.) Turning the tide: the eradication of invasive
species, pp. 1–3. Occasional Paper SSC no. 28. IUCN SSC Invasive
Species Specialist Group. IUCN, Gland, Switzerland and Cambridge,
UK.
Veitch, C.R., Clout, M.N. and Towns, D.R. (2011). ‘Invasive alien species:
losses, gains, and engaging the public’. In: C.R. Veitch, M.N. Clout and
D.R. Towns (eds.) Island invasives: eradication and management, p.
XI. Occassional Paper SSC no. 42. Gland, Switzerland: IUCN and
Auckland, New Zealand: CBB.
Will, D.J., Campbell, K.C. and Holmes, N.D. (2015). ‘Using digital
data collection tools to improve overall cost-efficiency and provide
timely analysis for decision making during invasive species eradication
campaigns’. Wildlife Research 41: 499–509.
Will, D.J., Swinnerton, K., Silander, S., Keitt, B., Griffiths, R., Howald,
G.R., Figuerola-Hernandez, C.E., and Herrera-Giraldo, J.L. (2019).
‘Applying lessons learnt from tropical rodent eradications: a second
attempt to remove invasive rats from Desecheo National Wildlife Refuge,
Puerto Rico’. In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and
C.J. West (eds.) Island invasives: scaling up to meet the challenge, pp.
154–161. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
G.E. Key and N.P. Moore
Key, G.E. and N.P. Moore. Tackling invasive non-native species in the UK Overseas Territories
Tackling invasive non-native species in the UK Overseas Territories
G.E. Key and N.P. Moore
GB Non-Native Species Secretariat, Animal and Plant Health Agency, Sand Hutton, York YO41 1LZ, UK.
<jillian.key@apha.gov.uk>.
Abstract The 16 UK Overseas Territories (OTs) together account for 94% of the UK’s unique biodiversity and make a
significant contribution to global biodiversity. Being predominantly islands, the OTs are very vulnerable to the introduction
of potentially harmful invasive non-native species, and pressures are increasing with the continual growth of international
trade and impact of climate change. Biosecurity is acknowledged as the most cost-effective means of addressing invasive
species threats for small islands, and yet the OTs face many challenges in the implementation of biosecurity controls.
In 2016 a UK Government funded project “Tackling Invasive Non-Native Species in the UK Overseas Territories” was
initiated to improve the biosecurity of the OTs against non-native species in order to improve their environmental resilience
and food security through technical assistance and capacity building. A gap analysis carried out in early 2017 assessed the
strengths and weaknesses for all 16 OTs along the biosecurity continuum in three areas: (1) prevention (2) early warning
and rapid response, and (3) long-term management. Overall, capacity is weakest in the area of prevention and greatest
in that of long-term management. Border activities, where implemented, are primarily linked to agricultural production
and animal health. Few OTs have carried out horizon scanning or comprehensive pathway analysis or have the capacity
to carry out pest risk analysis. Greatest capacity is seen in the relatively well resourced Antarctic and sub-Antarctic
territories, and in St Helena Island which was the subject of a 4-year project in anticipation of air access. Legislation
is generally weak, and few OTs have developed territorial biosecurity policies or strategies. Officers responsible for
biosecurity often have a range of functions in addition to their biosecurity roles, lack access to specialist expertise and
diagnostic facilities, and may also lack access to appropriate training. This compromises their ability to deliver effective
biosecurity. This situation is common to many small island states.
Keywords: biosecurity, capacity, gap analysis, horizon-scanning, pathway-analysis, prevention
INTRODUCTION
The 16 UK Overseas Territories (OTs) together
account for 94% of the UK’s unique biodiversity and as
such make a significant contribution to global biodiversity
(Churchyard, et al., 2014). Despite this, involvement of
the UK government in the OTs with regards provision
of financial and other resources is minimal, with the OTs
receiving only project funding from the UK (e.g. Vaas,
et al., 2017). Being predominantly islands, the OTs are
very vulnerable to the introduction of potentially harmful
invasive non-native species, recognised as the biggest
threat to island biodiversity, as well as to food security and
sustainable development (Copsey, et al., 2018). Pressures
are increasing with the continual growth of international
trade, the main driver of the spread of invasive species,
resulting in higher numbers of individuals of more species
being moved around the world, both deliberately and
accidentally. The chances of a new potentially harmful
species arriving and establishing in a new area are therefore
greater. The implementation of biosecurity measures is
aimed at minimising this risk (Copsey, et al, 2018), and
contributes towards achievement of Strategic Goal B of
the Convention on Biological Diversity, Reduce the direct
pressures on biodiversity and promote sustainable use, and
specifically Aichi Target 9 (UNEP, 2011).
Biosecurity, defined as measures to reduce the risk of
introducing or spreading invasive non-native species (and
other harmful organisms such as diseases) in the wild, has
long been acknowledged as the most cost-effective means
of addressing invasive species threats for small islands (for
example Tye, 2009). To be effective, actions need to be
implemented across the biosecurity continuum, with preborder controls at the country of origin, inspections and
interceptions at the border, and post-border surveillance
and interventions in the wider environment, all applied
to both deliberate (legal and illegal) and accidental
introductions. Once implemented, biosecurity actions must
be maintained as part of normal government practice.
The IUCN announced the Honolulu Challenge at the
World Conservation Congress 2016, calling for greater
action to tackle the issue of invasive non-native species
across the globe, with particular attention to preventative
action and the development of effective biosecurity policies
(IUCN, 2017).
As part of the UK Government’s response the 3-year
project Tackling Invasive Non-Native Species in the UK
Overseas Territories was initiated. Its objective is “to
improve the biosecurity of the OTs against invasive nonnative species to improve their environmental resilience
and food security; achieved through reducing the risk
and impact of invasion and natural hazards via technical
assistance and capacity building”.
In order to plan the appropriate capacity building
activities, a gap analysis was carried out in January 2017
on biosecurity practices and capacity in all 16 UK OTs
(Fig. 1) (information from McPherson, 2016):
Anguilla: one main and a number of smaller islands
in the Caribbean region with a total area of 90 km2 and
population of 13,572.
Ascension Island: a single main island in the South
Atlantic, with an area of 87 km2 and population of 1,000.
Bermuda: eight connected islands and over 190 smaller
islands in the wider Caribbean with a total area of 53.7 km2
and population of 65,038.
British Antarctic Territory (BAT): the Antarctic
Peninsula and two groups of nearby islands, with a total
area of 1,709,400 km2 and no permanent population.
British Indian Ocean Territory (BIOT); archipelago of
over 50 small low-lying islands in the Indian Ocean, with
a total area of 50 km2 and no permanent population, but a
large permanent military presence.
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 637–642. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
637
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
British Virgin Islands (BVI): Four main islands and
over 50 small islets and cays in the Caribbean, with a total
area of 151 km2 and population of 28,882.
Cayman Islands: three islands in the Caribbean, with a
total area of 264 km2 and population of 54,397.
Cyprus Sovereign Base Areas (CSBA): two separate
areas, Akrotiri-Episkopi (the Western SBA) and Dhekelia
(the Eastern SBA), on the island of Cyprus in the
Mediterranean, with a total area of 254 km2 and population
of 15,700.
Falkland Islands: two main islands and over 770
smaller islands in the South Atlantic, with a total area of
12,173 km2 and population of 2,841.
Gibraltar: a peninsula at the southern coast of Spain,
with an area of 6.8 km2 and population of 31,465.
Montserrat: a single island in the Eastern Caribbean,
with an area of 102 km2 and population of 4,922.
Pitcairn Islands: four islands in the South Pacific, with
a total area of 48.7 km2 and population of 47, all resident
on the main island.
St Helena Island: a single main island in the South
Atlantic, with an area of 121 km2 and population of 4,534.
South Georgia and South Sandwich Islands (SGSSI):
one main island and several small ones in the South
Georgia group and a group of 11 small islands in the South
Sandwich Islands Group, all in the sub-Antarctic. Total
area is 3,903 km2 with no permanent population.
Tristan da Cunha: four islands in the South Atlantic,
with a total area of 207 km2 and population of 268.
Turks and Caicos Islands: two island groups of over
120 small islands in the Caribbean, with a total area of 417
km2 and population of 49,000.
METHODS
A questionnaire was designed, identifying the
components required for an effective biosecurity
programme along the biosecurity continuum. Emphasis was
given to the pre-border and post-border activities targeted
Fig. 1 The 16 UK Overseas Territories.
638
by the project, grouped in three areas: 1) Prevention; 2)
Early Warning and Rapid Response (EWRR); and (3)
Management, Prioritisation and Frameworks (MPF) in the
components defined as follows:
Prevention
Pest Risk Analysis (PRA): system established and in
use to evaluate the likelihood of the entry, establishment,
or spread of a pest or disease, and the associated potential
biological and economic consequences. Both phytosanitary
and zoosanitary risks covered.
Non-Native Species Risk Analysis (NNRA):
comprehensive risk assessment frameworks exist to assess
the risk of non-native species (plant and animal) becoming
invasive.
Pathway Analysis: prioritised pathways of entry
identified, and results used as the basis for procedures.
Horizon Scanning: horizon scanning exercise carried
out to identify invasive species most likely to invade via
identified pathways.
Contingency Planning: formalised generic contingency
plan or plans in place to deal with priority invasive species
that are likely to arrive. This is divided into (i) Plants,
including both plants and plant health risks (non-native
plant pests and diseases); (ii) Animals, including both
vertebrates and animal health risks (non-native vertebrates,
animal diseases and parasites); and (iii) Other risks
(invertebrates other than plant pests, and marine species).
Border Operations: in-place and operational,
considering staffing, provision of dedicated facilities,
procedures and protocols in place, public awareness, and
levels of compliance. Both phytosanitary and zoosanitary
risks covered.
Early warning and rapid response
Alert System: clear system in place for reporting
incursions or new species, for both plant and animal
(vertebrate and invertebrate) risks.
Surveillance: generic and/or incursion specific
programmes in place for surveillance of priority invasive
Key & Moore: Invasive species in UK Overseas Territories
species. This is divided into (i) Plants, including both
plants and plant health risks (non-native plant pests
and diseases); (ii) Animals, including both vertebrates
and animal health risks (non-native vertebrates, animal
diseases and parasites); and (iii) Other risks (invertebrates
other than plant pests, and marine species).
Monitoring: generic and/or incursion specific
programmes in place for monitoring established priority
invasive species.
Rapid Response Capacity: capacity (capability and
resources) to provide rapid response to incursions. This
is divided into (i) Plants, including both plants and plant
health risks (non-native plant pests and diseases); (ii)
Animals, including both vertebrates and animal health risks
(non-native vertebrates, animal diseases and parasites);
and (iii) Other risks (invertebrates other than plant pests,
and marine species).
Management, prioritisation and frameworks
Prioritisation: prioritised established invasive species
for control/eradication based on global risk management
best practice, such as the Guidelines for invasive species
management in the Pacific (Tye, 2009).
Baseline Data: baseline inventories available for plants
(native and non-native), animals (terrestrial vertebrates and
invertebrates), and other (marine species).
Territorial Framework: biosecurity legislation in place
and enforced; biosecurity strategy or policy in place or
endorsed, and being implemented.
Contacts were established in each OT for both
agriculture and environment sectors, and territory capacity
assessed though a combination of email, telephone
interviews and face to face interviews. At least two people
were involved in each territory, with the exception of BAT
where there was only one. Capacity for each component
was rated and scored as follows:
None - No action taken / Nothing in place. Score of 0
Basic - Some actions taken / Basic framework or actions
in place / Actions planned in near future and expected to
take place. Score of 1.
Some - Some substantial advances while other actions
remain to be done / Actions being actively implemented
along a planned timeframe. Score of 2.
Good - Substantive actions taken / Substantial
framework or actions in place / Action being implemented
/ Action achieved. Score of 3.
Scores were summed across the components and
territories to provide a simple index for comparison
purposes. The text and ratings assigned to the components
were in all cases agreed and approved by the contacts incountry for each territory. The resulting scores were then
cross-checked by Dr Niall Moore of the GB Non-Native
Species Secretariat to ensure that the ratings matched
the comments; any adjustments were then discussed and
agreed by the relevant contacts.
Final scores were checked by visitors from the RSPB,
IUCN, Animal and Plant Health Agency (APHA), UK, and
South Georgia Heritage Trust with recent experience of
the relevant territory. Again, any discrepancies were then
discussed and agreed by the relevant contacts before the
scores were finalised.
RESULTS
Responses were obtained from all 16 OTs. Overall,
respondents welcomed the project and expressed frustration
where they identified gaps in their territory.
Table 1 Overall scores in the areas of Prevention, Early Warning and Rapid Response (EWRR) and
Management, Prioritisation and Frameworks (MPF) and total scores for each of the 16 OTs and the UK
in ascending order, out of a maximum overall score of 66. Maximum possible scores per area are 24
(Prevention and EWRR) and 18 (MPF). Overall mean score excludes that for the UK.
Territory
Turks and Caicos
BIOT
CSBA
Montserrat
Ascension
Anguilla
Bermuda
Tristan da Cunha
Pitcairn
Falkland Islands
Cayman
BVI
Gibraltar
BAT
St Helena
SGSSI
UK
Overall mean score for the OTs
Prevention
4
3
3
5
5
8
5
7
9
11
11
10
3
17
14
14
21
8.1
EWRR
8
5
7
8
8
4
9
7
10
10
9
14
17
11
18
19
20
10.3
MPF
7
12
11
9
10
12
12
12
7
10
13
10
17
17
13
18
17
11.9
Overall score
19
20
21
22
23
24
26
26
26
31
33
34
37
45
45
51
58
639
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
Differences between territories
Scores for each territory in the three categories of
Prevention, EWRR and MPF are shown in Table 1, with
the territories listed from the lowest overall score (weakest
capacity) to the highest (most capacity). The estimated
score for the UK is given for comparison.
The three highest scoring territories are the two
Antarctic and sub-Antarctic territories, and St Helena. BAT
and SGSSI (with total scores of 45 and 51 respectively)
benefit from their unique environmental status and
considerable research input. St Helena (with a total score
of 45) has been the subject of a 4-year project to strengthen
biosecurity in anticipation of air access. The total score for
SGSSI (51) is closest to that estimated for the UK (58).
A group of four territories have total scores between 31
and 36, comprising in ascending order: Falkland Islands,
Cayman Islands, BVI and Gibraltar; Gibraltar has a score
accounting for less than 20% of their overall score in the
area of Prevention, but scores highly in the other areas.
A group of nine territories have the lowest totals, with
scores between 19 and 26 and only one or two points
between each, comprising in ascending order: Turks and
Caicos Islands, BIOT, CSBA, Montserrat, Ascension
Island, Anguilla, Bermuda, Tristan da Cunha and Pitcairn.
Three territories are particularly weak in the area of
Prevention, with scores accounting for less than 20% of
their overall score: BIOT, CBSA and Bermuda. All four
have ratings of Basic or None for all components in this
area with only two exceptions: Bermuda with a rating of
Some for border operations, and CBSA with a rating of
Some for contingency planning for animals and animal
health risks. Anguilla has a total score accounting for less
than 20% of overall in EWRR, with all ratings in this area
of Basic or None.
Components of biosecurity
The overall capacity is weakest in the area of Prevention,
with an average score of 8.1, and strongest in the area of
Management, Prioritisation and Frameworks, with an
average score of 11.9 (Table 1). Table 2 shows total scores
by component out of a maximum possible score of 48.
The highest scoring components are the group
encompassing baseline inventories. This is generally
good and especially for plants, with a total score of 43.
Baseline knowledge for animals (terrestrial vertebrates and
invertebrates) and other (marine species) both had a total
score of 35.
The next highest scoring component is a group of four
with scores of 28 to 30: alert system, prioritisation, legal
framework and border operations.
The greatest capacity gaps are those of horizon
scanning and contingency planning for other risks, both
with total scores of 8. The second greatest gap is a group
of three components: rapid response for other risks,
surveillance of other risks and non-native risk analysis, all
with scores of 12. Only five OTs have carried out horizon
scanning, rated as Good only for BAT which has benefitted
from considerable research input. The other OTs did not
Table 2 Total scores for each component: the maximum possible score is 48.
Component
Prevention
Risk Analysis (PRA)
Risk Analysis (NNRA)
Pathway Analysis
Horizon scanning
Contingency Planning
Border operations
Early Warning and Rapid Response
Alert System in Place
Surveillance
Monitoring
Rapid response Capacity
Long-term management
Prioritisation
Baseline
Framework
640
Total score
Plants and plant health risks
Animals and animal health risks
Other risks
16
12
17
8
15
23
8
30
Plants and plant health risks
Animals and animal health risks
Other risks
30
23
17
12
23
23
24
12
Plants
Animals
Other
Legal
Territorial policy or strategy
29
43
35
35
28
20
Plants and plant health risks
Animals and animal health risks
Other risks
Key & Moore: Invasive species in UK Overseas Territories
understand what horizon scanning was. “Other risks”
comprises non-crop pest invertebrates and marine species,
for which capacity is clearly weaker than for crop pests or
plants; even for the UK where surveillance for other risks
was the only component which was rated Basic, all the
other components being rated as Some or Good.
DISCUSSION
The relatively small population size of the OTs means
that biosecurity officers often have a range of functions
and responsibilities in addition to their biosecurity roles,
lack access to specialist expertise and diagnostic facilities,
and may also lack access to appropriate training. This
compromises their ability to deliver effective biosecurity.
There is a dependence on community support, itself
dependant on good levels of awareness and understanding.
Officers carrying out biosecurity functions work closely
with customs, and this is clearly an important partnership.
Biosecurity practices tend to be based on historic
legislation inherited from their colonial pasts and not
updated, with procedures aimed at protecting agriculture
and production, focusing primarily on managing deliberate
introductions to reduce the introduction of crop pests and
livestock diseases, with a few exceptions (e.g. BAT and
SGSSI). Legislation is weak and scattered across a number
of regulations relating to customs, plant health and animal
health. The broader threat posed by non-native invasive
species to the environment is not being recognised, and
extension of biosecurity approaches to species, which are
not crop pests or livestock diseases, is generally poor or
non-existent.
For many OTs, actions such as border operations and
post-border surveillance are focused on easily-identifiable
species such as Pacific lionfish Pterois volitans, brown
tree snake Boiga irregularis and Tephritid fruit flies
(Diptera: Tephritidae). While this is a good starting point
for biosecurity teams, actions need to go further, and
target more cryptic species identified as priority, as well
as taking a generic approach to detect the unexpected.
Biosecurity actions across the continuum are particularly
weak for non-crop pest invertebrates, except where there
has been a historic incident of note, such as the jacaranda
bug (Orthesia insignis) outbreak on endangered endemic
gumwood trees (Commidendrum robustum) in St Helena in
the mid-1990s, which raised attention within the Territory
to the issue of invasive non-native species.
BAT is distinct in being one of the few OTs which is
not an island but one of 29 national Antarctic programmes.
As such, BAT has no control over what is done on other
stations, or what the tourism industry does with regard to
biosecurity unless they come to BAT stations, rendering
it vulnerable to intra-Antarctic transfer of non-native
species. This issue is recognised as a concern in the
Antarctic and included by the Antarctic Treaty Committee
for Environmental Protection (CEP) in the 2016 CEP Nonnative Species Manual (Anon., 2016a).
CSBA and Gibraltar are also not islands and consist
of enclaves adjacent to EU countries (Spain and Cyprus).
CSBA has relatively few resources dedicated to biosecurity,
and with relatively long leaky land borders with the
Republic of Cyprus this is to be expected. Gibraltar puts
most attention into actions in the areas of Early Warning
and Rapid Response, and Management, Prioritisation and
Frameworks, with comprehensive monitoring programmes
for existing invasive species, and surveillance programmes
and rapid response capability in the event of an incursion.
Actions are detailed in the Biodiversity Action Plan (Perez,
2006).
Where OTs have rated capacity as Basic or above
in these components it is primarily due to the outcome
of a specific research project, usually UK-funded by a
competitive research grant such as a Darwin Plus award,
or builds on a topical invasive species issue such as the
Pacific lionfish (Pterois volitans) and pink hibiscus
mealybug (Maconellicoccus hirsutus) invasions in the
wider Caribbean (Morris & Whitfield, 2009; <http://www.
cabi.org/isc/datasheet/40171>).
Risk Analysis (PRA and NNRA) comes quite low, with
scores of 16 and 12 for PRA and NNRA respectively. Risk
analysis, when done correctly, is a time-consuming and
complex procedure which requires access to taxonomic
and other expertise and, in most cases, funding to bring
experts together. The small, resource-limited OTs are
challenged to achieve this, and most carry out simplified
forms of risk analysis as well as they can, on an ad-hoc
basis, with heavy reliance on published databases such
as the CABI Invasive Species Compendium and Global
Invasive Species Database, and on assessments carried out
for Florida, Hawaii and the Pacific Islands for plant species
(<http://www.hear.org/pier/wra.htm>). While these make a
good match for Pitcairn, their suitability to the other OTs
is less certain. Comprehensive, published assessments
specifically for the island groups in the Caribbean and
South Atlantic would be very helpful.
The introduction of new exotic species as pets is of
concern, particularly to the Caribbean territories, due to
the risk of escapes or deliberate dumping of potentially
invasive species in the wild. In the Caribbean, at least
some introductions are linked to hurricanes: in Anguilla
it is known that at least two monkeys escaped from an
individual, who had them as pets, after a hurricane in
1999, and the green iguana was first introduced on logs of
wood during a hurricane in 1995 (R. Connor, Government
of Anguilla, pers. comm.). Escapes of exotic fish are not
considered a big problem, probably due to the lack of large
bodies of fresh water inland in the OTs. Escapes of exotic
birds are also not considered a big issue. Currently, one of
the commonest domestic species of concern is the cat (Felis
catus) (R. Connor. pers. comm.). Unwanted kittens are
frequently dumped in the wild and form feral populations,
threatening wildlife such as the native Anguilla racer snake
(Alsophis rijgersmae), endemic Antillean iguana (Iguana
delicatissima), or endemic St Helena wirebird (Charadrius
sanctaehelenae) (Varnham, 2006).
With the exception of CSBA, all the OTs carry out
biosecurity border operations to a greater or lesser extent,
and 12 out of the 16 rated this as “Some” or “Good”.
Focusing limited resources on border inspections and
interceptions is cost-effective for islands where the border
is clearly defined and defendable. However, in a continental
context with leaky borders which cannot be readily
defended, an alternative strategic approach is to identify
the priority species or pathways of concern and work more
widely across the biosecurity continuum, particularly
post-border. Tactics adopted are based on the results of
pathway analysis and horizon scanning. In this context,
high scores across the board for all components aren’t
necessarily appropriate, instead a package of activities is
adopted designed to minimise the identified risks. CSBA
and Gibraltar are not island territories and have different
priorities. In CSBA, the focus is on the zoosanitary risks
of new animal disease outbreaks and public health issues,
routine monitoring is of aerial insect vectors, specifically
mosquitoes, and rapid response capacity exists to respond
in the event of human or animal health outbreak. Gibraltar
benefits from strong post-border monitoring, surveillance
and prioritisation actions to protect its unique biodiversity,
as laid out in the Gibraltar Biodiversity Action Plan and
Reserve Management Plan (Anon., 2016b; Perez, 2006).
641
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
Ascension Island and BIOT also rated border operations
as “Basic”. Both territories have limited or no agricultural
production and consequently little political incentive in the
past to invest in biosecurity border controls. The limited
resources available to biosecurity are targeted at postborder actions directed towards the highest risk species,
namely mosquitoes of human health concern and fire ants
in Ascension Island, and brown tree snake in BIOT. This
approach emphasises the importance of horizon scanning,
pathway analysis and accurate assessment of risks in the
first place, and the need to build capacity in these areas to
provide information on where to target resources.
PRIORITIES AND RECOMMENDATIONS
Aiming to build capacity for all OTs so that they have
high scores across the board is neither realistic nor suitable.
Whereas for many OTs an appropriate strategy would be
to devote a substantial proportion of available resources to
border operations, for others, such as CSBA or Ascension
Island, a more cost-effective strategy instead would be to
establish post-border surveillance programmes targeted
at identified priority species or pathways. In all the OTs
resources are limited, and officers must be very focused
in their activities. To do this effectively, each OT needs
basic information on the range of potential invasive species
(horizon scanning), how they might arrive (pathway
analysis) and how to assess risk (PRA and NNRA).
Capacity in these fundamentals was found to be lowest in
this gap analysis, and initial activities will concentrate in
this area:
Building fundamentals:
● Horizon scanning linked with pathway analysis:
to determine what potential invasive species are
out there and the different ways they can arrive.
The information is used to design an appropriate
package of responses which guides how the available
resources should be best divided up between
preventative actions, such as pathway or border
operations, and reactive actions, such as surveillance
and rapid response.
● Risk analysis: the process of assessing biosecurity
risks. OTs need access to support for risk analysis,
and a harmonised approach across the OTs to guide
practices on-island for:
● Assessment of plant or animal species for potentially
invasive characteristics;
● Assessment of the risks of a plant or animal species
carrying potentially harmful pests, parasites or
diseases.
Establishing the framework:
● Territorial policy or strategy: agreed actions to
achieve the appropriate package of response,
including a communications strategy for awareness
to improve compliance and internal advocacy to
promote government support.
● Legislation: regulate across the biosecurity
continuum, including actions to contain, control and
eradicate established invasive species. Provision
of model legislation would allow a harmonised
approach across OTs; assistance with drafting to
apply it at the territory level is also required.
Delivery:
● Training: on all aspects of biosecurity, with specific
needs varying with the Territory. This provides
essential underpinning to deliver the fundamentals
and framework outlined above.
642
Adding value:
● Regional coordination: use regional coordination
bodies where they exist and are active, linking
among the UKOTs and also to appropriate
independent countries and other territories.
● Build networks, either strengthening existing
or developing new ones, to promote sharing
and exchanges, and promote the confidence and
inspiration which result from peer-learning networks.
Building capacity in the activities outlined above
will equip the officers responsible for biosecurity in the
OTs with the capacity to develop other actions such as
contingency and rapid response planning.
ACKNOWLEDGEMENTS
We would like to take this opportunity to thank all the
people in the UKOTs who responded for their collaboration
with the gap analysis work, for taking valuable time and
effort to compile information in order to make sure the
results are as useful and accurate as possible.
REFERENCES
Anon. (2016a). Biosecurity Handbook, 4th Edition. BAS Environment
Office.
<https://www.bas.ac.uk/wp-content/uploads/2016/12/BASBiosecurity-Handbook-2016-FINAL.pdf>.
Anon. (2016b). Gibraltar Nature Reserve Management Plan 2016.
Consultative Draft. <https://www.gibraltar.gov.gi/new/sites/default/
files/HMGoG_Documents/2016-GNR_Management_Plan_FINALv2.
pdf>.
Churchyard, T., Eaton, M., Hall, J., Millett, J., Farr, A., Cuthbert, R. and
Stringer, C. (2014). The UK’s Wildlife Overseas: A Stocktake of Nature
in our Overseas Territories. Sandy, UK: RSPB
Copsey, J.A., Black, S.A., Groombridge, J.G. and Jones, C.G. (2018).
Species Conservation: Lessons from Islands. Cambridge: Cambridge
University Press.
IUCN (2017). International Union for Conservation of Nature Annual
Report 2016. Gland, Switzerland and Cambridge, UK: IUCN.
McPherson, S. (2016). Britain’s Treasure Islands. A Journey to the UK
Overseas Territories. Poole, UK: Redfern Natural History Productions.
Morris, J.A., Jr., and Whitfield, P.E. (2009). Biology, Ecology, Control
and Management of the Invasive Indo-Pacific Lionfish: An Updated
Integrated Assessment. NOAA Technical Memorandum NOS NCCOS
99.
Perez, C.E. (2006). Biodiversity Action Plan, Gibraltar: Planning for
Nature. Gibraltar: The Gibraltar Ornithological & Natural History
Society.
Tye, A. (2009). Guidelines for Invasive Species Management in the
Pacific: A Pacific Strategy for Managing Pests, Weeds and Other
Invasive Species. Apia, Samoa: SPREP.
UNEP (2011). The Strategic Plan for Biodiversity 2011–2020 and the
Aichi Biodiversity Targets. UNEP/CBD/COP/DEC/X/2, 29 October
2010, Nagoya, Japan. COP CBD Tenth Meeting.
Vaas, J., Driessen, P.P.J., Giezen, M., van Laerhoven, F. and Wassen,
M.J. (2017). ‘Who’s in charge here anyway? Polycentric governance
configurations and the development of policy on invasive alien species
in the semisovereign Caribbean’. Ecology and Society 22(4): 1.
Varnham, K. (2006). Non-native Species in UK Overseas Territories: A
Review. JNCC Report No. 372.
D.A. Knapp, J.J. Knapp, K.A. Stahlheber and T. Dudley
Knapp, D.A.; J.J. Knapp, K.A. Stahlheber and T. Dudley. A little goes a long way when controlling invasive plants for biodiversity conservation
A little goes a long way when controlling invasive plants
for biodiversity conservation
D.A. Knapp1, J.J. Knapp2, K.A. Stahlheber3 and T. Dudley4
Santa Barbara Botanic Garden, 1212 Mission Canyon Road, Santa Barbara, CA 93105, USA. <dknapp@sbbg.org>.
The Nature Conservancy, Ventura, CA, USA. 3University of Wisconsin, Green Bay, WI, USA. 4University of California,
Santa Barbara, CA, USA.
1
2
Abstract Invasive species, particularly animals, are being eradicated from islands at ever more ambitious scales. In order
to protect island biodiversity and the essential ecosystem functions that it provides, however, plant invasions should be
given more management attention. While many advances have been made, plant eradication is inherently more difficult
than animal eradication due to persistent seed banks, and eradication may not be possible for more extensive populations.
While maintenance control has been successful, critics question the sustainability and priority of these efforts, and targets
vary widely. Developing consistent and informed targets requires an understanding of how biodiversity varies with invader
cover, yet little is known about this topic. Our research suggests that limited control efforts may be highly beneficial. We
conducted a meta-analysis of 54 studies to investigate the effects of plant invasions on invertebrate diversity, incorporating
invader cover and residence time as potential causal mechanisms. We also contrasted restored plots with otherwise native
plots. We found that invertebrate species richness was 31% lower in exotic plots than in native plots, and that there is a
threshold at around 70% invader cover after which the negative effects are significant across all studies. Furthermore,
these negative effects tended to decrease with time, and invertebrate richness was even greater in restored plots. The
implication is that by removing 30% or less of invasive plant cover and restoring natives, we can achieve many of our
conservation goals. We argue that by maintaining invasive patches at or below 70% exotic cover at a site in the near
term, we can buy time for both the islands’ insect herbivores to adapt to use the invader, and for managers to continue
improving plant eradication technologies. By retaining native diversity in this way, we can help to increase the resistance
and resilience of these systems to global change and other stressors.
Keywords: invader management, invasive plants, invertebrates, island biodiversity, meta-analysis, threshold
INTRODUCTION
Islands support many organisms found nowhere else
in the world, and contribute disproportionately to global
biodiversity (Kier, et al., 2009). They also provide critical
habitat for 45% of the IUCN-listed species (Keitt, et al.,
2011). To protect this extraordinary biological diversity,
invasive species are being eradicated from islands at ever
more ambitious scales (Clout & Veitch, 2002; Burbidge,
2011), and eradication is increasingly promoted as an
important direction for island conservation. This is
promoting recovery of many rare and endangered species,
and of biodiversity as a whole (e.g. Klinger, et al., 2002;
Rauzon, et al., 2002). This retained biodiversity can
increase the stability of a system (Hautier, et al., 2015),
its resistance and resilience to global change (Mori, et
al., 2013; Isbell, et al., 2015) and its resistance to further
invasion (Tilman, 1999).
Island eradication and control efforts overwhelmingly
target invasive vertebrates, as an analysis of previous
Island Invasives conference proceedings reveals (Veitch
& Clout, 2002; Veitch, et al. 2011; 87% and 97% of the
papers, respectively). Yet plant invaders are also key
factors in native biodiversity decline (Wilcove, et al., 1998;
Gaertner, et al., 2009), with their impacts to disturbance
regimes, nutrient cycling, and fluxes of materials and
energy altering ecosystem structure and function (Mack
& D’Antonio, 1998; Liao, et al., 2008; Ehrenfeld, 2010).
Furthermore, invasive animal removals often result in the
ecological release of invasive plants (e.g., Klinger, et al.,
2002; Zavaleta, et al., 2001). In order to protect island
biodiversity and the essential ecosystem functions that it
provides, plant invasions should be given more management
attention. Yet eradication, the widely preferred alternative
to control (Clout & Veitch, 2002; Burbidge, 2011), is often
problematic for invasive plants.
Plant eradication is inherently more difficult, and
generally more expensive, than animal eradication due to
persistent seed banks, although many advances have been
made. Under the right conditions, seeds can persist for
several hundred years or more (Jha, 2005). On the Pacific
Islands of French Polynesia, Hawaii, and New Caledonia,
eradication of the invasive alien tree Miconia calvescens
has not yet been completed despite more than 15 years
of intensive control, due to a prolific and persistent seed
bank (Meyer, et al., 2011). Similar issues have plagued
an eradication programme for Sagina procumbens on
Gough Island in the South Atlantic, despite an impressive
array of innovative control techniques (Cooper, et al.,
2011). Invasive plant eradication can be achieved, but it
typically involves small populations, treated early in the
invasion process, with a swift and strong response (Mack
& Lonsdale, 2002; Rejmanek & Pitcairn, 2002).
Where eradication is not feasible, maintenance
control may be implemented. Maintenance control is
the “coordinated and consistent management of invasive
plants in order to maintain the plant population at low
levels” (University of Florida, 2018). This approach has
been successful, but typically requires a large labour force,
and critics question the sustainability and priority of these
efforts (Simberloff, 2009). Furthermore, targets for native
cover vary widely and invasive cover targets are typically
highly stringent. For example, a survey of 21 California
habitat restoration plans containing specified thresholds
(gathered via a Google search) reveals native cover targets
ranging from 15% to 90%, with an average target of 62%
(n=20). Exotic cover targets, on the other hand, were never
greater than 10% (n=7). It is also unclear how these targets
were derived. Developing consistent and informed targets
requires an understanding of how biodiversity varies with
invader cover, however little is known about this topic.
An important link between plant communities and the
greater food web is the invertebrate fauna. Invertebrates
are a key component of biodiversity, comprising 97% of all
animal species (Spelman, 2012) and playing key roles in
nutrient recycling, pollination, seed dispersal, energy flow,
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 643–650. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
643
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
and structuring plant and animal communities (Gullan &
Cranston, 2005). They also respond quickly, sensitively,
and locally to environmental changes (Kremen, et al.,
1993), and are thus excellent indicators of the consequences
of plant invasions and other disturbances. Analysis of
invertebrate responses to plant invasions can help delineate
the drivers of biodiversity and community patterns, thus
guiding the conservation and restoration of diverse native
ecosystems (Lodge, 1993; McMahon, et al., 2006).
We conducted a meta-analysis to investigate the
effects of plant invasions on invertebrate diversity (as
a whole, including both native and non-native species),
incorporating invader cover and residence time in the
system as potential explanatory variables. We also
contrasted the type of sites (restored or intact) used as the
native comparison. A meta-analysis approach can be used
to combine multiple studies and detect overall trends in
biotic responses to environmental factors. Our research
suggests that in control efforts, a little may go a long way.
METHODS
We compiled studies through both database queries
and subsequent surveys of the references cited in compiled
papers. We searched ISI Web of Science in November
2012, using the search string “Topic = (invasive OR
exotic AND plant) AND Topic = (arthropod* OR insect*
OR invertebrate*). From these searches, we assembled
106 published studies which compared insect, arthropod,
or other invertebrate diversity in invaded versus native
habitats. These studies were from both island and mainland
environments and included dissertations. Studies included
by richness and other diversity indices, which were
analysed separately. We extracted the data directly from
tables or from graphs using the programme DigitizeIt v. 1.5
(Island Bormann, Braunschweig, Germany: <http://www.
digitizeit.de>).
Fifty-four studies were eligible for testing using a
meta-analysis approach (means, variances and sample
sizes were reported) and are included in our meta-analysis
(Appendix 1). These studies represent a variety of habitat
types throughout the world, ranging from grassland to scrub
to riparian. Fifty-two of these studies reported invertebrate
richness, and fifteen studies reported values for diversity
indices incorporating evenness, with 12 reporting results
for the Shannon index, two for the Simpson’s index, and
one for Fisher’s alpha. Insects were the focus of 26 studies,
while 16 studies reported results for entire arthropod
assemblages, and 12 studies described results for other
invertebrate groups.
We extracted descriptor variables, where available, from
each study, including latitude, time since establishment of
the non-native plant at both the local (study site) and/or
regional (hundreds of square kilometres) scale, invader
cover, and whether or not the native-dominated site was
restored habitat. Where time since establishment was not
reported for a given study, we obtained this information
from other sources where possible. In order to utilise the
studies which reported cover classes or ranges rather than
exact values (over half of them), we placed invader cover
into six cover classes. We used natural breaks in the data
to develop the following classes: <10%, 10–30%, 30–50%,
50–70%, 70–90%, and >90%. Cover was thus considered
‘absolute’ and not relative. Studies reporting that the
invasive plant “formed a monoculture”, was “dense and
continuous,” or “completely dominated the landscape”
were conservatively classified into the 70–90% group. We
found that model results were not changed by reclassifying
these into either 50–70% or >90% cover.
We used the response ratio as an estimator of
effect size; in this case, the natural log of the ratio
644
(Xexotic/Xnative), where X represents the mean of either
invertebrate species richness or diversity index (analysed
separately) for a given study in either the ‘exotic’ or the
‘native’ locations. We chose the response ratio for several
reasons: first, we were interested in the magnitude of the
relative difference in invertebrate diversity between exotic
and native vegetation; second, use of the logarithm ensures
that deviations in these two variables are treated equally
(Hedges, et al., 1999). Lastly, it allowed us to assess both
the model and residual variation, giving an estimate of the
importance of the variables analysed here.
We calculated a single effect size per study by averaging
data collected over multiple years or seasons. When we
compared invertebrate richness or diversity in one native
area to those in multiple invaded areas or vice versa, we
calculated separate effect sizes for each comparison. When
studies included multiple levels of descriptor variables
(e.g. two or more establishment times), we calculated
an average effect size to determine the overall effect of
invasion (vs. native plant communities) but calculated
separate effect sizes for each level of the descriptor
variables when analysing the effects of these descriptor
variables on invertebrate richness or diversity.
We performed meta-analyses using the metafor
(Viechtbauer, 2010) package for R 2.15.0, and used
random effects models to calculate overall effect sizes for
invertebrate richness and diversity (Viechtbauer, 2010;
Gurevitch & Hedges, 1999). To estimate the variation in
the effect size described by different categorical variables
(cover, study scale, and type of control plot), we used
mixed-effects models using the Q statistic. This analysis
treats the variables as fixed but includes a random variance
component to account for variability across the studies.
In one case (invader cover), we also report results from
a fixed-effects model, which restricts our inferences to
the studies examined. For continuous descriptor variables
(latitude, invader time since establishment) we used
weighted generalised least squares regression to test their
relationships with effect size.
After accounting for the variation attributable to
descriptor variables, we estimated residual variation
(τ2) using a restricted maximum likelihood estimator
(Viechtbauer, 2005). For studies which reported results
for all descriptor variable groups (22), we used the Akaike
information criterion (AIC) to determine the model that
best fit the data.
RESULTS
Invertebrate species richness was 31% lower in exotic
plots than in native plots (effect size = -0.37 ± 0.10 on a
0–1 scale; Z = -5.48, p < 0.01; Fig. 1). There was a high
amount of variation in the studies using richness to indicate
diversity, however (Q = 111, p < 0.001). Invertebrate
diversity indices that incorporate evenness were less
strongly affected than richness values, but still 14% lower
in exotic plots (effect size = -0.15 ± 0.10; Z = -3.42, p <
0.01). Unlike the effect sizes for species richness, there was
not much variation among studies using diversity indices
(Q = 13, p > 0.50). The absolute value of latitude did not
explain a significant amount of heterogeneity in effect sizes
for species richness (Q = 1.09, p = 0.30), nor did study
scale (Q = 0.06, p = 0.97).
Using just data from native plots that had not undergone
habitat restoration, invaded plots had lower invertebrate
richness compared to native plots (-0.35 ± 0.07; Z = -5.02,
p < 0.01). There was a stronger effect when plots restored
to native species were used for comparison (-0.61 ± 0.17;
Z = -1.73, p = 0.08), although this was just a statistical
trend, likely due to both low sample size (n=11) and high
variability. When analysed together, effect sizes were
Knapp, et al.: Controlling invasive plants
significantly more negative for the comparisons between
invaded and restored sites than invaded vs. otherwise native
sites (Q = 5.1, p = 0.02; Fig. 1), indicating that invertebrate
diversity was even greater in restored plots than in native
plots that did not undergo habitat restoration.
At the local scale, the negative effects of invasive
plants on invertebrate richness were greatest at the shortest
time since establishment and decreased with time, but
this pattern relies on a few key data points and was only
marginally significant (Q = 3.0, p =0.08, Fig. 2). At the
regional scale, time since invader establishment was not
related to effect size (Q = 0.40, p > 0.50).
The impact of exotic plants on invertebrate species
richness was highly variable below 70% invader cover, and
only cover classes above 70% had confidence intervals that
did not overlap zero (Fig. 1). When the cover classes below
70% were combined into a single category, the difference
in effect sizes between exotic plant cover classes was
marginal in a mixed-model analysis (Q = 4.7, p = 0.09),
while the groups were very different when the data were
fitted to a fixed effects model (Q = 176, p < 0.0001).
In all models except time since establishment, residual
heterogeneity was significant (p < 0.01), indicating
substantial amounts of variation in the effects that were not
explained by the models. The effects of descriptor variables
on effect sizes for diversity indices were not analysed, both
because low sample sizes prevented it and because low
residual heterogeneity obviated the need for it.
DISCUSSION
Our results showed a clear negative effect of plant
invasions on invertebrate richness and diversity. This has
important implications for the diversity and function of the
system as a whole, since insects and other invertebrates
perform so many important roles in an ecosystem –
including food provisioning for higher trophic levels such
as reptiles and amphibians, birds, and small mammals
(Weisser & Siemann, 2004).
Fig. 1 Mean invertebrate richness effect sizes (± 95%
confidence limits) across all studies (top panel), as
well as between studies contrasting effect sizes where
native plots represented restored or intact habitats
(middle panel). The bottom panel shows mean richness
effect sizes for exotic plant cover classes. Numbers in
parentheses indicate the number of effect sizes and the
total number of studies, respectively (some studies had
more than one comparison).
Furthermore, the most consistent and significant
negative effects of plant invaders on invertebrate richness
occur when invasive plants comprise over 70% of cover.
One likely reason for this threshold is a decline in the
diversity of other plant species when an invader comes
to dominate; for instance, Almeida-Neto, et al. (2011)
found that only host plant richness explained the unimodal
relationship they found between insect herbivore richness
and invasive grass cover. Many previous studies have
shown that insect and arthropod diversity is positively
related to plant species richness, presumably owing to
structural and food diversity as well as abiotic variables
(e.g., temperature, moisture) (Price, et al., 2011).
The implication of these results is that, in general,
with a moderate reduction of invasive plant cover and
restoration of native plants to at least 30% cover, we
can achieve meaningful progress towards the goal of
biodiversity conservation. While some invasive plants
will have impacts below this threshold (e.g. Knapp,
2014) this provides a general guideline in the absence of
species-specific impact information. If a critical level of
plant and invertebrate diversity can be maintained, then so
can key ecosystem functions such as nutrient cycling and
pollination (Gullan & Cranston, 2005).
Many will be legitimately concerned about indefinite
“maintenance management” of plant invaders. Invasive
plant management is challenging, and requires a longterm commitment (e.g. Mack & Lonsdale, 2002; Meyer,
et al., 2011). However, holding that 70% line by removing
invaders and, when needed, restoring at least 30% native
plant cover will buy time, both: 1) to allow the islands’
insect herbivores to adjust to using the invader, and 2) for
managers to continue improving plant control technologies
and eradication strategies. We elaborate on these points
below.
A novel plant species may be avoided by insect
herbivores because it differs from native plants in
characteristics such as nutritional quality, chemical
composition, and architecture (Strong, et al., 1984; Kuhnle
& Muller, 2009). Even a plant that can technically be eaten
may be avoided because it is not recognised as a food source
(Lankau, et al., 2004; Dudley, et al., 2012). The number of
different herbivores using a novel plant tends to increase
with the invader’s time since establishment, however
(Kennedy & Southwood, 1984; Brandle, et al., 2008).
Fig. 2 Relationship between effect size for invertebrate
richness and time since invader establishment at a site
for the 22 studies for which these data were available.
Dashes indicate line of best fit.
645
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
In our meta-analysis, where we consider the richness of
invertebrates as a whole including multiple feeding guilds
in addition to herbivores, we found a trend for invertebrate
richness to increase with time since invader establishment
(Fig. 2). This effect was only marginally significant –
perhaps because it was driven by just a few key points, or
perhaps because the effect of residence time is not as strong
for invertebrates as a whole as it is for insect herbivores
alone.
While these natural enemies are adapting to utilise
invasive plant species over time, our control techniques are
improving – allowing for both larger and more efficient,
effective projects. For instance, a transition from ground
to helicopter shooting enabled the eradication of goats on
Western Australian islands (Burbidge & Morris, 2002), as
did Judas goat technologies (Campbell & Donlan, 2005).
Aerial surveys help with plant detection and eradication as
well (Coulston, 2002; Knapp, et al., 2011), and treatment
techniques have improved to avoid vectoring plant material
(Coulston, 2002). Experimentation with techniques from
hand-pulling to herbicide to heat and saltwater applications
have improved the efficacy of invasive grass control efforts
on Laysan Island (Flint & Rehkemper, 2002). Similarly,
better herbicides and mapping systems have improved
invasive plant control in New Zealand (Wotherspoon &
Wotherspoon, 2002). Improvement in baiting technology
has enabled the eradication of rats in multiple locations
(Thomas & Taylor, 2002; Howald et al., 2007). Lastly,
targeting multiple species at one time has proven to be both
efficient and effective (Griffiths, 2011; Morrison, 2011).
It is heartening that our results showed restored plots
containing even more invertebrate species than other
native plots relative to invaded plots (although with
greater variability). Flower visitors can be more diverse
at restoration than reference sites, even after < one year
(Waltz & Covington, 2004; Lomov, et al., 2010). This may
be because early-colonising butterflies can be attracted
to more open, sunny restored areas disturbed by earth
moving, invasive plant removal, and outplanting (Magoba
& Samways, 2010; Hanula & Horn, 2011a). Conversely,
butterfly richness can decrease as percent plant cover rises
(Florens, et al., 2010). Higher invertebrate richness in
restored areas is likely also related to greater plant richness
and cover (Hanula & Horn, 2011), perhaps due to elements
of both early- and later-successional communities being
present. In this case, richness would also decrease with
time as succession occurs.
CONCLUSION
The theme of this conference is “Scaling Up to Meet
the Challenge.” Invasive species eradication successes
are being achieved at ever-increasing scales, but more
attention should be paid to the significant threat of plant
invasions. Although invasive plant control is challenging,
our research suggests that reducing invader density to just
70% cover can have significant benefits for invertebrate
biodiversity and thus ecosystem function. Furthermore,
habitat restoration can give that diversity an extra boost.
While the existence of seed banks dictates that this is a
long-term proposition, we argue that, over time, insect
herbivores will adapt to using the invader, while land
managers develop ever-better control technologies.
The biodiversity that is thus conserved will increase
the resistance and resilience of these systems to further
invasion and other stressors such as global climate change
(Millennium Ecosystem Assessment, 2003; Haddad, et al.,
2011), and allow us to truly achieve island conservation.
646
REFERENCES
Almeida-Neto, M., Prado, P.I. and Lewinsohn, T.M. (2011).
‘Phytophagous insect fauna tracks host plant responses to exotic grass
invasion’. Oecologia 165: 1051–1062.
Ando, Y., Utsumi, S., and Ohgushi, T. (2010). ‘Community structure of
insect herbivores on introduced and native Solidago plants in Japan’.
Entomologia Experimentalis et Applicata 136: 174–183.
Bailey, J.K., Schweitzer, J.A. and Whitham, T.G. (2001). ‘Salt cedar
negatively affects biodiversity of aquatic macroinvertebrates’. Wetlands
21: 442–447.
Bartomeus, I., Vila, M., and Santamaria, L. (2008). ‘Contrasting effects of
invasive plants in plant-pollinator networks’. Oecologia 155: 761–770.
Bassett, I.E., Paynter, Q. and Beggs, J.R. (2012). ‘Invertebrate community
composition differs between invasive herb alligator weed and native
sedges’. Acta Oecologica–International Journal of Ecology 41: 65–73.
Bickel, T.O. and Closs, G.P. (2009). ‘Impact of partial removal of the
invasive macrophyte Lagarosiphon major (Hydrocharitaceae) on
invertebrates and fish’. River Research and Applications 25: 734–744.
Bock, C.E., Bock, J.H., Jepson, K.L. and Ortega, J.C. (1986). ‘Ecological
effects of planting African love-grasses in Arizona’. National
Geographic Research 2: 456–463.
Brandle, M., Kuhn, I., Klotz, S., Belle, C. and Brandl, R. (2008). ‘Species
richness of herbivores on exotic host plants increases with time since
introduction of the host’. Diversity and Distributions 14: 905–912.
Burbidge, A.A. and Morris, K.D. (2002). ‘Introduced mammal
eradications for nature conservation on Western Australian Islands: A
review’. In: C.R. Veitch and M.N. Clout (eds.) Turning the tide: the
eradication of invasive species, pp. 64–70. Occasional Paper SSC
no. 28. IUCN SSC Invasive Species Specialist Group. IUCN, Gland,
Switzerland and Cambridge, UK.
Burbidge, A.A. (2011). ‘2001 to 2010 and beyond: trends and future
directions in the eradication of invasive species on islands’. In: C.R.
Veitch, M.N. Clout and D.R. Towns (eds.) Island invasives: eradication
and management, pp. 515–519. Occassional Paper SSC no. 42. Gland,
Switzerland: IUCN and Auckland, New Zealand: CBB.
Burghardt, K.T., Tallamy, D.W. and Shriver, W.G. (2009). ‘Impact of
native plants on bird and butterfly biodiversity in suburban landscapes’.
Conservation Biology 23: 219–224.
Cameron, G.N. and Spencer, S.R. (2010). ‘Entomofauna of the introduced
Chinese tallow tree’. Southwestern Naturalist 55: 179–192.
Campbell, K. and Donlan, C.J. (2005). ‘Feral goat eradications on
islands.’ Conservation Biology 19: 1362–1374.
Chey, V.K., Holloway, J.D., Hambler, C. and Speight, M.R. (1998).
‘Canopy knockdown of arthropods in exotic plantations and natural
forest in Sabah, north-east Borneo, using insecticidal mist-blowing’.
Bulletin of Entomological Research 88: 15–24.
Christopher, C.C. and G.N. Cameron (2012). ‘Effects of invasive Amur
honeysuckle (Lonicera maackii) and white-tailed deer (Odocoileus
virginianus) on litter-dwelling arthropod communities’. American
Midland Naturalist 167: 256–272.
Clout, M.N. and Veitch, C.R. (2002). ‘Turning the tide of biological
invasion: the potential for eradicating invasive species’. In: C.R. Veitch
and M.N. Clout (eds.) Turning the tide: the eradication of invasive
species, pp. 1–3. Occasional Paper SSC no. 28. IUCN SSC Invasive
Species Specialist Group. IUCN, Gland, Switzerland and Cambridge,
UK.
Cooper, J., Cuthbert, R.J., Gremmen, N.J.M., Ryan, P.G. and Shaw, J.D.
(2011). ‘Earth, fire and water: applying novel techniques to eradicate the
invasive plant, procumbent pearlwort Sagina procumbens, on Gough
Island, a World Heritage Site in the South Atlantic’. In: C.R. Veitch,
M.N. Clout and D.R. Towns (eds.) Island invasives: eradication and
management, pp. 162–165. Occassional Paper SSC no. 42. Gland,
Switzerland: IUCN and Auckland, New Zealand: CBB.
Cord, E.E. (2011). ‘Changes in Arthropod Abundance and Diversity with
Invasive Grasses’. MSc thesis. Kingsville, TX: Texas A&M.
Coulston, G.J. (2002). ‘Control of invasive plants on the Poor Knights
Islands, New Zealand’. In: C.R. Veitch and M.N. Clout (eds.) Turning
the tide: the eradication of invasive species, pp. 79–84. Occasional
Paper SSC no. 28. IUCN SSC Invasive Species Specialist Group.
IUCN, Gland, Switzerland and Cambridge, UK.
de Groot, M., Kleijn, D. and Jogan, N. (2007). ‘Species groups occupying
different trophic levels respond differently to the invasion of seminatural vegetation by Solidago canadensis’. Biological Conservation
136: 612–617.
Dudley, T.L., Bean, D.W., Pattison, R.R. and Caires, A. (2012). ‘Selectivity
of a biological control agent, Diorhabda carinulata Desbrochers, 1870
(Coleoptera: Chrysomelidae) for host species within the genus Tamarix
Linneaus, 1753’. Pan-Pacific Entomologist 88: 319–341.
Knapp, et al.: Controlling invasive plants
Durst, S.L., Theimer, T.C., Paxton, E.H. and Sogge, M.K. (2008).
‘Temporal variation in the arthropod community of desert riparian
habitats with varying amounts of saltcedar (Tamarix ramosissima)’.
Journal of Arid Environments 72: 1644–1653.
Ehrenfeld, J.G. (2010). ‘Ecosystem consequences of biological invasions’.
Annual Review of Ecology, Evolution, and Systematics 41: 59–80.
Flint, E. and Rehkemper, C. (2002). ‘Control and eradication of the
introduced grass, Cenchrus echinatus, at Laysan Island, Central Pacific
Ocean’. In: C.R. Veitch and M.N. Clout (eds.) Turning the tide: the
eradication of invasive species, pp. 110–115. Occasional Paper SSC
no. 28. IUCN SSC Invasive Species Specialist Group. IUCN, Gland,
Switzerland and Cambridge, UK.
Florens, F.B.V., Mauremootoo, J.H., Fowler, S.V., Winder, L. and Baider,
C. (2010). ‘Recovery of an indigenous butterfly community following
control of invasive alien plants in a tropical island’s wet forests’.
Biodiversity and Conservation 19: 3835–3848.
Gaertner, M., Den Breeyen, A., Hue, C. and Richardson, D.M. (2009).
‘Impacts of alien plant invasions on species richness in Mediterraneantype ecosystems: a meta-analysis’. Progress in Physical Geography 33:
319–338.
Gerber, E., Krebs, C., Murrell, C., Moretti, M., Rocklin, R. and Schaffner,
U. (2008). ‘Exotic invasive knotweeds (Fallopia spp.) negatively
affect native plant and invertebrate assemblages in European riparian
habitats’. Biological Conservation 141: 646–654.
Gossner, M. and Ammer, U. (2006). ‘The effects of Douglas fir on treespecific arthropod communities in mixed species stands with European
beech and Norway spruce’. European Journal of Forest Research 125:
221–235.
Gremmen, N.J.M., Chown, S.L. and Marshall, D.J. (1998). ‘Impact of
the introduced grass Agrostis stolonifera on vegetation and soil fauna
communities at Marion Island, sub-Antarctic’. Biological Conservation
85: 223–231.
Griffiths, R. (2011). ‘Targeting multiple species – a more efficient
approach to pest eradication’. In: C.R. Veitch, M.N. Clout and D.R.
Towns (eds.) Island invasives: eradication and management, pp. 172–
176. Occassional Paper SSC no. 42. Gland, Switzerland: IUCN and
Auckland, New Zealand: CBB.
Gullan, P.J. and Cranston, P.S. (2005). The Insects: An Outline of
Entomology. Malden, MA, USA: Blackwell Publishing Ltd.
Gurevitch, J. and Hedges, L.V. (1999) ‘Statistical issues in ecological
meta-analyses’. Ecology 80: 1142–1149.
Haddad, N.M., Crutsinger, G.M., Gross, K., Haarstad, J. and Tilman, D.
(2011). ‘Plant diversity and the stability of foodwebs’. Ecology Letters
14: 42–46.
Hagen, E.N., Bakker, J.D. and Gara, R.I. (2010). ‘Aerial arthropod
communities of native and invaded forests, Robinson Crusoe Island,
Chile’. Environmental Entomology 39: 1159–1164.
Hanula, J.L. and Horn, S. (2011a). ‘Removing an exotic shrub from
riparian forests increases butterfly abundance and diversity’. Forest
Ecology and Management 262: 674–680.
Hanula, J.L. and Horn, S. (2011b). ‘Removing an invasive shrub (Chinese
privet) increases native bee diversity and abundance in riparian forests
of the southeastern United States’. Insect Conservation and Diversity
4: 275–283.
Hartley, M.K., Rogers, W.E. and Siemann, E. (2010). ‘Comparisons of
arthropod assemblages on an invasive and native trees: Abundance,
diversity and damage’. Arthropod–Plant Interactions 4: 237–245.
Harvey, K.J., Britton, D.R. and Minchinton, T.E. (2010). ‘Insect diversity
and trophic structure differ on native and non-indigenous congeneric
rushes in coastal salt marshes’. Austral Ecology 35: 522–534.
Hautier, Y., Tilman, D., Isbell, F., Seabloom, E.W., Borer, E.T. and Reich,
P.B. (2015). ‘Anthropogenic environmental changes affect ecosystem
stability via biodiversity’. Science 348: 336–340.
Hedges, L.V., Gurevitch, J. and Curtis, P.S. (1999). ‘The meta-analysis
of response ratios in experimental ecology’. Ecology 80: 1150–1156.
Herrera, A.M. and Dudley, T.L. (2003). ‘Reduction of riparian arthropod
abundance and diversity as a consequence of giant reed (Arundo donax)
invasion’. Biological Invasions 5(3): 167–177.
Hills, N., Hose, G.C., Cantlay, A.J. and Murray, B.R. (2008). ‘Cave
invertebrate assemblages differ between native and exotic leaf litter’.
Austral Ecology 33: 271–277.
Holmquist, J.G., Schmidt-Gengenbach, J. and Slaton, M.R. (2011).
‘Influence of invasive palms on terrestrial arthropod assemblages in
desert spring habitat’. Biological Conservation 144: 518–525.
Howald, G., Donlan, C.J., Galvan, J.P., Russell, J.C., Parkes, J.,
Samaniego, A., Wang, Y., Veitch, D., Genovesi, P. Pascal, M., Saunders,
A. and Tershy, B. (2007). ‘Invasive rodent eradication on islands’.
Conservation Biology 21: 1258–1268.
Hugel, S. (2012). ‘Impact of native forest restoration on endemic
crickets and katydids density in Rodrigues island’. Journal of Insect
Conservation 16: 473–477.
Isbell, F., Craven, D., Connolly, J., Loreau, M., Schmid, B., Beierkuhnlein,
C., Bezemer, T.M., Bonin, C., Bruelheide, H., de Luca, E., Ebeling, A.,
Griffin, J.N., Guo, Q.F., Hautier, Y., Hector, A., Jentsch, A., Kreyling,
J., Lanta, V., Manning, P., Meyer, S.T., Mori, A.S., Naeem, S., Niklaus,
P.A., Polley, H.W., Reich, P.B., Roscher, C., Seabloom, E.W., Smith,
M.D., Thakur, M.P., Tilman, D., Tracy, B.F., van der Putten, W.H., van
Ruiven, J., Weigelt, A., Weisser, W.W., Wilsey, B. and Eisenhauer, N.
(2015). ‘Biodiversity increases the resistance of ecosystem productivity
to climate extremes’. Nature 526: 574–577.
Jha, A. (2005). ‘How long can a seed stay alive?’ The Guardian (16
June
2005).
<https://www.theguardian.com/science/2005/jun/16/
thisweekssciencequestions1>. Accessed May 2018.
Kappes, H., Lay, R. and Topp, W. (2007). ‘Changes in different trophic
levels of litter-dwelling macrofauna associated with giant knotweed
invasion’. Ecosystems 10: 734–744.
Keitt, B., Campbell, K., Saunders, A., Clout, M. Wang, Y., Heinz,
R., Newton, K. and Tershy, B. (2011). ‘The global islands invasive
vertebrate eradication database: a tool to improve and facilitate
restoration of island ecosystems’. In: C.R. Veitch, M.N. Clout and
D.R. Towns (eds.) Island invasives: eradication and management, pp.
74–77. Occassional Paper SSC no. 42. Gland, Switzerland: IUCN and
Auckland, New Zealand: CBB.
Kennedy, C.E.J. and Southwood, T.R.E. (1984). ‘The number of species
of insects associated with British trees – a re-analysis’. Journal of
Animal Ecology 53: 455–478.
Kier, G., Kreft, H., Lee, T.M., Jetz, W., Ibisch, P., Nowicki, C., Mutke,
J. and Barthlott, W. (2009). ‘A global assessment of endemism and
species richness across island and mainland regions’. Proceedings of
the National Academy of Sciences 106: 9322–9327.
Klinger, R.C., Schuyler, P. and Sterner, J.D. (2002). ‘The response of
herbaceous vegetation and endemic plant species to the removal of feral
sheep from Santa Cruz Island, California’. In: C.R. Veitch and M.N.
Clout (eds.) Turning the tide: the eradication of invasive species, pp.
141–154. Occasional Paper SSC no. 28. IUCN SSC Invasive Species
Specialist Group. IUCN, Gland, Switzerland and Cambridge, UK.
Knapp, J.J., Schuyler, P.T., Walker, K.N., Macdonald, N.L. and Morrison,
S.A. (2011). ‘Benefits of supporting plant and animal eradication
projects with helicopters’. In: C.R. Veitch, M.N. Clout and D.R.
Towns (eds.) Island invasives: eradication and management, pp. 188–
191. Occassional Paper SSC no. 42. Gland, Switzerland: IUCN and
Auckland, New Zealand: CBB.
Knapp, D.A. (2014). ‘Effects of an Exotic Plant Invasion on Arthropod
Assemblages.’ PhD thesis. Santa Barbara, CA, USA: University of
California, Santa Barbara.
Kremen, C., Colwell, R.K., Erwin, T.L., Murphy, D.D., Noss, R.F. and
Sanjayan, M.A. (1993). ‘Terrestrial arthropod assemblages: Their use in
conservation planning’. Conservation Biology 7: 796–808.
Kuhnle, A. and Muller, C. (2009). ‘Differing acceptance of familiar and
unfamiliar plant species by an oligophagous beetle’. Entomologia
Experimentalis et Applicata 131: 189–199.
Lankau, R.A., Rogers, W.W. and Siemann, E. 2004. ‘Constraints on the
utilisation of the invasive Chinese tallow tree Sapium sebiferum by
generalist native herbivores in coastal prairies’. Ecological Entomology
29: 66–75.
Liao, C., Peng, R., Luo, Y., Zhou, X., Wu, X., Fang, C., Chen, J. and
Li, B. (2008). ‘Altered ecosystem carbon and nitrogen cycles by plant
invasion: a meta-analysis’. New Phytologist 177: 706–714.
Lodge, D.M. (1993). ‘Biological invasions: Lessons for ecology’. Trends
in Ecology and Evolution 8: 133–137.
Lomov, B., Keith, D.A. and Hochuli, D.F. (2010). ‘Pollination and plant
reproductive success in restored urban landscapes dominated by a
pervasive exotic pollinator’. Landscape and Urban Planning 96: 232–
239.
McMahon, S.M., Cadotte, M.W. and Fukami, T. (2006). ‘Tracking the
Tractable: Using Invasions to Guide the Exploration of Conceptual
Ecology’. In: M.W. Cadotte, S.M. McMahon and T. Fukami (eds.)
Conceptual Ecology and Invasion Biology, pp. 3–14. Dordrecht, the
Netherlands: Springer.
Mack, M.C. and D'Antonio, C.M. (1998). ‘Impacts of biological invasions
on disturbance regimes’. Trends in Ecology and Evolution 13: 195–198.
Mack, R.N. and Lonsdale, W.M. (2002). ‘Eradicating invasive plants:
hard-won lessons for islands’. In: C.R. Veitch and M.N. Clout (eds.)
Turning the tide: the eradication of invasive species, pp. 164–172.
Occasional Paper SSC no. 28. IUCN SSC Invasive Species Specialist
Group. IUCN, Gland, Switzerland and Cambridge, UK.
Magoba, R.N. and Samways, M.J. (2010). ‘Recovery of benthic
macroinvertebrate and adult dragonfly assemblages in response to
large scale removal of riparian invasive alien trees’. Journal of Insect
Conservation 14: 627–636.
Magoba, R.N. and M.J. Samways (2012). ‘Comparative footprint of
alien, agricultural and restored vegetation on surface-active arthropods’.
Biological Invasions 14: 165–177.
647
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
Magura, T., Tothmeresz, B. and Bordan, Z. (2000). ‘Effects of nature
management practice on carabid assemblages (Coleoptera: Carabidae)
in a non-native plantation’. Biological Conservation 93: 95–102.
Meyer, J-Y, Loope, L. and Goarant, A-C. (2011). ‘Strategy to control the
invasive alien tree Miconia calvescens in Pacific Islands: eradication,
containment or something else?’. In: C.R. Veitch, M.N. Clout and
D.R. Towns (eds.) Island invasives: eradication and management, pp.
91–96. Occassional Paper SSC no. 42. Gland, Switzerland: IUCN and
Auckland, New Zealand: CBB.
Millennium Ecosystem Assessment (2003). Ecosystems and human wellbeing: Biodiversity synthesis. Washington, DC, USA: World Resources
Institute.
Mori, A.S., Furukawa, T. and Sasaki, T. (2013). ‘Response diversity
determines the resilience of ecosystems to environmental change’.
Biological Reviews 88: 349–364.
Moron, D., Lenda, M., Skorka, P., Szentgyorgyi, H., Settele, J. and
Woyciechowski, M. (2009). ‘Wild pollinator communities are
negatively affected by invasion of alien goldenrods in grassland
landscapes’. Biological Conservation 142: 1322–1332.
Morrison, S.A. (2011). ‘Trophic considerations in eradicating multiple
pests’. In: C.R. Veitch, M.N. Clout and D.R. Towns (eds.) Island
invasives: eradication and management, pp. 208–212. Occassional
Paper SSC no. 42. Gland, Switzerland: IUCN and Auckland, New
Zealand: CBB.
Osunkoya, O.O., Polo, C. and Andersen, A.N. (2011). ‘Invasion impacts
on biodiversity: Responses of ant communities to infestation by
cat’s claw creeper vine, Macfadyena unguis-cati (Bignoniaceae) in
subtropical Australia’. Biological Invasions 13: 2289–2302.
Parr, C.L., Ryan, B.J. and Setterfield, S.A. (2010). ‘Habitat complexity
and invasive species: The impacts of Gamba grass (Andropogon
gayanus) on invertebrates in an Australian tropical savanna’. Biotropica
42: 688–696.
Pinto, C., Sousa, J.P., Graca, M.A.S. and da Gama, M.M. (1997). ‘Forest
soil Collembola: Do tree introductions make a difference?’ Pedobiologia
41: 131–138.
Price, P.W., Denno, R.F., Eubanks, M.D., Finke, D.L. and Kaplan, I.
(2011). Insect Ecology: Behavior, Populations and Communities.
Cambridge, MA, USA: Cambridge University Press.
Pryke, J.S. and M.J. Samways (2009). ‘Recovery of invertebrate diversity
in a rehabilitated city landscape mosaic in the heart of a biodiversity
hotspot’. Landscape and Urban Planning 93: 54–62.
Rauzon, M.J., Forsell, D.J. and Flint, E.N. (2002). ‘Seabird re-colonisation
after cat eradication on Equatorial Jarvis, Howland, and Baker islands,
USA, Central Pacific’. In: C.R. Veitch and M.N. Clout (eds.) Turning
the tide: the eradication of invasive species, p. 411. Occasional Paper
SSC no. 28. IUCN SSC Invasive Species Specialist Group. IUCN,
Gland, Switzerland and Cambridge, UK.
Rejmanek, M. and Pitcairn, M.J. (2002). ‘When is eradication of exotic
pest plants a realistic goal?’ In: C.R. Veitch and M.N. Clout (eds.)
Turning the tide: the eradication of invasive species, pp. 249–253.
Occasional Paper SSC no. 28. IUCN SSC Invasive Species Specialist
Group. IUCN, Gland, Switzerland and Cambridge, UK.
Robertson, M.P., Harris, K.R., Coetzee, J.A., Foxcroft, L.C., DippenaarSchoeman, A.S. and van Rensburg, B.J. (2011). ‘Assessing local scale
impacts of Opuntia stricta (Cactaceae) invasion on beetle and spider
diversity in Kruger National Park, South Africa’. African Zoology 46:
205–223.
Samways, M.J. and N.J. Sharratt (2010). ‘Recovery of endemic
dragonflies after removal of invasive alien trees’. Conservation Biology
24: 267–277.
Samways, M.J., Sharratt, N.J. and Simaika, J.P. (2011). ‘Effect of
alien riparian vegetation and its removal on a highly endemic river
macroinvertebrate community’. Biological Invasions 13: 1305–1324.
Sax, D. (2002). ‘Equal diversity in disparate species assemblages: a
comparison of native and exotic woodlands in California’. Global
Ecology and Biogeography 11: 49–57.
Schirmel, J., Timler, L. and Buchholz, S. (2011). ‘Impact of the
invasive moss Campylopus introflexus on carabid beetles (Coleoptera:
Carabidae) and spiders (Araneae) in acidic coastal dunes at the southern
Baltic Sea’. Biological Invasions 13: 605–620.
Schoeman, C.S. (2008). ‘Synergistic Impact of Invasive Alien Plants and
the Alien Argentine Ant on Local Ant Assemblages in the Western Cape’.
MSc thesis. Stellenbosch, South Africa: University of Stellenbosch.
Simao, M.C.M., Flory, S.L. and Rudgers, J.A. (2010). ‘Experimental
plant invasion reduces arthropod abundance and richness across
multiple trophic levels’. Oikos 119: 1553–1562.
Simberloff, D. (2009). ‘We can eliminate invasions or live with them.
Successful management projects’. Biological Invasions 11: 149–157.
Spelman, L. 2012. National Geographic Animal Encyclopedia.
Washington, DC: National Geographic Society.
St John, M.G., Wall, D.H. and Hunt, H.W. (2006). ‘Are soil mite
assemblages structured by the identity of native and invasive alien
grasses’? Ecology 87: 1314–1324.
648
Strong, D.R., Lawton, J.H. and Southwood, R. (1984). Insects on Plants:
Community Patterns and Mechanisms. Cambridge, MA, USA: Harvard
University Press.
Tallamy, D.W. and Shropshire, K.J. (2009). ‘Ranking Lepidopteran use
of native versus introduced plants’. Conservation Biology 23: 941–947.
Talley, T.S., Nguyen, K.C. and Nguyen, A. (2012). ‘Testing the effects of
an introduced palm on a riparian invertebrate community in southern
California’. PLOS ONE 7(8): e42460.
Theel, H.J., Dibble, E.D. and Madsen, J.D. (2008). ‘Differential
influence of a monotypic and diverse native aquatic plant bed on a
macroinvertebrate assemblage: An experimental implication of exotic
plant induced habitat’. Hydrobiologia 600: 77–87.
Thomas, B.W. and Taylor, R.H. (2002). ‘A history of ground-based
rodent eradication techniques developed in New Zealand, 1959–1993’.
In: C.R. Veitch and M.N. Clout (eds.) Turning the tide: the eradication
of invasive species, pp. 301–310. Occasional Paper SSC no. 28. IUCN
SSC Invasive Species Specialist Group. IUCN, Gland, Switzerland and
Cambridge, UK.
Tilman, D. (1999). ‘The ecological consequences of changes in
biodiversity: A search for general principles’. Ecology 80: 1455–1474.
Triet, T., Man, L.C. and Nga, N.P. (2004). ‘Impacts of Mimosa pigra on
Native Plants and Soil Insect Communities in Tram Chim National
Park, Vietnam’. In: M. Julien, G. Flanagan, T. Heard, B. Hennecke, Q.
Paynter and C. Wilson (eds.) Research and Management of Mimosa
pigra, pp. 22–27. Canberra, Australia: CSIRO Entomology.
Ulyshen, M.D., Horn, S. and Hanula, J.L. (2010). ‘Response of beetles
(Coleoptera) at three heights to the experimental removal of an invasive
shrub, Chinese privet (Ligustrum sinense), from floodplain forests’.
Biological Invasions 12: 1573–1579.
University of Florida, Center for Aquatic and Invasive Plants. (2018).
‘Maintenance Control Strategy’. <https://plants.ifas.ufl.edu/manage/
developing-management-plans/maintenance-control-strategy/>.
Accessed May 2018.
Veitch, C.R. and Clout, M.N. (eds.) (2002). Turning the tide: the
eradication of invasive species. Occasional Paper SSC no. 28. IUCN
SSC Invasive Species Specialist Group. IUCN, Gland, Switzerland and
Cambridge, UK.
Veitch, C.R., Clout, M.N. and Towns, D.R. (eds.) (2011). Island invasives:
eradication and management. Proceedings of the International
Conference on Island Invasives. Gland, Switzerland: IUCN and
Auckland, New Zealand: CBB.
Viechtbauer, W. (2010). ‘Conducting meta-analyses in R with the metafor
package’. Journal of Statistical Software 36: 1–48.
Waltz, A.E.M. and Covington, W.W. (2004). ‘Ecological restoration
treatments increase butterfly richness and abundance: Mechanisms of
response’. Restoration Ecology 12: 85–96.
Webb, C.E., Oliver, I. and Pik, A.J. (2000). ‘Does coastal foredune
stabilization with Ammophila arenaria restore plant and arthropod
communities in southeastern Australia’? Restoration Ecology 8: 283–
288.
Weisser, W.W. and Siemann, E. (eds.) (2004). Insects and Ecosystem
Function. Berlin, Germany: Springer-Verlag.
White, E.M., Wilson, J.C. and Clarke, A.R. (2008). ‘Diversity and
abundance of arthropod floral visitor and herbivore assemblages on
exotic and native Senecio species’. Plant Protection Quarterly 23:
90–98.
Wilcove, D.S., Rothstein, D., Dubow, J., Phillips, A. and Losos, E.
(1998). ‘Quantifying threats to imperiled species in the United States’.
BioScience 48: 607–615.
Wotherspoon, S.H. and Wotherspoon, J.A. (2002). ‘The evolution and
execution of a plan for invasive weed eradication and control, Rangitoto
Island, Hauraki Gulf, New Zealand’. In: C.R. Veitch and M.N. Clout
(eds.) Turning the tide: the eradication of invasive species, pp. 381–388.
Occasional Paper SSC no. 28. IUCN SSC Invasive Species Specialist
Group. IUCN, Gland, Switzerland and Cambridge, UK.
Wu, Y.-T., Zhang, X.-D., Zhao, B., Jiang, L.-F., Chen, J.-K. and Li,
B. (2009). ‘Effects of saltmarsh invasion by Spartina alterniflora on
arthropod community structure and diets’. Biological Invasions 11:
635–649.
Zavaleta, E.S., Hobbs, R.J. and Mooney, H.A. (2001). ‘Viewing invasive
species removal in a whole-ecosystem context’. Trends in Ecology and
Evolution 16: 454–459.
Zuefle, M.E., Brown, W.P. and Tallamy, D.W. (2008). ‘Effects of nonnative plants on the native insect community of Delaware’. Biological
Invasions 10: 1159–1169.
Reference
Ando, et al., 2010
Bailey, et al., 2001
Bartomeus, et al., 2008
Bassett, et al., 2012
Bickel & Closs, 2009
Bock, et al., 1986
Brandle, et al., 2008
Burghardt, et al., 2009
Cameron & Spencer, 2010
Chey, et al., 1998
Christopher & Cameron, 2012
Cord, 2011
de Groot, et al., 2007
Durst, et al., 2008
Florens, et al., 2010
Gerber, et al., 2008
Gossner & Ammer, 2006
Gremmen, et al., 1998
Hagen, et al., 2010
Hanula & Horn, 2011a
Hanula & Horn, 2011b
Harris, et al., 2004
Hartley, et al., 2010
Harvey, et al., 2010
Herrera & Dudley, 2003
Hills, et al., 2008
Holmquist, et al., 2011
Location
Central Japan
Arizona, USA
Spain
New Zealand
New Zealand
Arizona, USA
Germany
Pennsylvania, USA
Texas, USA
Sabah, Borneo
Ohio, USA
Texas, USA
Slovenia
Arizona, USA
Mauritius
Switzerland, Germany, & France
Germany
Marion Island, SubAntarctic
Robinson Crusoe Island, Chile
Georgia, USA
Georgia, USA
New Zealand
Texas, USA
Australia
California, USA
Australia
California, USA
Latitude
35.07
34.67
42.32
35.02
45.03
31.65
51
40.25
29.53
5.42
39.12
27.49
46.05
33.65
20.4
47
48.18
46.83
33.63
33.88
33.88
41.2
29.53
34
38.23
33.82
36.45
Native Habitat
Experimental forest field
Riparian woodland
Mediterranean shrubland
Lake margin
Littoral
Semidesert grassland
Multiple
Suburban residences
Coastal prairie
Tropical rainforest
Hardwood forest
Grassland
Agricultural fields & ruderal areas
Floodplain
“Indigenous forest”
Grassland, scrub
Spruce forest
“Drainage lines”
Lower montane forest
Riparian hardwood forest
Riparian hardwood forest
Kanuka scrub
Tree plantations
Coastal salt marsh
Riparian forest
Cave trees
Desert spring
X
X
X
X
X
X
X
X
X
X
X
X
X
Time?
X
X
X
X
X
X
X
X
X
1
X
X
X
X
X
X
X
X
X
X
X
X
Cov?
restored
restored
restored
restored
Control
#Exotics
1
1
2
1
1
1
1
1
1
5
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
1
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Richn.
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Div.
Appendix 1 Studies used in the meta-analysis and their attributes. “Time?” indicates whether or not time since establishment was reported in the reference, and “Cov?” indicates
whether or not cover of the invader was reported. “Control” indicates if the native comparison included restored habitat. “#Exotics” indicates the number of different exotic plant
species that were included in the study. “Richn.” Indicates if the study evaluated invertebrate species richness, while “Div.” indicates if the study evaluated invertebrate diversity.
Knapp, et al.: Controlling invasive plants
649
650
Location
Rodrigues Island, SW Indian Ocean
Germany
South Africa
South Africa
Hungary
Poland
Australia
Australia
Portugal
South Africa
South Africa
South Africa
South Africa
California, USA
Germany
South Africa
Indiana, USA
Kansas, USA
Eastern USA
California, USA
Mississippi, USA
Vietnam
Georgia, USA
Australia
Australia
China
Delaware, USA
Reference
Hugel, 2012
Kappes, et al., 2007
Magoba & Samways, 2010
Magoba & Samways, 2012
Magura, et al., 2000
Moron, et al., 2009
Osunkoya, et al., 2011
Parr, et al., 2010
Pinto, et al., 1997
Pryke & Samways, 2009
Robertson, et al., 2011
Samways & Sharrat, 2010
Samways, et al., 2011
Sax, 2002
Schirmel, et al., 2011
Schoeman, 2008
Simao, et al., 2010
St John, et al., 2006
Tallamy & Shropshire, 2009
Talley, et al., 2012
Theel, et al., 2008
Triet, et al., 2004
Ulyshen, et al., 2010
Webb, et al., 2000
White, et al., 2008
Wu, et al., 2009
Zuefle, et al., 2008
33.55
33.3
37.88
54.53
34.05
39.22
39.1
36.5 to 45
32.75
33
10.7
33.88
35.4
27.83
31.52
39.7
25
Latitude
19.72
51.15
23.02
18.9
48.47
50.05
27.83
12.72
40.28
33.95
Native Habitat
Time?
Tropical forest
Floodplains
Riparian
Fynbos scrub
Oak-hornbeam forest
X
Wet meadow
X
Eucalyptus & subtropical rainforest
X
Mesic eucalyptus savanna
X
Riparian
Southern afrotemperate forest,
fynbos scrub
Savanna, Sabie-crocodile thorn
X
thicket
Riparian
Riparian, fynbos scrub
Oak woodland
X
Coastal dunes within heath
X
Fynbos scrub, renosterveld
X
Experimental
X
Prairie
X
Multiple
Riparian woodland
Aquatic
Seasonally inundated grassland
X
Floodplain forest
X
Coastal foredunes
X
Pasture & grassland
X
Salt marsh
X
Common garden
X
X
X
X
X
X
X
X
X
X
X
X
Cov?
restored
restored
restored
restored
restored
Control
restored
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
#Exotics
1
1
1
1
1
1
1
1
2
1
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Richn.
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Div.
Appendix 1 (Cont’d) Studies used in the meta-analysis and their attributes. “Time?” indicates whether or not time since establishment was reported in the reference, and “Cov?”
indicates whether or not cover of the invader was reported. “Control” indicates if the native comparison included restored habitat. “#Exotics” indicates the number of different exotic
plant species that were included in the study. “Richn.” Indicates if the study evaluated invertebrate species richness, while “Div.” indicates if the study evaluated invertebrate diversity.
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
X. Lambin, J.C. Horrill and R. Raynor
Lambin, X.; J.C. Horrill and R. Raynor. Achieving large scale, long-term invasive American mink control in northern Scotland despite short term funding
Achieving large scale, long-term invasive American mink control
in northern Scotland despite short term funding
X. Lambin1, J.C. Horrill2 and R. Raynor3
School of Biological Sciences, University of Aberdeen, Tillydrone Avenue, Aberdeen AB24 2TZ, UK. <x.lambin@
abdn.ac.uk>. 2Rivers and Fisheries Trusts of Scotland, 11 Rutland Square, Edinburgh EH1 2AS, UK. 3Scottish Natural
Heritage, Great Glen House, Leachkin Road, Inverness IV3 8NW, UK.
1
Abstract The American mink (Neovison vison) has invaded most of the United Kingdom following escapes from furfarms over decades. Its escalating impact on riparian and coastal biodiversity, including seabirds and water voles, is well
documented. Starting in north-east Scotland in 2004, long-term, multi-institution mink control efforts have harnessed
the enthusiasm of volunteer conservationists to push back the mink invasion over a vast area. Rather than the outcome
of a single project with secured long-term funding, this achievement resulted from four successive joined up projects
each with short-term funding. The beginnings of the project (2004–2006), under the auspices of the north-east Scotland
Biodiversity Partnership were small scale (30 km2) and centred upon a lowland remnant water vole meta-population.
Mink control efforts were scaled-up to 6,000 km2 of mostly marginal mink habitat as part of the Cairngorms Water Vole
Conservation Project (2006–2009) centred on the newly established Cairngorms National Park. The project, led by the
University of Aberdeen, was funded by a charity, a UK Research council and Scottish Natural Heritage and involved the
national park authority, and three local fisheries trusts. The approach was to deploy a “rolling carpet” of mink control
based on the use of mink rafts operated by volunteers and that facilitated mink detection and removal. Substantial funding
was then secured for a successor project, the Scottish Mink Initiative (2010–2014) involving, all previous partners plus
14 local fisheries trusts coordinated by the Rivers and Fisheries Trusts of Scotland. Mink were pushed back over a vast
area (29,000 km2) and their spread in coastal areas of north-west Scotland was countered. After a period with minimal
bridge funding, coordinated mink control efforts resumed, thanks to the newly funded Scottish Invasive Species Initiative
(2017–2021) seeking to extend the approach used with mink to other riparian invasives. Mink remain scarce or absent and
water voles are recovering spectacularly. Coordinated mink control delivered tangible conservation benefits and improved
understanding of the socio-ecological system despite the challenges of short-term funding.
Keywords: adaptive management, American mink, Arvicola, Neovison vison, participation, Scotland, water vole
INTRODUCTION
While there have been enormous achievements and
improvements in the eradication of a small number of
invasive mammalian species (brown (Rattus norvegicus)
black (R. rattus) and Pacific (R. exulans) rats, house mice
(Mus musculus), rabbits (Oryctolagus cuniculus), feral cats
(Felis catus)), on islands of increasing size (DIISE, 2015),
there has been comparatively little progress with efforts and
guidelines on how to durably control invasive species in
those areas where eradication is presently an unattainable
goal. Yet, prevention has failed in many areas, such that
focussing invasive management efforts exclusively on
islands where eradication can be achieved leaves much
valued biodiversity impacted by invasive species. Thus,
when considering whether to expand resources to protect
native biodiversity against the impact of invasive species, a
key unknown is what, if anything short of eradication, can
be achieved cost effectively and what management regimes
might be both ecologically effective and sustainable over
the long term.
Eradication can only be achieved where immigration
can be prevented or managed (Bomford & O’Brien, 1995).
Where this condition is not met, as is the case on continental
mainland and large island areas, control of invasives must
be the management objective. New Zealand’s so-called
‘mainland islands’ are areas where intensive conservation
adaptive and integrated pest management regimes are
applied and outcomes are closely monitored (Saunders
& Norton, 2001). They are adjacent to other areas where
invasives are not managed to the same extent, hence
subjected to immigration that, if not dealt with, could lead
to recolonisation.
A key feature of mainland islands is that conservation
management must be designed so as to last in perpetuity to
ensure that the biodiversity and socio-community gains are
not lost. It is therefore especially crucial that siting considers
all features that may make a mainland island defensible.
This may include topography (e.g. presence of peninsulas),
ecological gradients or socio-economic interest of the local
community that may affect their willingness to participate
in ongoing management and adopt biosecurity measures
and even the erection of conservation fences (Glen, et al.,
2013). An unavoidable corollary of planning for the very
long term, is the need for long-term funding commitments.
This is crucial to negate the risk that ecosystem restoration
will one day be undone should a lack of resources preclude
a rapid and decisive reaction following incursion by
invasives into a mainland island. In this respect, the fact
that New Zealand’s mainland islands are operated by
the Department of Conservation, a government agency,
provides a degree of continuity lacking elsewhere.
Owing to a lack of reported successful instances of
control of invasive species in mainland areas, and to a
few well publicised failures (e.g. Sheail, 2003; Santulli,
et al., 2014), managers have little guidance as to the
circumstances under which a mainland island approach
might prove successful. Of particular interest is how
complex institutional and funding environments need to be
navigated when planning long term control of invasives. In
the UK, for instance, protected areas are largely privatelyowned; conservation legislation incentivises rather than
mandates conservation management activities; a significant
proportion of conservation action is initiated in a bottom
up fashion by non-governmental organisation or local
communities (often enabled by government agencies); and
funding for projects rarely exceeds 3–5 years in duration.
In this paper, we present an account of the development
of a mainland island invasive control effort that grew in
spatial extent over 15 years from a localised communityled effort to operate on a vast scale (29,000 km2) in the north
of Scotland. It progressed from pilot, to demonstration
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 651–657. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
651
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
and, eventually, mainstreaming stages without secured
long-term funding but as an enduring partnership between
academic researchers and practitioners under an adaptive
management framework.
Invasive American mink threatening Ratty the water
vole, a British cultural icon
The initial motivation for the project was the protection
of the water vole (Arvicola amphibius), riparian rodents that
used to be very abundant in Britain but that experienced a
cumulative mean loss of occupied sites of 98.7% across
all regions of England, Scotland, and Wales by 1998 from
the 1939 baseline (Moorhouse, et al., 2015). Thus, the
water vole was included amongst Species Action Plans and
devolved Local Biodiversity Action Plans (LBAPs) when
the UK Government launched those plans for the recovery
of threatened species and habitats as part of the UK
Biodiversity Action Plan in response to the Convention on
Biological Diversity in 1994 (UK Biodiversity Partnership,
1995; JNCC, 2006). One of several suggested causes for
the catastrophic decline of the water vole was the American
mink (Neovison vison) that had invaded all but the northwesternmost corner of the UK following historical escapes
from fur farms (Fraser, et al., 2015). Its overriding influence
became clearer over time (Aars, et al., 2001; Moorhouse, et
al., 2015). Accordingly, LBAPs included controlling mink,
but little guidance or prescriptions on how this should be
implemented were included.
SCALING UP MINK CONTROL: FOUR PHASES
OF SPATIAL EXPANSION
Water voles in the catchment of the River Ythan
(1995–2007): 100–644 km2
Research into metapopulation processes by ecologists
from the University of Aberdeen funded by the Natural
Environment Research Council (NERC) (1995) identified
a handful of highly fragmented remnant water vole
populations in a 100 km2 portion of an intensely farmed
lowland area north of Aberdeen in NE Scotland (Telfer, et
al., 2001) (Fig. 1). Structured surveys revealed that water
voles had become regionally scarce or absent where they
were once common (Lambin, et al., 1996; Lambin, et
al., 1998; Lambin, et al., 2002). The intensively studied
metapopulation network was gradually shrinking under
the influence of American mink predation, causing the
extinction of multiple adjacent colonies (Lambin, et al.,
1996; Telfer, et al., 2001) (Table 1).
With funding secured by north-east Scotland’s Local
Biodiversity Action Plan group from Scottish Natural
Heritage (SNH), the government agency tasked with
promoting, caring for and improving Scotland’s natural
heritage (£145,000 over eight years, Fig. 2), the first stage
of the northern Scotland control mink project was initiated
in 2002. Its modest objective was to safeguard the remnant
lowland water vole metapopulations by preventing further
encroachment by mink. Initially, a member of staff from
the local Ythan District Fishery Board, a statutory body
empowered to protect, enhance and conserve Atlantic
salmon and sea trout within the Ythan catchment, was
employed on a part time basis to control mink (2002–
2003). Subsequently, a full-time member of staff, employed
by the University of Aberdeen (UoA), was appointed
over five consecutive one-year contracts (2003–2007) as
mink control activities were extended to the entire 644
km2 area of the catchment of the river Ythan as evidence
accumulated that it was possible to protect remnant water
vole colonies from encroachment by mink (Fig. 1).
This early step was arguably an instance of last ditch
conservation, focussed on safeguarding a fast-shrinking
isolated remnant water vole metapopulation. It was
nevertheless influential in shaping ways of working that
became crucial as the project area was expanded 45-fold
over the next 10 years.
Key features were:
i) Close links between research on water vole and mink
population dynamics and conservation delivery;
ii) Systematic deployment of mink rafts that make it
possible to detect the presence of mink and to target
cage trapping to those sections of waterways where
current mink presence is confirmed (Reynolds, et al.,
2004);
iii) Involvement of local residents who were encouraged
to volunteer to monitor and report the appearance of
signs of mink on mink rafts in their neighbourhoods,
allowing a single project officer to effectively
control mink of an entire catchment through targeted
trapping.
Fig. 1 Map of northern Scotland showing the five stages
of expansion of successive mink control projects from
a sub-catchment of the River Ythan (thick black line,
numbered 1), the entire catchment of the River Ythan
(Grey dashed lined, numbered 2), the Cairngorms
National Park (thin black line, numbered 3), the area
of the expanded Cairngorms project (dashed black
line, numbered 4) and the area where the Scottish
Mink Initiative operated (Continuous thick black line,
numbered 5).
652
Fig. 2 Annualised expenditure of all projects relevant to
water vole conservation and mink control broken down
as funding for enabling or evaluating research (white
bars) or conservation delivery (black bars).
Lambin, et al.: Long-term mink control in Scotland
iv) Partnership with organisations tasked with the
management, conservation and enhancement of
native freshwater fish and their environments in
Scotland and increasingly involved in invasive
species management.
In 2009, the river trust in the adjoining catchment of the
River Deveron, emulated the project and obtained funding
from SNH for an integrated package of invasives control,
including American mink. The likely disappearance of
the Ythan water vole population was averted, and this
population is now thriving and extends across the entire
lowland NE Aberdeenshire plain (W Morgan, E McHenry,
X Lambin unpublished data).
The Cairngorms Water Vole Conservation Project
(2007–2009): 5,500–10,570 km2
Further surveys of water voles in the uplands of
NE Scotland commissioned by SNH and research into
metapopulation genetics processes by UoA (1998–2000)
uncovered large water vole metapopulation networks in
the area that was to become the Cairngorms National Park
(CNP) in 2003 (Aars, et al., 2001; Lambin, et al., 1998;
WildCRU, 2004) (Table 1). These populations, while in
slow decline, had not yet been affected by the American
mink invasion to the same extent as lowland populations,
owing to the low density of alternative prey for mink
in the uplands (Oliver, et al., 2009). They presented the
opportunity to preserve functioning metapopulations
and the associated ecosystem functions arising from the
ecosystem engineering activities of water voles on upland
riparian vegetation (Bryce, et al., 2013) as opposed to
the more desperate task of rescuing critically endangered
survivors.
The CNP encompasses a mountain massif, dominated
by heather moorland where shooting of red deer (Cervus
elaphus), red grouse (Lagopus lagopus) and fly fishing of
salmon (Salmo salar) provide much needed income to the
rural economy. In order to make these leisure activities
possible, a large number of game keepers and fishing
ghillies are employed to intensively manage heather
moorland through rotational burning, killing predators
of grouse and accompanying anglers. These individuals
were recognised as a potential trained workforce that
already culled ~ 60–70 mink annually in CNP, hence had
the expertise and a professional interest in the issue. Their
willingness to step up and coordinate hitherto patchy mink
control was ascertained through consultation funded by the
newly established CNP in 2004. Thus, we reasoned that the
CNP was a potential defensible mainland island stronghold
for water voles where mink control could be sustained in
perpetuity.
Funding bids to the Tubney Charitable Trust, a
charitable organisation for projects that conserve the
natural environment in the UK, and to the NERC outlining
the ambition to implement a formal active adaptive
management approach to control American mink on the
large scale of CNP (encompassing 5,500km2, Fig. 1) were
prepared. SNH had again committed match funding should
either bid succeed. The Cairngorms National Park Authority
(CNPA) and three river trusts managing important salmon
rivers flowing from the Cairngorms (River Dee Trust, Spey
Foundation and Deveron, Bogie & Isla Rivers Charitable
Trust) were also formal partners committing in-kind staff
time. Both bids were funded and substantial funding
was in place for three years (2006–2009), facilitating the
employment of three project officers and one postdoctoral
research fellow by UoA.
A detailed account of the project’s approach and
achievements is given in Bryce, et al. (2011) and Oliver,
et al. (2016) and a brief summary only is given here. The
approach was to deploy mink rafts with an approximate
spacing of 2 km in a ‘rolling carpet’ fashion to first remove
mink from upland areas and subsequently expand coverage
Table 1 Sequence and main findings of research at the University of Aberdeen that enabled the next step of mink control
efforts by characterising the system to be managed, that evaluated the achievements of mink control efforts or that
provided a strategic evaluation of different ways or working.
Research Project
S. Telfer PhD
Years
1996–
1999
Scope
Enabling
Main finding
Water voles metapopulation processes are disrupted by
mink causing spatially correlated colony extinction
Funder
UK Research
Council
J. Luque Larena
Postdoc fellowship
2003–
2004
Enabling
Cairngorms Mountains are invaded by mink owing to
presence of rabbits in abandoned hill farms
European
Union
A. Zalewski
Postdoc fellowship
2005–
2006
Enabling
Cairngorms Mountains are a partial obstacle to mink
dispersal but mink circumvent hills and nevertheless
spread
European
Union
M.K. Oliver
R. Bryce
Postdoc fellowships
E. Fraser PhD
2006–
2009
Evaluation
UK Research
Council
2010–
2013
Enabling
Strong lowland–highland source–sink dynamics and
high mobility between catchments influencing capture
rates
Mink spread in sparsely populated coastal areas is
heavily constrained by topography and boat-based
ecotourism operators are potential volunteers
M.K. Oliver Postdoc
fellowship
2010
Evaluation
Mink control reduces captures to almost zero in three
years. Mink dispersal large-scale (31 km for females),
male biased, and links adjacent river catchments
UK Research
Council
Y. Melero
Postdoc fellowship
2011–
2014
Evaluation
No evidence of mating failure at low density causing
Allee effect but instead compensatory increase in
fecundity at low density
European
Union
E. McHenry PhD
2014–
2018
Strategic
Doing more with less: optimising investment in
detection and control
UK Charity
W. Morgan PhD
2014–
2018
Evaluation
Patterns of recovery in water voles
UoA
SNH
653
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
downstream to remove mink from an increasingly large
area, hence protecting the uplands with increasing depth;
a version of the ‘remove and protect model’ with depth
(Bell, et al., 2019). The systematic use of mink rafts was
made possible by the work of 208 volunteers. We sought
volunteers willing to adopt a mink raft and report to a
project officer or trapper in their local community whether
a mink was present. Only when fresh mink signs were
detected was a cage live trap set, hence minimising the time
wasted checking empty traps at least once every 24 hours
as mandated by law. If a mink was caught (as occurred
following 10–22% of detections according to season),
it was humanely killed before the raft was returned to
monitoring mode. The project officer played a crucial role
in coordinating the efforts of volunteers, not all of whom
were equipped or qualified to humanely despatch a mink.
On detection of the presence of a mink by a volunteer,
the full capabilities of the larger volunteer force could be
called upon to effectively trap and despatch the mink.
Project officers sought permission to access the land
and deploy a mink raft and then recruited local volunteers
to operate the raft. Game-keepers who are licenced to carry
fire-arms were partners of choice to adopt and operate mink
rafts, although it proved difficult to dissuade them from
their traditional practice of deploying traps irrespective of
evidence of the presence of the focal species (Fig. 3). Two
of three project officers had prior family or professional
associations with the local game keeping community
and this undoubtedly facilitated building constructive
relationships. The adoption of rafts by local residents was
key to allowing project officers to deploy further rafts
downstream in the more biologically productive parts of
the CNP and where landownership is more fragmented and
residents with a wider diversity of professions live. Here,
we adopted a functional approach to participation (Pretty,
1995) whereby local people were co-opted to meet the
predetermined objectives of achieving coordinated mink
control. Thus, recruitment of volunteers to operate rafts
was targeted toward individuals with an interest in nature
conservation and natural resource management, such as
forest or local government rangers, fishing ghillies, bailiffs,
nature reserve managers, but also included numerous local
residents made aware of opportunities to become involved
in the project through community talks and publication in
the local press. Where required, project officers would check
mink traps or despatch mink themselves but volunteers
were always encouraged to step up their involvement from
monitoring rafts only, to trapping or becoming a trained
despatcher.
The large project area was subdivided into subcatchment management units encompassing major
tributaries of main catchments (median size: 55 km2).
Analyses of the impact of culling on the population used,
as a reference point, the time when mink raft deployment
was deemed complete in a sub-catchment by the local
project officer. The number of mink captured per km of
waterway decreased from an average of 0.16 to 0.06 to 0.01
for sub-catchments in the first, second and third years after
inception of control, respectively. This was despite higher
fecundity amongst mink that had survived culling (Melero,
et al., 2015; Oliver, et al., 2016). Most mink caught in the
third year after inception of control were males, reflecting
their high propensity to disperse from the natal area. This
was also reflected in the high proportion of juvenile males
amongst the few mink caught in the higher elevations of
the CNP which were cleared of mink by the end of 2007.
No mink at all were caught in 3,417 km2 of montane and
moorlands areas of CNP but 376 mink were removed
from 5,381 km2 covering moorland and pastoral areas of
lower altitude. There was further evidence of high mink
mobility within and between river catchments resulting in
compensatory immigration, as mink capture rate in a subcatchment increased with connectivity to mink still present
in other sub-catchments (Bryce, et al., 2011; Oliver, et al.,
2016).
The key lessons from the ongoing evaluation of
management efforts were:
i) The presence of large-scale lowland-highland sourcesink dynamics in mink such that most mink impacting
upland biodiversity had dispersed from more
productive lowland areas. This motivated a change
in the scope of the project when the management
group endorsed downstream expansion from 5,500
to 10,570 km2 at the end of the second year of the
project (2007) so as to deplete mink where most were
born (Fig. 1).
ii) Deploying a large number of mink rafts and
recruiting volunteers is a gradual process and a
pool of volunteers must be replenished to make up
for volunteer turn-over (Beirne & Lambin, 2013).
Different communities and river trusts vary in their
ability to embrace conservation volunteering and the
resulting asynchrony in the inception of mink control
delayed region-wide eradication.
iii) Mink disperse widely and dispersal connects major
river catchments, implying an inter-dependence
between river catchments and the organisations that
manage them. Thus, high mobility of mink dictates
that control should be on a very large scale so as to
avoid the effects of compensatory immigration.
The Scottish Mink Initiative (2011–2015): 10,570
–29,000 km2
Fig. 3 Temporal dynamics of the number of mink caught
per year (black line, black circles), the number of mink
rafts deployed (grey line, grey circles) and the number of
volunteers contributing to the projects (black line, white
circles).
654
The achievement of the Cairngorms Water Vole
Conservation Project elicited much enthusiasm from
volunteers who had been part of a rare conservation good
news story, as well as from private and public land managers
(e.g. CNPA) and Scottish Natural Heritage. As the threeyear funding period was coming to an end, there was a
real risk that the project would fall from a funding cliff
edge such that not only would all biological gains be lost
but the volunteer community would become despondent if
abandoned. SNH had also been working with the Scottish
Wildlife Trust (SWT) and local fisheries’ trusts in the northwest Highlands to remove mink in that area, so there was an
opportunity to develop a more strategic approach to mink
control across the north of Scotland by amalgamating and
expanding the various projects into a single, much larger
Lambin, et al.: Long-term mink control in Scotland
scheme. SNH, along with two other key funders (CNPA and
the Tubney Trust) expressed their willingness to renew their
funding commitments for a further three years (£478,000;
£8,932; £100,000, respectively). However, the partnership
research grant scheme run by the UK research council
had been discontinued and funding commitments did not
include the overheads universities expect from research
grants. This made it impossible for UoA to continue as
the lead partner of what was increasingly an ambitious
conservation delivery project rather than a combination of
this and research. Furthermore, it was evident that local
organisations managing common natural resources and
representing private entities gaining economic benefits
from harvesting salmon would be more appropriate
long-term custodians of a mainland island project than a
university and thereby ensure it had a long-term legacy.
Accordingly, a new partnership was formed involving
Rivers and Fisheries Trusts of Scotland (RAFTS) and SWT.
RAFTS was a charity with a formal objective comprising
“the conservation and enhancement of native freshwater
fish and their environments in Scotland”. Twenty-six river
trusts and foundations were members of RAFTS and it was
already actively involved in (mostly riparian plant) invasive
management. It had a strong track record in fundraising
and project management for its members. It proved to be
the ideal body to lead an expanded project and to ensure
coordinated action using best practice by its member
river trusts at a scale commensurate with the biological
challenge posed by mink. Nine river trusts in northern and
north-east Scotland were enlisted in a new partnership and
they committed in kind resources to removing mink from
their river catchments. The renewed funding commitments
were critical in allowing an application to the EU-funded
LEADER scheme operated by the Scottish Government.
The aim of LEADER is to increase support to local rural
community and business networks to build knowledge and
skills, and encourage innovation and cooperation, in order
to tackle local development objectives. A competitive
application involving multiple local areas was assembled
and further funds (£229,000 from LEADER, and £14,000
from river trusts) were secured, facilitating the appointment
of three project officers and a coordinator employed
by RAFTS. For the second time, mink control efforts in
northern Scotland bounced back from a financial cliff edge.
Owing to the time required for the evaluation of
the funding bid and recruiting new project staff, mink
volunteers had been left without support or certainty on
the future of mink control efforts during the 19-month gap
that elapsed between the end of the Cairngorms project
in October 2009 and the start of the new Scottish Mink
Initiative (SMI) in April 2011. Over that period, a skeleton
staff was retained from previous projects to maintain the
volunteer and associated mink raft network prior to further
expansion (Raynor, et al., 2016). This included one parttime member of staff from the north-west Highlands
project. It had adopted a “cordon sanitaire” approach,
comprising a double line of mink rafts intended to prevent
mink from invading northern Scotland, following from
recommendations in an unpublished report to SNH
(Harrington, et al., 2008). That approach turned out to be
flawed owing to mink dispersal abilities, evident in data
collected as part of the Cairngorms project but that were
unpublished at that time (Oliver, et al., 2016), and to the
importance of the coastal environment in driving invasion
range expansion (Fraser, et al., 2015).
Four newly appointed SMI staff had to be trained and
build new trust relationships with volunteers previously
supported by other staff. While some volunteers had
continued with their activities in the intervening time and
caught a minimum of 139 mink in 2010, many no doubt
concluded that the project had come to an abrupt end and
ceased their activities. This led to reinvasion of some of
the project areas, especially in the vicinity of the crucial
catchment of the River Don where inadequate local
support had prevented progress with mink control as part
of the Cairngorms project (contrast figure 2 in Oliver, et al.,
(2016) and figure 3 in Melero, et al., (2015)).
Once the full complement of project officers was
again embedded in the local community and supported
by local river trusts, the approach refined in the previous
project was scaled up substantially by SMI resulting in
837 volunteers operating up to 1022 rafts and removing
a minimum of 646 mink between 2011 and 2014. This
resulted in a vast area encompassing ~29,000 km2 bounded
by seas becoming free of breeding mink as determined by
the absence of footprints on mink detection rafts, the metric
chosen by the steering group to gauge the effectiveness of
the project (Fig. 1), hence increasing our ability to deal
with the constant flux of mink moving up from the south.
Mink were regularly detected in the southern and western
edges of the project area (51 in 2014) especially, reflecting
primarily immigration by males during the rut period. A
more detailed account of its achievements and of some
of the challenges encountered is found in Raynor, et al.
(2016).
The Transition to Scotland’s Invasive species Initiative
(2018-2022): 29,500 km2
One ultimate objective of SMI was to engender a
sense of ownership of the mink management and wider
biosecurity, considering the threat posed by aquatic
invasives such as the salmon fluke (Gyrodactylus salaris)
and the giant hogweed (Heracleum mantegazzianum),
amongst the local fisheries trusts as appropriate to any
mainland island project. There was also an aspiration to
further build on the partnership by involving more and
more trusts, as resources allowed. Thus, during the second
half of the funded period (September 2013–August 2015),
there was a process of hand-over of all local processes to
10 local rivers and fisheries trusts. This included transfer of
responsibility for managing existing networks of volunteers
and mink rafts, including all access agreements with land
owners, health and safety and standard operating protocols,
and all relevant databases. A project coordinator remained
employed by RAFTS and each participating river trust
received payment to cover costs incurred in undertaking
a combination of mink raft checking and maintenance, as
well as data collection and support and coordination for the
local volunteer network.
Two main limitations to the effectiveness of the
handover have been: 1) not all areas of high mink
productivity on the lowland coastal plain in the extreme
corner of NE Scotland have sufficient salmon resources
to maintain functional river trusts. Without additional
resources, such areas could again become a source of
dispersing mink into adjacent better controlled areas; 2)
maintaining mink raft coverage in remote areas of northwest Scotland, where the low human population density;
a predominance of red deer over grouse as the primary
game species; difficult topography including many coastal
islands; and a limited road network all placed significant
restrictions on the ability to maintain required coverage
for surveillance. The handover arrangements have been
severely tested, with mixed results, by the absence of any
financial support to any of the trusts between August 2015
and November 2017. During this period, a major reform
of freshwater fisheries governance that would have led to
river trusts and boards being disbanded was mooted by the
Scottish Government and this precluded the submission
of grant applications for the successor project by RAFTS.
655
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
The proposed reform was ultimately abandoned but led to
the demise of RAFTS as an organisation. Scottish Natural
Heritage, a key long-term supporter of the project from its
very outset, stepped in as lead partner for an application
to the Heritage Lottery Fund and an award of £1.59M was
announced in August 2017. Thus, after a protracted period
without secure funding, a successor to SMI, centred on
applying the citizen conservationist approach to a suite of
riparian invasives and prepared by RAFTS, will operate
from 2018–2022. The new project, the Scottish Invasive
Species Initiative, will tackle the challenge of reviving
the volunteer network and undoing unavoidable partial
reinvasion of the project area for another four years and
further increase engagement in invasive management by
local communities (Horrill, et al., 2019).
DISCUSSION
Over 15 years, a vast mainland island area has been
established in northern Scotland that protects native
riparian biodiversity including water voles from the
destructive influence of the invasive American mink. The
endeavour is the outcome of a succession of research and
implementation projects conducted in partnership that
optimised mink control effort so they could be scaled-up.
Implementation projects progressed from a small-scale
pilot phase (in the Ythan), to a two-stage demonstration
phase, first evidencing the feasibility of scaling up mink
raft deployment and enthusing volunteers to become
citizen conservationists (the Cairngorms project) and then,
the feasibility of devolving management of such a large
scale project to local organisations engaged in natural
resource management (the SMI) according to a wider,
more strategic framework. The later stage of SMI was the
beginning of embedding mink control within the activities
of rivers trusts working autonomously but in a coordinated
manner. The most recently funded successor project has
the ambition to extend the approach refined with mink
to a suite of containable riparian plant invasives that are
widespread in Scotland.
Long-term invasive species management was achieved
despite short-term funding as a result of a succession of
fixed-length short-term discrete projects each of three to
four years duration, rather than the result of any integrated
long-term joined-up endeavour underpinned by secured
funding or any strategic decision on the size of any area
where mink could be controlled on Scotland’s mainland.
As the feasibility of controlling mink on a large scale
was demonstrated and the endeavour’s spatial ambition
grew, the very existence of the project was in jeopardy
on multiple occasions and some of its achievements were
eroded during four gaps between funding cycles. Its future
is secured for another four years after the latest two and
a half year funding gap since the end of SMI. Although
the large spatial reach of the project, its cost effectiveness
and hence attractiveness, results from the use of volunteer
citizen conservationists, the lack of continuity in funding
has been highly detrimental to the trust relationship built
between the project and volunteers giving their time freely
for conservation. Invasive species control in mainland
areas is, by definition, an open-ended commitment and it
is paramount the limited resourcing required to maintain
what has been achieved should be in place conditionally
on evidence of success and sustainability being presented.
The cumulative cost of all components of the project,
including the research by EU-funded fellows and four PhDs
that enabled the project or contributed to its evaluation
under the adaptive management, was £2,800,000. The
cost-effectiveness of the project resulted from the use
of a workforce of 866 unpaid “citizen-conservationist”
volunteers. Based on the assumption that their time
656
contribution amounted to 30 min/2weeks = 13 hours
per year per volunteer, the total 2,652 volunteer years is
equivalent to 21.6 standard person years, crudely valued
at £1,404,00 using the assumptions of Robertson, et al.
(2019). Arguably, the value of their contribution is greater
still because of the increased awareness of the issues
caused by invasives and community cohesion benefits
(Evely, et al., 2011).
Although the volunteer approach is relatively cheap, it
is not cost-free as volunteers require a degree of support,
encouragement, information and re-supplying by project
staff. The successive incarnations of the mainland mink
control efforts have involved an increasing number of
volunteers (peaking at 612 in 2014 Fig. 3) supported by
a fixed and small number of project officers. Volunteer
retention over time is less than 100 % such that it is
constantly necessary to recruit new volunteers. Despite
project staff consistently reinforcing the message that
“no mink is good news”, it remains that the enduring
absence of mink on a volunteer’s raft contributes to some
volunteers dropping out (Beirne & Lambin, 2013). The
greatest risk causing volunteer drop-out is the perception
that the project has come to an untimely end in the absence
of communication from project staff, as arose during the
funding gaps, even if efforts to fund-raise for a successor
project are underway.
SNH, Scotland’s governmental organisation responsible
for the management of natural heritage including the threat
posed by invasive species, has been an enduring and crucial
funder at all stages of mainland mink endeavour ever since
1995. It contributed 45 % of the total £2,803,950 cash cost
over 21 years and 62% of the subset (£1,900,000) spent
on conservation delivery. SNH is also the main funder of
the Hebridean Mink Project (Macleod, et al., 2019), hence
is committing substantial resources to managing American
mink. However, as with all government agencies,
including in New Zealand and the USA, it is constrained
by its inability to commit long-term funding for managing
established invasive species. Even SNH’s Species
Action Framework scheme that made sizeable financial
contributions to SMI (£710,000 including extensions) was
a five-year programme of targeted species management.
Furthermore, contributions from SNH were conditional on
funding being secured from other funders. Fund-raising by
UoA and RAFTS was successful but time-consuming and
added complexity to project management and reporting. It
is a major concern that given EU funds covered 20 % of
total costs and provided for 40% of the research work, the
departure of the UK from the EU in 2019 will potentially
leave a major hole in funding.
Through all phases of the project, the programmes of
research that enabled and evaluated the development of
large scale invasive control were always funded by separate
funding streams to those used for conservation delivery
(such as species recovery or habitat management). This
was in response to implicit or explicit indications that while
funders of conservation delivery like the sound of adaptive
management, they are less keen to pay for it. The modicum
of adaptive management achieved resulted largely from
universities having access to lots of (predominantly)
young, enthusiastic people keen to gain qualifications
in conservation through applied research. For adaptive
management to be a reality and not just an aspiration, there
is a clear need for more integrated (research–management)
funding streams delivering vital continuity of support.
Our work demonstrates there is no technical difficulty
in expanding working with citizen conservationists for
pushing back huge scale invasion. Partnerships and
relationships had a critical role in achieving this work
(across all project phases). The outcomes have been
Lambin, et al.: Long-term mink control in Scotland
achieved through those networks and the empowerment
of volunteers and interested/affected ‘stakeholders’, all in
spite of repeated uncertainty of funding. There is little doubt
that even more could have been achieved had continuous
funding been in place. Indeed, short-term funding is a
major impediment to efficiency and increases the overall
cost of long-term invasive control as lost ground must
be recovered. Repeated gaps in funding, associated staff
turnover and re-badging of projects are all damaging to
the trust relationships built with volunteers. It makes no
economic sense to embrace long-term control of invasives
without funding it. Scotland, like other countries, needs
a long-term stream of funding if it is going to manage
invasive species. Thus, the future will tell whether our
efforts were bold and trail blazing or overly ambitious
and ultimately wasted if the SMI’s ambition to become
embedded within local management practice in perpetuity
is not borne out.
REFERENCES
Aars, J., Lambin, X., Denny, R. and Griffin, A.C. (2001). ‘Water vole
in the Scottish uplands: distribution patterns of disturbed and pristine
populations ahead and behind the American mink invasion front’.
Animal Conservation 4: 187–194.
Beirne, C. and Lambin, X. (2013). ‘Understanding the determinants of
volunteer retention through capture-recapture analysis: Answering
social science questions using a wildlife ecology toolkit’. Conservation
Letters 6: 391–401.
Bell, P., Nathan, H. and Mulgan, N. (2019) ‘Island eradication within
large landscapes: the remove and protect model’. In: C.R. Veitch, M.N.
Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) Island invasives:
scaling up to meet the challenge, pp. 604–610. Occasional Paper SSC
no. 62. Gland, Switzerland: IUCN.
Bomford, M. and O’Brien, P. (1995). ‘Eradication or control for vertebrate
pests.’. Wildlife Society Bulletin 23: 249–255.
Bryce, R., Oliver, M.K., Davies, L., Gray, H., Urquhart, J. and Lambin,
X. (2011). ‘Turning back the tide of American mink invasion at an
unprecedented scale through community participation and adaptive
management’. Biological Conservation 144: 575–583.
Bryce, R., van der Wal, R., Mitchell, R. and Lambin, X. (2013).
‘Metapopulation dynamics of a burrowing herbivore drive spatiotemporal dynamics of riparian plant communities’. Ecosystems 16:
1165–1177.
DIISE (2015). The Database of Island Invasive Species Eradications,
developed by Island Conservation, Coastal Conservation Action
Laboratory UCSC, IUCN SSC Invasive Species Specialist Group,
University of Auckland and Landcare Research New Zealand. <http://
diise.islandconservation.org>.
Evely, A.C., Pinard, M., Reed, M.S. and Fazey, I. (2011). ‘High levels of
participation in conservation projects enhance learning’. Conservation
Letters 4: 116–126.
Fraser, E.J., Lambin, X., Travis, J.M.J., Harrington, L. A., Palmer,
S.C.F., Bocedi, G. and Macdonald, D.W. (2015). ‘Range expansion of
an invasive species through a heterogeneous landscape—the case of
American mink in Scotland’. Diversity and Distribution 21: 888–900.
Glen, A.S., Pech, R.P. and Byrom, A.E. (2013). ‘Connectivity and invasive
species management: towards an integrated landscape approach’.
Biological Invasions 15: 2127–2138.
Harrington, L., Hughes, J. and Macdonald, D. (2008). Management
of American Mink in the Northern Highlands: A Proposed Cordon
Sanitaire Approach. Unpublished report to SNH, the Environment
Agency, the People’s Trust for Endangered Species and the European
Nature Trust. Oxford.
Lambin, X., Fazey, I., Sansom, J., Dallas, J., Stewart, W., Piertney, S.B.,
Palmer, S.C.F., Bacon, P.J. and Webb, A. (1998). ‘Aberdeenshire Water
Vole Survey: The Distribution of Isolated Water Vole Populations in
the Upper Catchments of the Rivers Dee and Don.’ Report to Scottish
Natural Heritage ref: C/LF1/BAT/97/2.
Lambin, X., Graziani, G. and Hon, J. (2002). Water Vole Survey in the
Deveron and Ythan Catchments 2002. Unpublished report to Aberdeen,
north-east Scotland LBAP Partnership and SNH.
Macleod, I.A., Maclennan, D., Withaker, S., Thompson, D.B.A., Raynor,
R. and Chaffer, R. (2019) ‘Large scale eradication of American mink,
Neovison vison, from the Outer Hebrides of Scotland’. In: C.R. Veitch,
M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) Island
invasives: scaling up to meet the challenge, pp. 261–266. Occasional
Paper SSC no. 62. Gland, Switzerland: IUCN.
Melero, Y., Robinson, E. and Lambin, X. (2015). ‘Density-and agedependent reproduction partially compensates culling efforts of invasive
non-native American mink’. Biological Invasions 17: 1–13.
Moorhouse, T.P., Macdonald, D.W., Strachan, R. and Lambin, X. (2015).
‘What Does Conservation Research Do, When Should it Stop, and What
Do we Do Then? Questions Answered with Water Voles’. In: D.W.
Macdonald and R.E. Feber (eds.) Wildlife Conservation on Farmland
Volume 1. Managing for Nature on Lowland Farms, pp. 269–290.
Oxford: Oxford University Press.
Oliver, M.K., Luque-Larena, J.J., and Lambin, X. (2009). ‘Do rabbits
eat voles? Apparent competition, habitat heterogeneity and large-scale
coexistence under mink predation’. Ecology Letters 12: 1201–1209.
Oliver, M.K., Piertney, S.B., Zalewski, A., Melero, Y. and Lambin,
X. (2016). ‘The compensatory potential of increased immigration
following intensive American mink population control is diluted by
male-biased dispersal’. Biological Invasions 18: 3047–3061.
Pretty, J.N. (1995). ‘Participatory learning for sustainable agriculture’.
World Development 23: 1247–1263.
Raynor, R., Horrill, C., Taylor, M. and Taylor, K. (2016). ‘American
Mink. Version 1.0’. In: M.J. Gaywood, P.J. Boon, D.B.A. Thompson
and I.M. Strachan (eds.) The Species Action Framework Handbook, pp.
287–313. Battleby, Perth: Scottish Natural Heritage.
Reynolds, J.C., Short, M.J. and Leigh, R.J. (2004). ‘Development of
population control strategies for mink Mustela vison, using floating rafts
as monitors and trap sites’. Biolological Conservation 120: 533–543.
Robertson, P., Mill, A., Shirley, M., Roy, S., Adriaens, T., Ward, A.,
Tatayah, V., and Booy, O. (2019) ‘Invasive species removals and scale
– contrasting island and mainland experience’. In: C.R. Veitch, M.N.
Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) Island invasives:
scaling up to meet the challenge, pp. 687–691. Occasional Paper SSC
no. 62. Gland, Switzerland: IUCN.
Santulli, G., Palazón, S., Melero, Y., Gosálbez, J. and Lambin, X. (2014).
‘Multi-season occupancy analysis reveals large scale competitive
exclusion of the critically endangered European mink by the invasive
non-native American mink in Spain’. Biological Conservation 176:
21–29.
Saunders, A. and Norton, D. (2001). ‘Ecological restoration at mainland
islands in New Zealand’. Biolological Conservation 99: 109–119.
Sheail, J. (2003). ‘Government and the management of an alien pest
species: A British perspective’. Landscape Research 28: 101–111.
Telfer, S., Holt, A., Donaldson, R. and Lambin, X. (2001). ‘Metapopulation
processes and persistence in remnant water vole populations’. Oikos 95:
31–42.
UK Biodiversity Partnership (1995). Biodiversity: the UK Steering Group
Report. Volume 2 : Action Plans. London: HMSO.
WildCRU, Oxford University (2004). Water vole surveys in Fife and
Central Cairngorms. Scottish Natural Heritage Commissioned Report
No. 058 (ROAME No. F01AC316).
Horrill, C., Oliver, M.K. and Stubbs Partridge, J. (2019) ‘Lessons on
effectiveness and long-term prevention from broad-scale control of
invasive alien species in Scotland’s rivers and lochs’. In: C.R. Veitch,
M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) Island
invasives: scaling up to meet the challenge, pp. 458–465. Occasional
Paper SSC no. 62. Gland, Switzerland: IUCN.
JNCC (2006). UK Biodiversity Action Plan. Peterborough: Department
for Environment, Food and Rural Affairs.
Lambin, X., Telfer, S., Cosgrove, P. and Alexander, G. (1996). Survey of
Water Voles and Mink on the Rivers Don and Ythan. Report to Scottish
Natural Heritage. ref: 70/96/F1A/007.
657
D. Moverley
Moverley, D. Battling invasive species in the Pacific
Battling invasive species in the Pacific
D. Moverley
Invasive Species Adviser, Secretariat of the Pacific Regional Environment Programme (SPREP). <davidm@sprep.org>.
Abstract This paper is a snapshot of how Pacific invasive species battlers are protecting their islands with the assistance
of the Global Environment Facility’s Pacific Alliance for Sustainability (GEF-PAS) project “Prevention, control and
management of invasive alien species in the Pacific”. The aims of the project are presented, along with examples of
how implementation increased awareness and capacity, and management of invasive plants and animals, throughout the
Pacific. Over 100 IAS activities took place across nine countries between 2011 and 2016. The project, one of the largest
investments in invasive species management in Pacific history, has raised the benchmark of invasive species management
in the Pacific and enhanced regional mechanisms. Hopefully the people within this story inspire and assist other battlers
to join the fight and protect our islands from invasive species.
Keywords: Cook Islands, Federated States of Micronesia, GEF-PAS Project, Kiribati, Niue, Palau, regional project,
Republic of the Marshall Islands, Samoa, SPREP, Tonga, Vanuatu
INTRODUCTION
The Pacific region is populated by diverse people
and spans a third of the earth’s surface and encompasses
about half of the global sea surface (Fig. 1). There are ca.
2,000 different languages and ca. 30,000 islands. Pacific
ecosystems are one of the world’s biodiversity hotspots,
with a large number of species found only in the Pacific
and nowhere else. There are 2,189 single-country endemic
species recorded to date. Of these species, 5.8% are already
extinct or exist only in captivity. A further 45% are at risk
of extinction (SPREP, 2013). The region faces some of the
highest extinction rates in the world. The largest cause of
extinction of single-country endemic species in the Pacific
is the impact of invasive alien species (IAS). Invasive alien
species also severely impact economies, ability to trade,
sustainable development, health, ecosystem services, and
the resilience of ecosystems to respond to natural disasters.
Fortunately, we can do something about it. Even in this
diverse region, many things are shared in common. The
people are self-reliant, rely heavily on their environment
to support their livelihoods and share many common IAS
issues as they are ultimately connected. Sharing what is
learnt regionally benefits the people and their families
economically, culturally, and in their daily lives. The 2013
State of Conservation in the Oceania assessment (SPREP,
2013) showed that IAS are the most important driver of
species loss in the region and contribute directly to the loss
of ecosystem function and loss of resilience, and ability to
respond to climate change threats. Invasive alien species
also severely impact Pacific economies, ability to trade,
sustainable development, health, ecosystem services, and
the resilience of ecosystems to respond to natural disasters.
The status of the IAS issue in the Pacific is “poor”
according to the report on the State of Conservation in
Oceania (SPREP, 2013).
Pacific strategy – overcoming challenges
Fig. 1 Island nations and groups in the Pacific. Island locations and sizes are not to scale.
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
658
up to meet the challenge, pp. 658–662. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Moverley: Battling invasive species in the Pacific
The “Guidelines for invasive species management in
the Pacific: a Pacific strategy for managing pests, weeds
and other invasive species” (SPREP, 2009) provide a
comprehensive framework for the Pacific Region to respond
to IAS at the regional and national levels, endorsed in 2009
by members of both the Secretariat of the Pacific Regional
Environmental Programme (SPREP), and the Pacific
Community (SPC). This framework is used throughout
the Pacific for structuring the National Invasive Species
Strategies and Action Plan (NISSAP) and the Territorial
Invasive Species Strategies and Action Plan (TISSAP).
The Guidelines were implemented to achieve the
objective of reducing the environmental, economic, and
human health impacts of IAS in both terrestrial and marine
habitats in the Pacific region. The classification of the IAS
management themes within the Guidelines allows current
and future IAS management activity and success to be
measured both nationally and regionally and enables the
identification of gaps which need to be addressed.
The project “Prevention, control and management of
invasive alien species in the Pacific Islands” (GEF-PAS)
was implemented by the United Nations Environment
Programme (UNEP) and executed by the SPREP and
national partner agencies from 2011 to 2016. The GEFPAS project goal was to ‘conserve ecosystems, species
and genetic diversity in the Pacific Region’. The GEF-PAS
project structure followed the ‘Guidelines for Invasive
Species Management in the Pacific’ (SPREP, 2009) with
three major components: (i) foundations; (ii) problem
definition, prioritisation and decision making; and (iii)
management action.
This regional approach has supported the establishment
of the Pacific Invasive Species Guidelines Reporting
Database, a database of national, territorial and regional
progress in implementing the “Guidelines”, with indicators
on priority IAS initiatives. SPREP coordinates two Pacific
IAS networks. The Pacific Invasives Partnership (PIP) is
an umbrella group of IAS experts from organisations who
work on IAS issues in more than one Pacific country. PIP
is focused on coordinating IAS assistance in the Pacific
region and aims to build cooperation among Pacific experts
who provide assistance to Pacific countries and territories.
SPREP also coordinates the Pacific Invasives Learning
Network (PILN), a peer network of cross-sectoral IAS
practitioners in the Pacific. The PILN aims to build
cooperation between Pacific countries and territories on
IAS issues. There are PILN teams in all but three of the
21 SPREP Pacific island member countries and territories.
A knowledge management system was initiated by the
development of the Pacific Invasive Species Battler Series
launched in 2016 with nine booklets to date focused on
common IAS issues, based on Pacific examples and serving
the Pacific Region. They are available from the Pacific
Invasive Species Battler Resource Base (<https://piln.
srep.org>), a searchable resource base providing the latest
information on IAS issues, case studies, and introductory
guides on common IAS issues. This resource is designed to
increase the capacity of Pacific countries and territories in
an effective and efficient manner, The “battler” brand has
developed from Pacific IAS practitioners’ internal/external
communications over the years and serves as an ongoing
reminder that if we don’t achieve any change on the ground
as practitioners we have been ineffective, it reminds us that
IAS management is a long-term challenge that most of
us will be working on for the rest of our lives and that as
a regional collective we are not alone on our individual
islands.
OUTCOMES OF THE GEF-PAS PROJECT
Ten Pacific Island countries originally participated in
the GEF-PASIAS project: the Cook Islands, Federated
States of Micronesia, Kiribati, Marshall Islands, Niue,
Palau, Papua New Guinea, Samoa, Tonga and Vanuatu.
These were reduced to nine countries with the withdrawal
of Papua New Guinea following the mid-term review,
due to issues related to the country’s readiness to engage
with the project. The project was therefore responsible
for delivering support to a culturally, geographically and
economically diverse set of Small Island Developing
States (SIDS) spread across the vast geographical scope of
the Pacific Ocean.
In-country subprojects and activities were facilitated
by National Project Coordinators and overseen by national
Invasive Species Coordinating Committees. The project’s
goal "to conserve ecosystems, species and genetic diversity
in the Pacific region" is broad and aspirational and is
backed by the objective "to reduce the environmental,
economic, and human health impacts of invasive alien
species in both terrestrial and marine habitats in the Pacific
region". The project budget was US$7,010,890 consisting
of US$3,031,818 in GEF funds and US$3,979,072 in
SPREP and country co-financing. The project consists of
five components, three of which can be described as core
components which relate to the three major areas of work
and nine thematic directions.
Component 1 – Foundations: generating support
This component addresses the limited understanding of
the threats posed by IAS to the environment, economies,
human health and cultural values of decision makers,
the private sector and the general public. It aims to raise
awareness across all sectors of society of the importance
of IAS risks and impacts, and of the benefits of IAS
management for biodiversity, the economy and human
health. It also aims to actively support IAS management.
With raised awareness, it is expected that sufficient
resources will become available to enable all national
and regional IAS priorities to be addressed and, most
importantly, enable capacity building efforts to flourish
alongside the development of supportive policy and
legislation.
The three thematic directions addressed by Component
1 are:
Generating support— Raising awareness of the
impacts of invasive species on biodiversity, the economy,
human health and socio-cultural values, and generating
support for action to manage and reduce them.
Building capacity— Developing the institutions, skills,
infrastructure, technical support, information management,
linkages, networks and exchanges required to manage
invasive species effectively.
Legislation, policy and protocols— Ensuring that
appropriate legislation, protocols, policies and procedures
are in place and operating, to underpin the effective
management of invasive species.
Component 2 – Problem definition, prioritisation and
decision-making: baseline and monitoring
This component aims at addressing the chronic lack
of information and data on IAS within the region which
impacts on the ability of governments to define priorities,
develop national strategies and establish supportive policies
and legislation. It aims to ensure that information and data
on IAS, their distribution and status is readily available to
support informed decision making, strategic planning and
effective management. Importantly, the component also
aims to address the potential biosecurity and economic
impacts of IAS through improved knowledge of transboundary movement and regional status of critical IAS.
659
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
The three thematic directions addressed by this
Component are:
Baseline and monitoring— Establishing a baseline
of information on the status and distribution of invasive
species and a programme for detecting change, including
range changes and emerging impacts.
Prioritisation— Establishing effective systems
for assessing risk and prioritising invasive species for
management.
Research on priorities— Understanding priority
invasives, including species biology and impacts, and
developing effective management techniques.
Component 3 – Management action
This component addresses the practical requirements of
managing IAS. Until management action is implemented,
no progress is made on addressing IAS. The three thematic
directions addressed by this component are:
Biosecurity—Preventing the trans-boundary and interisland movement of IAS in the region by encouraging the
establishment of cost-effective biosecurity measures (e.g.
rapid response protocols) aimed at reducing the need for
costly post-invasion control measures. It aims to assist
the establishment of effective systems throughout the
Pacific to regulate introductions and to detect and manage
unauthorised or accidental introductions across borders or
to new islands within countries.
Management of established invasives—Reducing or
eliminating the impacts of established invasive species,
by eradication, containment, exclusion, or population
reduction by physical, chemical or biological control.
Restoration—Restoring native biodiversity or
ensuring recovery of other values, after invasive species
management.
Each of the above three components dovetails directly
with the priority thematic areas of the Guidelines which
were developed as a result of an extensive regional
stakeholder consultation process in 2007/2008. As such,
they reinforce the rationale and justification for the IAS
project and its legitimacy in the eyes of the regional IAS
stakeholders and their international partners and networks.
Together, the three components also address the IAS
management weaknesses identified in the Guidelines.
Components 4 and 5 – Project management &
monitoring and evaluation
These management-related components establish
SPREP as the designated project Executing Agency
and support a Project Facilitator and half-time Financial
Manager for this purpose. SPREP funding covered the
costs of the Project Manager. SPREP also had designated
responsibility to ensure an effective monitoring and
evaluation framework is established at inception. This role
is consistent with SPREP’s regional mandate and role to
foster national and Pacific-wide strategies consistent with
international best practices. SPREP is also able to engage
the member organisations of the umbrella coordinating
body the Pacific Invasives Partnership to further the goals
of the project through provision of advice and the PIP
members’ own IAS management and capacity building
interventions. The project activities strengthened capacity
by improving IAS outreach, policies, laws, prevention and
management. The project helped participating countries
and others in the Pacific region to address existing and
future biological invasions.
660
ACTIONS
More than 100 IAS activities took place across nine
countries between 2011 and 2016 (SPREP, 2016). Here
the scope and range of the purpose of actions undertaken
and examples of those actions are highlighted.
Awareness raising and capacity building
Awareness of the impacts of IAS was increased at the
local, governmental and political level. As an example, a
royal visit to Toloa Rainforest by His Majesty King Tupo
VI and Her Majesty Queen Nanasipau’u raised the profile
of IAS management in Tonga. School scholarships were
also presented by Her Majesty Queen Nanasipau’u to the
top three Tupou College forest restoration team members
at a national school prize-giving.
Awareness of IAS is important to create or support
actions. For countries to take control of their responses
to invasives, the first steps were to develop awareness
in communities (local to national, and across a range of
social roles), to mainstream IAS issues, to create or access
long-term external funding mechanisms, and to generally
increase the support for IAS issues. As an example,
engaging posters were made by teams in Palau, Vanuatu,
and the Cook Islands to communicate which species
were invasive, what they affected, and boost the idea that
individuals can take action. Outreach is a vital component
of battling IAS because an educated, engaged community
produces fast, effective action.
Capacity building of institutions, skills, infrastructure,
technical support, information management, networks
and exchanges required to manage IAS effectively were
developed. Particularly given the strong customary land
ownership in the Pacific, on-site management requires
whole-of-community engagement, and the strong
community ties in the Pacific are a strength for IAS
early detection and rapid response. Local people with
site knowledge and experience were integral to project
implementation and benefited from learning new field
techniques and scientific approaches, enhancing regional
capacity. Direct engagement with field action makes local
communities more likely to maintain site management,
value their environment, and support or generate future
conservation.
As examples, biosecurity training was provided in
Kiribati, and a multi-country workshop was held in Samoa
to support the prevention of IAS movements between
islands. Training to detect and manage little fire ants was
conducted in Vanuatu. A workshop on eradicating rodents
from small islands was held in 2015, with participants
from Kiribati, Republic of the Marshall Islands, Tonga, and
Wallis and Futuna practising the eradication techniques on
Malinoa and Motutapu islands in Tonga. The removal of
the rats (Rattus spp.) has already boosted bird populations,
such as the fuleheu or wattled honeyeater (Foulehaio
carunuculata) and misi or Polynesian starlings (Aplonis
tabuensis). Black-naped terns (Sterna sumatrana) were
nesting and had eggs on the beaches of both islands.
The Pacific region is under-resourced regarding
research capability and IAS, biodiversity, and ecosystem
data. The limited resources available for IAS management
demand that achievable goals are prioritised based on
research and available data and that priorities meet the
expectations of all stakeholders. Further, as Parties to
international environmental agreements such as the
Convention on Biodiversity, the region needs to show
progress and success in meeting their obligations under
these agreements.
Moverley: Battling invasive species in the Pacific
Invasive plant species actions
Weed risk assessments can be costly and timeconsuming, and vital information for assessments such
as seed viability may not be known. Given the limited
resources available to Pacific island countries and territories
and the existence of almost 2,000 species with existing
weed risk assessments for the Pacific, the most effective
first step is to ensure that existing weed risk assessments
are being used. This was a focus of the project and resulted
in the Battler series publication “Find answers online to
common invasive species questions”.
Weed risk assessments contribute to prioritisation
of target species, areas, and activities in combination
with stakeholder consultation and local knowledge
during NISSAP formulation. There are ongoing priority
weed programmes operating in the Pacific which are
showing success towards eradication. Widespread weeds
can sometimes be targeted using biological rather than
chemical or physical control: 36 natural enemies have
established on 19 weed species in the Pacific. Since 1911,
17 countries and territories have deliberately released
biological control agents on weeds in the Pacific. There are
many opportunities for distributing existing agents further
around the Pacific and opportunities to target new species.
Unlike biological control agents that were introduced
to target invasive animals but were devastating to those
islands to which they were introduced, such as the Indian
mongoose for controlling snakes and the rosy wolf snail
for controlling the giant African snail (Lissachatina fulica),
the use of biological control to manage widespread weeds
has been much more successful and much safer following
standard international protocols such as host-specificity
testing on other possible desirable plants. Internationally,
483 agents have been released with none resulting in
unpredictable non-target effects (M. Day & L. Hayes, pers.
comm.).
The development of biological control agents for
weeds can be initially expensive. However, once agents are
researched and located, they can provide an endless service
of controlling invasive plants to a degree where their
impact is greatly reduced. Further, once the initial agents
are confirmed as effective, it is relatively cheap to move
them to new countries or locations following any additional
location-specific host-testing that may be required.
Vanuatu was fortunate to benefit from two new agents
being developed for African tulip tree under a Landcare
Research New Zealand project with the Cook Islands. The
agents required minimal further host specificity testing and
are due to be introduced to Vanuatu in the near future.
Palau has benefited from many years work on Mikania
micrantha, which has a natural rust enemy already
established in many countries. Accordingly, previous host
specificity testing has been carried out extensively for
many countries, leaving just one plant for which Palau was
required to do tests on. The rust was introduced to Palau
but was not successful for undetermined reasons. It is
planned that the rust will be moved to some states within
the Federated States of Micronesia, such as Yap, where
Mikania micrantha is a serious weed. Mikania micrantha
is located in 20 Pacific countries and territories; however,
the rust agent has only been introduced to six to date.
Invasive animal species actions
There are good examples of sustainable control projects
operating in the Pacific. A key action for environmental
protection is to prevent the spread of IAS across
international or internal borders. The four main stages are
pre-export control, pre-border control, at-border control,
and post-border rapid response.
Niue created a harmonised Biosecurity Bill which
allows environmental concerns to be addressed along
with the traditional agricultural and trade concerns. Early
detection and rapid response (EDRR) plans have been
created for the Cook Islands, Kiribati, and Samoa. The
plans detail the staff and funding requirements, identify
best practices regarding known target species, and convey
decisions made about the country’s approach to the known
and potential IAS. Simulation exercises were completed to
identify gaps before a response becomes necessary.
The Battler series booklet “Catch it early: invasive
species early detection and rapid response” outlines the
components of effective IAS response systems. The creation
of response plans, training of staff, and procurement of
equipment needs to be supported by ongoing engagement,
regular refresher simulation exercises, and greater public
awareness to maintain fast responses to incursions. The
accidental introduction of five mongooses to Tonga in
2016 demonstrated the need for rapid, planned response
action. Long-term management is often required for IAS
that cannot be eradicated due to their value as a cultural
or livelihood resource or simply the amount of resources
required to do so. Such management requires ongoing
resourcing but may be the only option available, so
the value that is being protected from the IAS needs to
outweigh the cost of management.
Managing pigs is a balance between cultural or food
needs and environmental needs. Domestic pigs are kept in
pens as an important food source, but pigs that get out of
pens cause a lot of damage. Investing in upskilling locals
to the level of professional hunters has paid off in Niue.
In 12 months, approximately 130 pigs have been hunted
by locals, reducing the population by one half from the
estimated pre-hunt total. Professional pig-hunting dogs,
global positioning systems, and expert mentoring have
resulted in a sustainable, low-cost method for managing
pigs on Niue. In Samoa and Vanuatu, crown of thorns
starfish is the target of on-going control. Although a native
species, crown of thorns can become invasive following
modifications to the environment by man or natural
disasters such as tsunamis and cyclones. In both countries,
the local communities are provided with tools, procedures,
and support to lower the impact that the outbreaks have on
their local marine ecosystem.
Many IAS are already widespread in the Pacific and
impacting biodiversity, including in protected natural
areas. When this is the case, there are still options to
protect these species and ecosystems with a site-led or
asset-based approach. Exclusion of IAS is an option if
the surrounding environment contains widespread or
otherwise unmanageable IAS which may affect high-value
areas. It can also be used as a short-term measure until
a solution becomes available. In Mount Talau National
Park in the Vava’u islands of Tonga, a rare plant Casearia
buelowii, which is endemic to Mt Talau and only survives
through fewer than 20 individuals, was being continually
undermined by pigs, exposing the roots to damage and
the heat of the sun. A Tongan pig fence was constructed
around the site to exclude the pigs until a long-term pig
management solution can be found. Widespread IAS may
also be contained to restrict their arrival in uninfested areas.
On Niue, the primary infestation of taro vine is situated
within the villages of Alofi and Alofi South. Isolated
infestations are targeted to contain the infestation to the
primary infestation site.
There are good examples of site-led or asset-based
restoration projects operating in the Pacific. Restoration
supports species recovery and the continued provision
of ecosystem services. Ideally, restoration involves the
community at multiple scales because restoration is a longterm, if not continual, process. The Kingdom of Tonga is
restoring two key ecological sites: the Toloa Rainforest and
661
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
Mt Talau. The Toloa Rainforest restoration efforts include
a Pacific (R. exulans) and ship rat (R. rattus) control
programme, via bait stations, for the whole forest to save
native bird and plant species from predation. Replanting
of native trees in the rainforest began in 2014 to improve
structure and size and to reintroduce plant species that
have gone extinct within the forest. Planting will continue
until the extension of the forest is complete and the forest
sub-canopy and disturbed sites are restored. Weed and rat
control will continue into the future. Toloa Rainforest is the
last remaining stand of native forest on the main Tongan
island of Tongatapu and serves as an educational resource
for the schools of Tonga. A key aspect of the project is
making information readily available for people who
visit the forest, and many of these informational products
explain native species and IAS threats. A trail, with rest
and wildlife viewing stops throughout, has also been
developed.
Hengahenga, or Tongan whistler (Pachycephala
jacquinoti), are recovering on Mt Talau following rat
control. Rodents have been controlled for four years with
statistically significant increases in the number of Tongan
whistler (endemic to Vava‘u) and other birds such as the
Polynesian triller (Lalage maculosa) and Polynesian
starling. Rats heavily impact the survival and productivity
of the Tongan whistler because the birds build an open bowl
nest that is easily accessed by rats. The control programme
is run by the local community with the assistance of the
Vava‘u Environmental Protection Association. It uses a rat
bait-take database that captures, stores, and reports on bait
take at each bait station during the programme and allows
analysis of bait take to inform success at lowering the rat
population, identify areas of high rat activity, and allow
for more economical use of the bait. Hengahenga are now
seen and heard in the surrounding area with many Tongans
witnessing this bird for the first time in their lives.
Samoa has also embarked on restoring two important
sites, Mount Vaea Reserve and O Le Pupu Pu‘e National
Park. On Mt Vaea, the focus has been on controlling
widespread weeds, which form 90 percent of all stems
within the regular survey sampling plots. Following weed
control, each area is re-vegetated with native trees which
quickly form a canopy, reducing the ability of the lightdemanding weed species to grow. The weed species that
are adapted to grow under low light conditions are regularly
managed. Six hectares of Mt Vaea have now been restored,
with the on-site nursery having provided19,000 trees for
volunteers and the local village to plant. The focus at O Le
Pupu Pu‘e has been on planting to suppress the Merremia
peltata infestations that are restricting regeneration
following disturbances such as cyclones. Again, the closed
canopy of the revegetated areas is supressing this lightdemanding species.
projects in the Pacific need to be extended for one year to
allow for this.
The “Guidelines for Invasive Species Management
in the Pacific” provide an effective framework to plan
IAS programmes and can be directly used to plan actions
within NISSAPs to measure and analyse actions between
countries with consistency. Regional support was vital
in achieving quality outcomes. The GEF-PAS project
allowed support to be increased substantially, and this
needs to continue. Empowering country coordinators and
stakeholders through coaching, capacity building, and
support in-country, is the most important means to sustain
IAS management capability. This needs continued support.
The Pacific Invasives Learning Network appears to
be the most effective means for Pacific practitioners to
work together and learn from each other. PILN creates and
supports the regional flow of information regarding IAS
management. PILN requires sustained support. Successful
projects resulted in increased visibility and support by
local communities, other related sectors, and at the political
level. This support is indicated by institutionalisation
within government agencies of a core IAS role, as has
been the case in some countries, and the commitment to
progressing IAS management within the following Global
Environment Facility replenishment cycles and other
funds.
The GEF-PAS project has made significant progress.
The implementation of GEF 6 and EDF 11 projects will
build substantially on these successes and continue to refine
and enhance the management of IAS in the Pacific Region.
Climate change is increasing the intensity and urgency of
the response to IAS by reducing the capacity and resilience
of Pacific ecosystems and societies to adapt to climate
change. Invasive alien species management needs to be an
accepted tool for Pacific ecosystems and communities to
adapt to climate change.
FUTURE PACIFIC PROJECTS
SPREP. (2009). Guidelines for invasive species management in the
Pacific: A Pacific strategy for managing pests, weeds and other invasive
species. Compiled by Alan Tye. Apia, Samoa. <https://www.sprep.org/
node/9588>.
SPREP. (2013). State of Conservation in Oceania Regional Report (2013)
Published by the Secretariat of the Pacific Regional Environment
Programme. <https://www.sprep.org/attachments/Publications/BEM/
state-conservation-oceania-report.pdf>.
SPREP. (2016). Battling Invasive Species in the Pacific: Outcomes of the
Regional GEF-PAS IAS Project; Prevention, control and management of
invasive species in the Pacific Islands Secretariat of the Pacific Regional
Environment
Programme.
<https://www.sprep.org/attachments/
Publications/BEM/battling-invasive-species-pacific.pdf>.
UNEP. (2017). Prevention, control and management of Invasive Alien
Species in the Pacific Islands. Evaluation Office of UN Environment,
Terminal Evaluation of UN Environment Project. <https://www.
google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&ved=2
ahUKEwjJpt2OhPPcAhXGgLwKHb4KBkcQFjAAegQIABAC&url=
https%3A%2F%2Fwww.thegef.org%2Fsites%2Fdefault%2Ffiles%2F
project_documents%2F3664_2017_TE_UNEP_Regional_BD_FSP_
UNEPSPEM_IAS_Pacific_1.pdf&usg=AOvVaw1XY9XpCiwpWr_
JBuKLlvR0>.
The GEF-PAS Project was evaluated by the United
Nations Environment Programme (UNEP, 2017). The
summary of project criteria “Strategic relevance”,
“Achievement of outputs”, “Communication and public
awareness”, “Supervision, guidance and technical
backstopping” all had overall ratings of Highly Satisfactory;
“Socio-political sustainability” had an overall rating of
Highly Likely; “Preparation and readiness” had an overall
rating of Unsatisfactory.
The increased establishment and support of existing
national institutionalised IAS programmes needs to be
carried forward, building upon the success of GEF-PAS.
Project under-preparedness and -readiness is common
to most Pacific projects due to the lack of capacity and
increased logistical challenges in such a large geographic
area. Future project preparation times for large, complex
662
ACKNOWLEDGEMENTS
The project outcomes were achieved by dedicated
Pacific country teams from the Cook Islands, the Federated
States of Micronesia, Kiribati, Niue, Palau, Republic of
the Marshall Islands, Samoa, Tonga and Vanuatu with
the assistance of the SPREP Invasive Species Team. The
project was funded by the Global Environment Facility
and implemented by the United Nations Environment
Programme. This paper was compiled by the SPREP
Invasive Species Team and has benefited from the
persistence of Dick Veitch and the review and editing of
James Russell. A big thanks and congratulations on the
achievements of this project.
REFERENCES
S. Oppel, S.J. Havery, L. John, L. Bambini, K. Varnham, J. Dawson and E. Radford
Oppel, S.; S.J. Havery, L. John, L. Bambini, K. Varnham, J. Dawson and E. Radford. Maximising conservation impact by prioritising islands for biosecurity
Maximising conservation impact by prioritising islands for biosecurity
S. Oppel1, S.J. Havery1, L. John1, L. Bambini1, K. Varnham1, J. Dawson2 and E. Radford1
Royal Society for the Protection of Birds, RSPB Headquarters, the Lodge, Sandy, Bedfordshire. SG19 2DL. U.K.
<Steffen.Oppel@rspb.org.uk> 2Durrell Wildlife Conservation Trust, Les Augrès Manor, La Profonde Rue, Trinity,
Jersey, British Channel Islands. JE3 5BP. U.K.
1
Abstract Invasive alien species are one of the primary threats to native biodiversity on islands worldwide, and their
expansion continues due to global trade and travel. Preventing the arrival and establishment of highly successful
invasive species through rigorous biosecurity is known to be more economic than the removal of these species once
they have established. However, many islands around the world lack biosecurity regulations or practical measures and
establishing biosecurity will require social and financial investments. Guiding these investments towards islands where
native biodiversity is at highest risk from potential invasions is of strategic importance to maximise conservation benefit
with limited resources. Here we implement an established prioritisation approach, previously used to identify which
islands will have the greatest conservation gains from the eradication of invasive species, to identify which islands
would benefit the most from establishing or improving biosecurity. We demonstrate this approach for 318 islands in
the Caribbean UK Overseas Territories and Bermuda where we considered all threatened native terrestrial vertebrates
that are vulnerable to the most harmful invasive vertebrates (black and brown rats, cats, small Indian mongoose, green
iguana). The approach calculates the increase in conservation threat score resulting from anticipated negative effects of
potential invaders on native biodiversity, and highlighted Sombrero (Anguilla) and Cayman Brac (Cayman Islands) as
important islands where threatened reptile species would likely be eliminated if rats, feral cats or mongoose invaded.
Feasibility and cost implications should now be investigated more closely on the highlighted islands. The prioritisation
presented here can be expanded to more islands and more invasive/native taxa (herbivores, plants and invertebrates), but
requires a classification of the severity of potential impacts between invasive and native species for which currently little
information exists. Besides highlighting opportunities for biosecurity, this approach also highlights where knowledge
gaps about population sizes of and threats to reptiles with restricted ranges exist.
Keywords: Caribbean, feral cat, iguana, invasive mammals, mongoose, rats, reptiles
INTRODUCTION
The majority of the world’s archipelagos have been
invaded by non-native species, some of which have
detrimental effects on native biodiversity (Atkinson, 1985;
McCreless, et al., 2016; Turbelin, et al., 2017). Although
some islands can be restored by eradicating certain invasive
species, such operations can be expensive (Martins, et al.,
2006; Holmes, et al., 2015). The limited amount of funding
available for island restoration efforts has motivated
managers to prioritise the islands where an eradication
would yield the greatest biodiversity benefits at global and
regional levels (Brooke, et al., 2007; Dawson, et al., 2015;
Stanbury, et al., 2017). However, current technologies
limit restoration via eradication to 15% of islands that have
been invaded (Keitt, et al., 2019), hence eradication is not
a universal solution to preserve global island biodiversity.
Preventing harmful species invading those islands
which still have globally significant biodiversity values
is an important and efficient avenue to prevent loss of
biodiversity (Broome, 2007; Russell, et al., 2008; Spatz,
et al., 2017). Biosecurity measures also require financial
investments, both initially and in perpetuity, to detect
and eliminate any potential invaders to islands (Oppel,
et al., 2011; Key & Moore, 2019). Because the costs for
biosecurity can be considerable, financial constraints can
also limit the number of islands that can be protected
with effective biosecurity measures (Moore, et al.,
2010; Greenslade, et al., 2013). Here we propose to use
established prioritisation approaches (Brooke, et al., 2007;
Dawson, et al., 2015; Stanbury, et al., 2017) to guide
the investment of resources for biosecurity to minimise
the risks of invasion of non-native vertebrates to islands
where they would cause the greatest loss of biodiversity.
We demonstrate this approach for 318 islands that belong
to United Kingdom Overseas Territories (UKOTs) in the
Caribbean and Bermuda.
The islands in the Caribbean UKOTs feature globally
important biodiversity (Forster, et al., 2011; Dawson, et
al., 2015; Churchyard, et al., 2016), with a large number
of endemic reptiles, birds, and plants. Due to centuries
of human habitation and inter-island trade, most islands
have been invaded by some non-native species (Hilton &
Cuthbert, 2010), but only a few islands contain the complete
suite of invasive vertebrate species present in the Caribbean
region. In addition, >100 small and uninhabited islands
are still free of invasive vertebrate species and function as
refugia for some globally threatened species that cannot
coexist with harmful invasive vertebrates (Dawson, et al.,
2015). Preventing the invasion of non-native vertebrates
that have caused significant declines to native species on
other islands could secure globally significant populations
of threatened vertebrates. Despite the recognised threat
of invasive species to endemic biodiversity, biosecurity
regulations and implementations are generally insufficient
to reduce the risk of further spread of invasive species
between islands in the Caribbean region (RSPB, 2017; Key
& Moore, 2019).
We conducted a prioritisation that identifies those
islands where the invasion of five potentially harmful
invasive vertebrates could cause the greatest loss to
biodiversity in the Caribbean UKOTs. We recommend
immediate investment in feasibility studies and biosecurity
on those islands to avoid the invasion of these five species
and the subsequent loss of native biodiversity, and we
recommend that similar approaches should be used in
other regions, or indeed globally, to identify islands where
investment in biosecurity is most urgently needed.
METHODS
Study area
We used all 318 islands in the five Caribbean UKOTs
(Anguilla, British Virgin Islands, Cayman Islands,
Montserrat, and Turks and Caicos Islands) and in Bermuda,
which is situated 1,500 km north of the Caribbean but
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 663–669. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
663
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
is climatically similar (Fig. 1). These islands are mostly
tropical and range from small sandy islets of 0.01 ha to
islands with mountain ranges and a variety of habitat types
> 20,000 ha. Only 14 islands are permanently inhabited
by human communities of up to 65,000 people, while
the remaining islands are either completely uninhabited,
function only as tourist resorts or destinations, or are
visited temporarily by fishermen.
Selection of potential invasive species
To assess biodiversity loss that could result from the
invasion of harmful animal species, we selected the five
most harmful invasive terrestrial vertebrates (McCreless,
et al., 2016) that are widespread in the Caribbean region.
Green iguanas (Iguana iguana) are known to hybridise
and compete with native reptiles (Gibbon, et al., 2000;
Vuillaume, et al., 2015), small Indian mongoose (Urva
auropunctata) are versatile predators considered one of the
worst invasive species (Hays & Conant, 2007; Barun, et al.,
2008), brown (Rattus norvegicus) and black rats (R. rattus)
and feral cats (Felis catus) are efficient predators that can
have detrimental effects on island biodiversity (Towns,
et al., 2006; Jones, et al., 2008; Medina, et al., 2011;
Nogales, et al., 2013). These five species are distributed
widely across islands in the Caribbean (Kairo, et al., 2003;
Dawson, et al., 2015) and are therefore potential invaders
of all islands in the region.
Distribution of native and invasive species
For each island we previously collated information on
the presence of native and invasive terrestrial vertebrate
species for an eradication prioritisation (Dawson, et al.,
2015) and a general inventory of biodiversity (Churchyard,
et al., 2016), and updated these previous compilations
with recent information and threat assessments (IUCN,
2017). We considered all globally threatened terrestrial
vertebrate species (including marine turtles) as listed on the
International Union for Conservation of Nature Red List of
Threatened Species (IUCN, 2017) and all colonial seabird
species and restricted range bird species. We also included
reptiles of conservation concern that are endemic to a single
territory or inhabit fewer than 15 islands across their range
(Dawson, et al., 2015). We updated this information with
new records shared by local partner organizations since
2013 (Hedges, 2017). We considered the green iguana that
exists on Montserrat as a genetically distinct conservation
management unit, because it is genetically closely related
to the iguana on Saint Lucia, which is treated as a native
species of conservation concern (Powell, 2004; Stephen,
et al., 2013; Vuillaume, et al., 2015). Due to the lack of
sufficient distribution data and limited existing knowledge
of interactions, native and invasive plant or invertebrate
species were not considered in this prioritisation.
Calculating the conservation threat score of islands
We followed the approach of Dawson, et al. (2015) to
calculate the conservation threat score (termed ‘conservation
value’ in Dawson, et al., 2015) of each island based on the
sum of each native species’ vulnerability. The vulnerability
was calculated as the product of the global threat status,
the irreplaceability, which indicates the global significance
of an island’s population, and the severity of impact of the
most harmful invasive vertebrate species already present
on an island (i.e. the species with the greatest severity of
impact score; Dawson, et al., 2015; Stanbury, et al., 2017).
We scored threat and impact categories on both a linear and
logarithmic scale to address the arbitrariness of assigning
quantitative values to normative categories (Game, et al.,
2013; Helmstedt, et al., 2016). The severity of impact
was classified in three categories, depending on whether
an invasive species had no impact on a native species (0),
small to moderate impact that would reduce population
size but allow the native species to persist (1), or a severe
impact that would eventually lead to the local extinction of
the native species (2). We classified unassessed reptiles as
‘At Risk’, which received a numerical value equivalent to
‘Vulnerable’ (Dawson, et al., 2015).
Simulating the invasion of islands to calculate increase
in conservation threat score
To quantify the magnitude of biodiversity loss that
could result from invasion, we first assessed which of the
five selected invasive species were already present on an
island in 2016, and then simulated the arrival and invasion
of those species that were not yet present in 2016. We then
re-calculated the conservation threat score of each island
as described above, where the vulnerability of each native
species was adjusted to reflect the most harmful invasive
species on the island, which may be one of the simulated
invaders. We assumed that all invasive species not yet
present on an island would invade, because biosecurity
measurements should, in our opinion, not be tailored for a
single species but guard against the arrival of a broad suite
of species. However, we emphasise that our prioritisation
could also be performed for single species invasions, but
assessing the merits of guarding against one or another
invasive species would require information about the
relative invasion risk of various species.
The calculation of the conservation threat score
depends on a classification of the threat posed by each
invasive species to each native species, but these threats
can be hypothetical for interactions between certain island
endemic species and invasive species that have so far not
invaded the respective island. Consequently, we drew on
taxonomically related or otherwise very similar species
to specify the potential threat that would result from
invasion. For example, if black rats adversely affect a small
Sphaerodactylus gecko on one island, we assumed that a
similarly sized Sphaerodactylus species that is endemic to
an island without any rats would suffer similar effects if the
island were invaded by rats (Case & Bolger, 1991).
Prioritising islands for biosecurity
Fig. 1 Location of 318 islands (black dots) in six United
Kingdom Overseas Territories where the priority for
biosecurity was assessed. Circles around islands
indicate the location of the highest priority islands listed
in this paper.
664
Islands that should receive the most immediate
investment into biosecurity are those where the native
fauna would face the greatest increase in conservation
threat score if the five selected vertebrate species invaded.
We therefore calculated the difference in conservation
Oppel, et al.: Prioritising islands for biosecurity
value at present and after the simulated invasion of the
five vertebrate species, and ranked islands based on the
magnitude of this difference. We present the results as a
ranking table and include information on island size and
human population size for each island. These aspects will
affect the complexity and cost of biosecurity measures,
as well as the probability of invasive species arrival and
establishment, but they did not factor into our prioritisation
of islands for biosecurity, which was entirely based on the
potential threat to native biodiversity. All calculations were
performed in R 3.2.5 (R Development Core Team 2015)
based on the code provided by Dawson, et al. (2015).
RESULTS
Of the 318 islands in our assessment, 125 did not have
any invasive species on them, and 150 (47%) did not have
any of the five focal invasive species. Of the islands with
any of the five focal invasive species, 31 (10%) had one
invasive, 117 (37%) had two, 12 (4%) had three, 6 (2%)
had four, and only two islands (Tortola and Virgin Gorda,
British Virgin Islands) had all five of the focal invasive
species. On 183 islands (57.5%) the invasion of any of the
five focal invasive species would not lead to an increase in
the conservation threat score, because the native vertebrates
on these islands were not at greater risk of predation from
those invasive species that have not yet invaded. Thus,
biosecurity measures to prevent the invasion of at least one
of the five focal species would be useful on 133 islands in
our assessment.
We identified several important islands across the
Caribbean UKOTs and Bermuda where biosecurity could
help prevent the loss of globally important biodiversity
(Table 1). Two islands emerged where an invasion of
non-native vertebrates could lead to an increase in the
conservation threat score more than five times greater
than on any other island included in our study, mostly due
to the potential loss of Critically Endangered endemic
reptiles (Table 1): Sombrero (Anguilla), and Cayman Brac
(Cayman Islands).
Among the most important islands we identified for
biosecurity, three were inhabited by >1000 people and
have existing populations of rats, feral cats, and green
iguanas (Cayman Brac, Grand Cayman, and Montserrat,
Table 1). However, the small Indian mongoose is so far
absent from those islands and reducing the risk of invasion
of this efficient predator on islands that already have
other harmful invasive species could help secure globally
important biodiversity. Together with Montserrat, Anegada
in the British Virgin Islands was among the top priorities
for biosecurity to reduce the risk of invasion of black rats
and small Indian mongoose, despite both islands also
being a high priority for the eradication of already existing
invasive species (Dawson, et al., 2015).
DISCUSSION
We show that effective biosecurity on islands in the
Caribbean UK Overseas Territories could reduce the risk
of further spread of harmful invasive vertebrates to islands
where globally threatened reptiles and birds would be at
risk. Investing in effective biosecurity procedures and
educating the public and policy makers about the risks
to their national heritage when no biosecurity is in place
should be the immediate next steps of UK and local
governments, private island owners, and international
funding bodies. Our approach offers the guidance to focus
on a limited number of vulnerable islands, as more than
half of the islands we evaluated are not at immediate risk
of further biodiversity loss from the invasion of the five
invasive vertebrate species that we selected.
Similar to other prioritisations identifying islands for
eradication of invasive species (e.g., Harris, et al., 2012;
Dawson, et al., 2015; Stanbury, et al., 2017), our list is
subject to incomplete information about the distribution of
both native and invasive species. The distribution of several
reptile species is poorly documented across many islands
of the Caribbean, and their threat status is also poorly
assessed on the IUCN Red List, both of which may affect
our assessment of their local importance and therefore
introduce bias to our projections of loss in conservation
value (Russell, et al., 2017). Further surveys to increase
the knowledge of native and invasive species on islands
would be beneficial but should not be used as an argument
to delay the immediate adoption of effective biosecurity
protocols to safeguard the most important islands that we
identified.
Besides thorough knowledge about the native and
invasive species occurring on an island, our approach also
requires a classification of the interactions between native
and invasive species. Because these interactions can be
hypothetical for single-island endemic native species that
have not been exposed to invasive species, due caution is
necessary when interpreting the output of our prioritisation.
We used the response of taxonomically similar species to
the same invasive species to predict biologically plausible
consequences of an invasion, but interactions between
native and invasive species are often complex and
unpredictable (Simberloff & Von Holle, 1999; Simberloff,
2006). We encourage researchers to provide robust and
reliable predictions about the potential consequences of
invasions to assist with strategic investment decisions for
reducing the risk of invasive species becoming established
on islands harbouring globally important biodiversity
(Moore, et al., 2010).
In summary, we demonstrated that biosecurity is not
only important on small uninhabited islands or privately
owned tourist resorts where natural habitats remain and
endemic and globally threatened species persist. Even
on large and populated islands such as Grand Cayman,
Cayman Brac, and Montserrat, the invasion of small
Indian mongoose could result in a significant deterioration
of the conservation status of several globally threatened
vertebrates (Hays & Conant, 2007). We therefore urge
local governments, private island owners (e.g. Mosquito
Island) and communities to carefully inspect all incoming
cargo and people and establish ongoing measures to detect
and remove any new invasive species. Training of border
officials and conservation staff, public education and
awareness campaigns targeting the accidental introduction
of invasive species onto uninhabited islands by visiting
people (e.g. fishermen, tourists) should also be implemented,
because international and domestic biosecurity measures
are currently weak across all Caribbean UK Overseas
Territories (Key, 2017; RSPB, 2017). Laws governing
biosecurity measures in the Caribbean UK Overseas
Territories and Bermuda are disjointed, not comprehensive
and scattered through various environmental, agricultural
and customs regulations. Collaboration under existing
national legislative mechanisms may improve the situation
quickly prior to enacting any new legislation (RSPB,
2017). We would also encourage regional collaboration in
developing biosecurity measures, information sharing and
learning from any existing biosecurity initiatives.
ACKNOWLEDGEMENTS
We greatly appreciate the support of all the UK Overseas
Territory governments and BirdLife International partner
organisations for providing data on the distribution of
native and invasive species. We are grateful to Montserrat
National Trust, Anguilla National Trust, National Parks
Trust of the British Virgin Islands, Turks and Caicos
665
666
0
0
200
Cayman
Islands
British Virgin
Islands
British Virgin
Islands
British Virgin
Islands
Bermuda
Grand
Cayman
Salt Island
Carval Rock
Anegada
Nonsuch
Island
0
53,160
0
Anguilla
Little Scrub
Island
0
Cayman Brac Cayman
Islands
British Virgin
Islands
2,098
Anguilla
Sombrero
Mosquito
Island
0
UKOT
Island
Human
popul’n
8.2
3,844.4
1.0
78.2
20,159.4
4.1
49.2
3,889.4
29.2
Island
area (ha)
106.0
11,103.9
0.0
161.7
1,249.5
0.0
50.6
1,364.7
0.0
Current
threat
310.0
11,313.9
242.7
406.6
1,986.5
737.6
978.1
6,879.3
5,729.2
Postinvasion
threat
204.0
210.0
242.7
244.9
737.0
737.6
927.5
5,514.5
5,729.2
Potential
increase in
conservation
threat score
Chelonia mydas, Eretmochelys imbricata,
Dermochelys coriacea, Cyclura pinguis,
Spondylurus anegadae
Plestiodon longirostris, Pterodroma cahow
Brown rat, black rat, feral
cat, green iguana, small
Indian mongoose
Amphisbaena fenestrata, Sphaerodactylus sp.
Eretmochelys imbricata, Chelonia mydas,
Caretta caretta, Dendrocygna arborea,
Crocodylus acutus, Cyclura lewisi, Anolis
conspersus, Crocodylus rhombifer, Typhlops
caymanensis, Tropidophis caymanensis
Eretmochelys imbricata, Spondylurus
semitaeniatus, Spondylurus sloanii,
Amphisbaena fenestrata
Ameiva corax
Black rat, small Indian
mongoose
Brown rat, black rat, feral
cat, green iguana, small
Indian mongoose
Brown rat, black rat, feral
cat, green iguana, small
Indian mongoose
Small Indian mongoose
Brown rat, black rat, feral
cat, green iguana, small
Indian mongoose
Brown rat, black rat, feral
cat, green iguana, small
Indian mongoose
Small Indian mongoose
Sphaerodactylus sp., Ameiva corvina
Brown rat, black rat, feral
cat, green iguana, small
Indian mongoose
Dendrocygna arborea, Crocodylus acutus,
Anolis luteosignifer, Typhlops epactius,
Tropidophis schwartzi, Anolis maynardii,
Celestus maculatus, Cyclura nubila
caymanensis
Spondylurus semitaeniatus, Sphaerodactylus
parthenopion, Cyclura pinguis, Amphisbaena
fenestrata
Globally threatened species at risk from
invasion
Potential invaders
Table 1 The top 25 islands in Caribbean UK Overseas Territories and Bermuda where the invasion of five common vertebrate species could potentially cause the greatest increase
in threats to native biodiversity. Note that some invasive species are already present on some islands (i.e. those with current threat score > 0), and only the potential new invaders
are listed; islands with a current threat score >0 would also benefit from the removal of already existing invasive species. Human population size and island area are provided for
information, as they will affect invasion risk and effort required for biosecurity. The current and post-invasion threat scores are calculated as the sum of all impact scores of invasive
species on all threatened native species present on an island before and after potential invasion.
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
6.4
British Virgin 0
Islands
Montserrat
Round Rock
Montserrat
and
Big Sand Cay Turks
Caicos
7.8
0
Turks and
Caicos
Bush Cay
0
54.5
10157.4
8.5
0
Turks and
Caicos
Fish Cay
4922
19.5
0.9
British Virgin 0
Islands
0
0.3
Fallen
Jerusalem
Bermuda
Inner Pear
Rock
0
102.2
Bermuda
Horn Rock
Island
area (ha)
Virgin 0
Ginger Island British
Islands
UKOT
Island
Human
populn
0.0
12571.3
0.0
0.0
0.0
0.0
0.0
105.3
106.0
Current
threat
131.7
12708.1
140.3
168.4
170.4
194.4
201.3
308.5
310.0
Postinvasion
threat
131.7
136.8
140.3
168.4
170.4
194.4
201.3
203.3
204.0
Potential
increase in
conservation
threat score
Chelonia mydas, Cyclura carinata,
Leiocephalus psammodromus
Green iguana, small Indian
mongoose
Brown rat, black rat, feral
cat, green iguana, small
Indian mongoose
Spondylurus semitaeniatus, Amphisbaena
fenestrata
Eretmochelys imbricata, Cyclura carinata
Eretmochelys imbricata, Cyclura carinata
Eretmochelys imbricata, Spondylurus
semitaeniatus, Amphisbaena fenestrata
Eretmochelys imbricata, Spondylurus
semitaeniatus, Amphisbaena fenestrata
Plestiodon longirostris, Pterodroma cahow
Plestiodon longirostris, Pterodroma cahow
Globally threatened species at risk from
invasion
Dermochelys coriacea, Eretmochelys
imbricata, Chelonia mydas, Caretta caretta,
Turdus lherminieri, Icterus oberi, Diploglossus
montisserrati, Leptodactylus fallax, Anolis
lividus, Mabuya montserratae, Iguana iguana
(Montserrat)
Brown rat, black rat, feral
cat, green iguana, small
Indian mongoose
Brown rat, black rat, feral
cat, green iguana, small
Indian mongoose
Brown rat, black rat, feral
cat, green iguana, small
Indian mongoose
Brown rat, black rat, feral
cat, green iguana, small
Indian mongoose
Brown rat, black rat, feral
cat, green iguana, small
Indian mongoose
Brown rat, black rat, feral
cat, green iguana, small
Indian mongoose
Brown rat, black rat, feral
cat, green iguana, small
Indian mongoose
Potential invaders
Table 1 (continued) The top 25 islands in Caribbean UK Overseas Territories and Bermuda where the invasion of five common vertebrate species could potentially cause the greatest
increase in threats to native biodiversity. Note that some invasive species are already present on some islands (i.e. those with current threat score > 0), and only the potential new
invaders are listed; islands with a current threat score >0 would also benefit from the removal of already existing invasive species. Human population size and island area are provided
for information, as they will affect invasion risk and effort required for biosecurity. The current and post-invasion threat scores are calculated as the sum of all impact scores of invasive
species on all threatened native species present on an island before and after potential invasion.
Oppel, et al.: Prioritising islands for biosecurity
667
668
0
0
0
0
0
0
Turks and
Caicos
Turks and
Caicos
Turks and
Caicos
Turks and
Caicos
Turks and
Caicos
Turks and
Caicos
Six Hills
West Cay
Middleton
Cay
Long Cay
(Turks)
White Cay
Indian Cay
Plandon Cay
Middle Creek Turks and
Cay
Caicos
0
0
UKOT
Six Hills East Turks and
Cay
Caicos
Island
Human
populn
41.6
15.5
2.9
2.4
18.4
4.6
6.2
3.1
Island
area (ha)
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Current
threat
110.3
110.3
111.3
111.8
117.0
121.5
123.0
123.0
Postinvasion
threat
110.3
110.3
111.3
111.8
117.0
121.5
123.0
123.0
Potential
increase in
conservation
threat score
Brown rat, black rat, feral
cat, green iguana, small
Indian mongoose
Brown rat, black rat, feral
cat, green iguana, small
Indian mongoose
Brown rat, black rat, feral
cat, green iguana, small
Indian mongoose
Brown rat, black rat, feral
cat, green iguana, small
Indian mongoose
Brown rat, black rat, feral
cat, green iguana, small
Indian mongoose
Brown rat, black rat, feral
cat, green iguana, small
Indian mongoose
Brown rat, black rat, feral
cat, green iguana, small
Indian mongoose
Brown rat, black rat, feral
cat, green iguana, small
Indian mongoose
Potential invaders
Cyclura carinata
Cyclura carinata
Cyclura carinata
Cyclura carinata
Cyclura carinata, Leiocephalus
psammodromus
Cyclura carinata, Tropidophis greenwayi
Cyclura carinata, Aristelliger hechti
Cyclura carinata, Aristelliger hechti
Globally threatened species at risk from
invasion
Table 1 (continued) The top 25 islands in Caribbean UK Overseas Territories and Bermuda where the invasion of five common vertebrate species could potentially cause the greatest
increase in threats to native biodiversity. Note that some invasive species are already present on some islands (i.e. those with current threat score > 0), and only the potential new
invaders are listed; islands with a current threat score >0 would also benefit from the removal of already existing invasive species. Human population size and island area are provided
for information, as they will affect invasion risk and effort required for biosecurity. The current and post-invasion threat scores are calculated as the sum of all impact scores of invasive
species on all threatened native species present on an island before and after potential invasion.
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
Oppel, et al.: Prioritising islands for biosecurity
National Trust, Bermuda Audubon Society, UKOTA
members, C. Stringer, J. Hall, J. Millett, G. Gerber, and K.
Newton for providing valuable assistance. Two anonymous
reviewers and Dick Veitch provided useful comments to
improve this contribution.
Keitt, B., Holmes, N., Hagen, E., Howald, G. and Poiani, K. (2019).
‘Going to scale: Reviewing where we’ve been and where we need to go
in invasive vertebrate eradications’. In: C.R. Veitch, M.N. Clout, A.R.
Martin, J.C. Russell and C.J. West (eds.) Island invasives: scaling up to
meet the challenge, pp. 633–636. Occasional Paper SSC no. 62. Gland,
Switzerland: IUCN.
REFERENCES
Key, J. (2017). Tackling Invasive Non-native Species in the UK Overseas
Territories – Gap Analysis of Biosecurity Capacity. Sand Hutton, UK:
Animal and Plant Health Agency.
Atkinson, I.A.E. (1985). ‘The spread of commensal species of Rattus to
oceanic islands and their effects on island avifaunas’. In: P.J. Moors
(ed.) Conservation of Island Birds. pp. 35–81. Cambridge, United
Kingdom: International Council for Bird Preservation.
Barun, A., Budinski, I. and Simberloff, D. (2008). ‘A ticking time-bomb:
The small Indian mongoose in Europe’. Aliens 26: 14–16.
Brooke, M.d.L., Hilton, G.M. and Martins, T.L.F. (2007). ‘Prioritizing
the world’s islands for vertebrate-eradication programmes’. Animal
Conservation 10: 380–390.
Broome, K. (2007). ‘Island biosecurity as a pest management tactic in
New Zealand’. In: G. Witmer, W. Pitt, K. Fagerstone and C. Clark (eds.)
Proceedings of an International Symposium Managing Vertebrate
Invasive Species. pp. 104–107. Lincoln: University of Nebraska.
Case, T.J. and Bolger, D.T. (1991). ‘The role of introduced species in
shaping the distribution and abundance of island reptiles’. Evolutionary
Ecology 5: 272–290.
Churchyard, T., Eaton, M.A., Havery, S., Hall, J., Millett, J., Farr, A.,
Cuthbert, R.J., Stringer, C. and Vickery, J.A. (2016). ‘The biodiversity
of the United Kingdom’s Overseas Territories: A stock take of species
occurrence and assessment of key knowledge gaps’. Biodiversity and
Conservation 25: 1677–1694.
Dawson, J., Oppel, S., Cuthbert, R.J., Holmes, N., Bird, J.P., Butchart,
S.H.M., Spatz, D.R. and Tershy, B. (2015). ‘Prioritizing islands for the
eradication of invasive vertebrates in the United Kingdom overseas
territories’. Conservation Biology 29: 143–153.
Forster, J., Lake, I.R., Watkinson, A.R. and Gill, J.A. (2011). ‘Marine
biodiversity in the Caribbean UK overseas territories: Perceived threats
and constraints to environmental management’. Marine Policy 35:
647–657.
Game, E.T., Kareiva, P. and Possingham, H.P. (2013). ‘Six common
mistakes in conservation priority setting’. Conservation Biology 27:
480–485.
Gibbon, J.W., Scott, D.E., Ryan, T.J., Buhlmann, K.A., Tuberville, T.D.,
Metts, B.S., Greene, J.L., Mills, T., Leiden, Y. and Poppy, S. (2000).
‘The global decline of reptiles, Déjà Vu amphibians’. BioScience 50:
653–666.
Greenslade, P., Burbidge, A.A. and Lynch, A.J.J. (2013). ‘Reply
to: Protecting islands from pest invasion: Optimal allocation of
biosecurity resources between quarantine and surveillance’. Biological
Conservation 157: 434.
Harris, D., Gregory, S., Bull, L. and Courchamp, F. (2012). ‘Island
prioritization for invasive rodent eradications with an emphasis on
reinvasion risk’. Biological Invasions 14: 1251–1263.
Hays, W.S.T. and Conant, S. (2007). ‘Biology and impacts of Pacific
Island invasive species. 1. A worldwide review of effects of the small
Indian mongoose, Herpestes javanicus (Carnivora : Herpestidae)’.
Pacific Science 61: 3–16.
Hedges, S.B. (2017). Caribherp: West Indian Amphibians and Reptiles
(<www.caribherp.org>).
Philadelphia,
Pennsylvania:
Temple
University.
Helmstedt, K.J., Shaw, J.D., Bode, M., Terauds, A., Springer, K.,
Robinson, S.A. and Possingham, H.P. (2016). ‘Prioritizing eradication
actions on islands: It’s not all or nothing’. Journal of Applied Ecology
53: 733–741.
Hilton, G.M. and Cuthbert, R. (2010). ‘The catastrophic impact of
invasive mammalian predators on birds of the UK Overseas Territories:
a review and synthesis’. Ibis 152: 443–458.
Holmes, N., Campbell, K., Keitt, B., Griffiths, R., Beek, J., Donlan,
C. and Broome, K. (2015). ‘Reporting costs for invasive vertebrate
eradications’. Biological Invasions 17: 2913–2925.
IUCN (2017). IUCN Red List of Threatened Species. Gland, Switzerland:
IUCN – The World Conservation Union. <http://www.iucnredlist.
org/>.
Jones, H.P., Tershy, B.R., Zavaleta, E.S., Croll, D.A., Keitt, B.S.,
Finkelstein, M.E. and Howald, G.R. (2008). ‘Severity of the effects of
invasive rats on seabirds: A global review’. Conservation Biology 22:
16–26.
Key, G.E. and Moore, N.P. (2019). ‘Tackling invasive non-native species
in the UK Overseas Territories’. In: C.R. Veitch, M.N. Clout, A.R.
Martin, J.C. Russell and C.J. West (eds.) Island invasives: scaling up to
meet the challenge, pp. 637–642. Occasional Paper SSC no. 62. Gland,
Switzerland: IUCN.
Martins, T.L.F., Brooke, M.d.L., Hilton, G.M., Farnsworth, S., Gould,
J. and Pain, D.J. (2006). ‘Costing eradications of alien mammals from
islands’. Animal Conservation 9: 439–444.
McCreless, E.E., Huff, D.D., Croll, D.A., Tershy, B.R., Spatz, D.R.,
Holmes, N.D., Butchart, S.H.M. and Wilcox, C. (2016). ‘Past and
estimated future impact of invasive alien mammals on insular threatened
vertebrate populations’. Nature Communications 7: 12488.
Medina, F.M., Bonnaud, E., Vidal, E., Tershy, B.R., Zavaleta, E.S.,
Josh Donlan, C., Keitt, B.S., Corre, M., Horwath, S.V. and Nogales,
M. (2011). ‘A global review of the impacts of invasive cats on island
endangered vertebrates’. Global Change Biology 17: 3503–3510.
Moore, J.L., Rout, T.M., Hauser, C.E., Moro, D., Jones, M., Wilcox, C.
and Possingham, H.P. (2010). ‘Protecting islands from pest invasion:
optimal allocation of biosecurity resources between quarantine and
surveillance’. Biological Conservation 143: 1068–1078.
Nogales, M., Vidal, E., Medina, F.M., Bonnaud, E., Tershy, B.R.,
Campbell, K.J. and Zavaleta, E.S. (2013). ‘Feral cats and biodiversity
conservation’. BioScience 63: 804–810.
Oppel, S., Beaven, B., Bolton, M., Vickery, J.A. and Bodey, T.W. (2011).
‘Eradication of invasive mammals on islands inhabited by humans and
domestic animals’. Conservation Biology 25: 232–240.
Powell, R. (2004). ‘Conservation of iguanas (Iguana delicatissima and I.
iguana) in the Lesser Antilles’. Iguana 11: 239–246.
RSPB (2017). Assessment of legislation supporting biosecurity in the
Caribbean UK Overseas Territories and Bermuda. Sandy, Bedfordshire,
UK: Royal Society for the Protection of Birds.
Russell, J.C., Towns, D.R. and Clout, M.N. (2008). Review of Rat invasion
Biology: Implications for Island Biosecurity. Science for Conservation
286. Wellington, New Zealand: Department of Conservation.
Russell, J.C., Meyer, J.-Y., Holmes, N.D. and Pagad, S. (2017). ‘Invasive
alien species on islands: Impacts, distribution, interactions and
management’. Environmental Conservation 44: 359–370.
Simberloff, D. and Von Holle, M. (1999). ‘Synergistic interactions of
nonindigenous species: invasional meltdown’. Biological Invasions 1:
21–32.
Simberloff, D. (2006). ‘Risk assessments, blacklists, and white lists
for introduced species: Are predictions good enough to be useful?’.
Agricultural and Resource Economics Review 35: 1–10.
Spatz, D.R., Holmes, N.D., Reguero, B.G., Butchart, S.H.M., Tershy,
B.R. and Croll, D.A. (2017). ‘Managing invasive mammals to conserve
globally threatened seabirds in a changing climate’. Conservation
Letters 10: 736–747.
Stanbury, A., Thomas, S., Aegerter, J., Brown, A., Bullock, D., Eaton, M.,
Lock, L., Luxmoore, R., Roy, S., Whitaker, S. and Oppel, S. (2017).
‘Prioritising islands in the United Kingdom and crown dependencies
for the eradication of invasive alien vertebrates and rodent biosecurity’.
European Journal of Wildlife Research 63: 31.
Stephen, C.L., Reynoso, V.H., Collett, W.S., Hasbun, C.R. and Breinholt,
J.W. (2013). ‘Geographical structure and cryptic lineages within
common green iguanas, Iguana iguana’. Journal of Biogeography 40:
50–62.
Towns, D., Atkinson, I. and Daugherty, C. (2006). ‘Have the harmful
effects of introduced rats on islands been exaggerated?’. Biological
Invasions 8: 863–891.
Turbelin, A.J., Malamud, B.D. and Francis, R.A. (2017). ‘Mapping the
global state of invasive alien species: patterns of invasion and policy
responses’. Global Ecology and Biogeography 26: 78–92.
Vuillaume, B., Valette, V., Lepais, O., Grandjean, F. and Breuil, M.
(2015). ‘Genetic evidence of hybridization between the endangered
native species Iguana delicatissima and the invasive Iguana iguana
(Reptilia, Iguanidae) in the Lesser Antilles: management implications’.
PLOS ONE 10: e0127575.
Kairo, M., Ali, B., Cheesman, O., Haysom, K. and Murphy, S. (2003).
Invasive Species Threats in the Caribbean Region. Arlington, VA: The
Nature Conservancy.
669
J. Pearson, P. St Pierre, L. Lock, P. Buckley, E. Bell, S. Mason, R. McCarthy, W. Garratt, K. Sugar and J. Pearce
Pearson, J.; P. St Pierre, L. Lock, P. Buckley, E. Bell, S. Mason, R. McCarthy, W. Garratt, K. Sugar and J. Pearce. Working with the local
community to eradicate rats on an inhabited island: securing the seabird heritage of the Isles of Scilly
Working with the local community to eradicate rats on an inhabited
island: securing the seabird heritage of the Isles of Scilly
J. Pearson1, P. St Pierre1, L. Lock1, P. Buckley1, E. Bell2, S. Mason3, R. McCarthy4, W. Garratt5, K. Sugar6 and J. Pearce7
Royal Society for the Protection of Birds, UK Headquarters, The Lodge, Sandy, Bedfordshire SG19 2DL, United
Kingdom. <jaclyn.pearson@rspb.org.uk>.2Wildlife Management International Ltd, PO Box 607, Blenheim, 7240, New
Zealand. 3Isles of Scilly Wildlife Trust, Trenoweth, St. Mary’s, Isles of Scilly, TR21 0NS, United Kingdom. 4Lowertown
Cottage, St Agnes, Isles of Scilly, TR22 0PL, United Kingdom. 5Duchy of Cornwall, Hugh House, St Mary’s, Isles of
Scilly, TR21 0HU, United Kingdom. 6Natural England, Polwhele, Truro, Cornwall, TR4 9AD, United Kingdom. 7Isles of
Scilly Area of Outstanding Natural Beauty, Council of the Isles of Scilly, Town Hall, St Mary’s, Isles of Scilly,
TR21 0LW, United Kingdom.
1
Abstract The inhabited Isles of Scilly, 45 km off the south-western tip of the UK, are home to 13 seabird species
including European storm petrel (Hydrobates pelagicus) and Manx shearwater (Puffinus puffinus), for which the UK has
a global responsibility. Between 1983 and 2006, the overall seabird population in Scilly declined by c.25%. This decline
triggered the establishment of the Isles of Scilly Seabird Recovery Project, a partnership with the aims to reverse seabird
decline and engage the local community and visitors in conserving Scilly’s seabird heritage. The eradication of brown
rats (Rattus norvegicus) from St Agnes and Gugh represented the result of over a decade of preparatory work, involving
raising awareness and gaining 100% support from the community. The two islands are home to 85 people. Therefore
additional, and somewhat unusual, preparations were required (including clearing sheds, communicating with school
children and taking precautions to ensure the safety of pets) during the ground-based baiting operation. In 2016 St Agnes
and Gugh were officially declared ‘rat-free’, meaning worldwide this is one of the largest community-based eradications
to have been successful. Biosecurity on inhabited islands is complex, so to ensure the project’s sustainability, efforts have
been community-led. The community has taken ownership of protecting its seabirds, with 100% saying rat removal and
the subsequent increase in seabirds has had, or will have, a positive effect on ecotourism, a key source of income for the
islands. No less than 68% of the community said their businesses have directly benefited. This project represents a case
study for other community-based projects, showcasing how eradications can gain community support and benefit both
wildlife and human populations.
Keywords: biosecurity, brown rat, eradication, Gugh, inhabited, public support, St Agnes
INTRODUCTION
The eradication of invasive species from islands has
become one of the most important tools for biodiversity
conservation but it can also improve local socioeconomics, human health and ecotourism. Rodents have
been successfully eradicated from islands throughout the
world, including a number of UK islands (Bell, et al.,
2000; Zonfrillo, 2001; Towns & Broome, 2003; Bell,
2004; Howald, et al., 2007; Bell, et al., 2011; Thomas,
et al., 2017; Bell, 2019). Most of these islands have been
uninhabited and many consider that islands with significant
human populations, an unreceptive local community or
occurrence of livestock and domestic animals are unlikely
to be feasible for eradication (Campbell, et al., 2015).
Given that an increasing number of eradications are
being investigated on inhabited islands, the importance of
the engagement and inclusion of local communities has
been highlighted in a number of recent eradication and
research projects (Oppel, et al., 2010; Bryce, et al., 2011;
Eason, et al., 2011, Walsh, et al., 2019). The opinions and
safety of the local community need to be a priority in any
eradication planned for inhabited islands (Stanbury, et al.,
2017). Without compliance of the full community, access
to properties may be denied which may result in the failure
of eradicating every rodent or following the eradication,
community members may compromise ongoing biosecurity
measures.
Human activities can affect the success of an eradication
campaign, particularly waste management, food storage,
buildings harbouring rat nesting materials, and limited
access to certain areas of the island. On the inhabited UK
islands where previous eradications have been completed,
they have been staffed by personnel working for the owners
of the island, for example Lundy, UK (Bell, 2004) and
Isle of Canna, UK (Bell, et al., 2011) whereby the parties
involved are working within the confines of employment
contracts. This is not the case with community members.
Other wildlife control projects may have seen decisionmakers ‘persuade the community’ to accept their decision,
e.g. the delayed rodent eradication programme for Lord
Howe Island (Australia) whereby many inhabitants felt
excluded from initial planning (Crowley, et al., 2017b).
The purpose of this paper is to set out the community
involvement through the various stages of the Isle of Scilly
Seabird Recovery Project, how the views of the local
community were collected and used in the design and
delivery of the project to establish and maintain community
support and evaluate how successful the project was in
achieving this.
Background to ‘Isles of Scilly Seabird Recovery
Project’
The Isles of Scilly are 45 km off the southwest tip of
the UK (Fig. 1). As an island group, they are made up
of five inhabited islands (St Mary’s, St Martin’s, Tresco,
Bryher and St Agnes and Gugh) and up to 190 uninhabited
islets and stacks (1,641 ha, Parslow, 2007). The Isles
of Scilly are nationally important for many species of
seabirds, supporting 20,000 birds of 13 native species
including the burrow-nesting species Manx shearwater
(Puffinus puffinus) and European storm petrel (Hydrobates
pelagicus) (Lock, et al., 2006). Declines of 25% had raised
significant conservation concerns about the future of the
seabirds on the islands. The Isles of Scilly ‘Seabird Liaison
Group’ (SLG) is a partnership between Royal Society for
the Protection of Birds (RSPB), Natural England, Isles
In:
670C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 670–678. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Pearson, et al.: Working with the community on the Isles of Scilly
of Scilly Wildlife Trust (IOSWT), Area of Outstanding
Natural Beauty (AONB) and Isles of Scilly Bird Group,
working within the ‘Isles of Scilly Seabird Conservation
Strategy’ since 2006 (Lock, et al., 2006; Lock, et al., 2009;
St Pierre, et al., 2014). This strategy describes the status
and context of the seabird populations on the Isles of Scilly
and identifies priority actions including current and future
measures to improve the available habitat for seabirds
(Lock, et al. 2006; Lock, et al. 2009; St Pierre, et al.,
2014). The eradication of brown rats (Rattus norvegicus)
from St Agnes and Gugh was identified as a priority
action to remove the threat of mammalian reinvasion on
the neighbouring uninhabited island of Annet and provide
the opportunity for Manx shearwaters and storm petrels to
breed successfully once St Agnes and Gugh were cleared
of rats.
St Agnes (105 ha) and Gugh (37 ha) are two islands
connected by a rock and sand bar at low tide and are
separated from the island of St Mary’s by a deep,1 km wide
channel. The main habitats are farmland, ponds, maritime
heathland and grassland, rocky shores and sandy beaches
(Parslow, 2007). Non-native Pittosporum crassifolium and
Coprosma repens were introduced as part of the flower
farming industry as shelter hedges in the late 1800s. There
is a pub, a Post Office and shop, two cafes, a campsite
and two community halls. Brown rats were accidentally
introduced to the Isles of Scilly from shipwrecks in the
1700s and were widespread and abundant across both
islands (McCann, 2005). The ‘community’ of St Agnes is
defined as the 85 residents who live full time on St Agnes,
plus two part-time residents who live on Gugh for six
months of the year in holiday homes.
Prior to and during the period of the ‘Isles of Scilly
Seabird Conservation Strategy 2006–2013’, seabird
conservation awareness activities were delivered on the
islands through community engagement by RSPB, IOSWT
and AONB. These activities were delivered through press
releases, articles and presentations updating residents
on the outcomes of annual seabird monitoring surveys,
seabird youth education, advocacy at the island fetes and
beach cleans. These activities represented a decade of
preparatory work enabling the community of St Agnes and
Gugh to learn about and take pride in protecting its seabird
heritage.
Fig. 1 Map of the Isles of Scilly, 45 km south-west off the
tip of the UK.
In 2010, the SLG held a workshop on St Mary’s, to
initially obtain the views of residents on all inhabited
islands regarding options for control, eradication and
the importance of seabirds. This workshop provided the
mandate for the SLG to commission a detailed assessment
into the feasibility of eradicating brown rats from St Agnes
and Gugh (Bell, 2011a). Due to eradication projects failing
on other inhabited island elsewhere in the world (Oppel,
et al., 2010), SLG required the feasibility assessment to
include social and economic evaluation. It was not known
how the community would feel about eradicating rats;
whether they would feel the proposed action necessary,
or how they would evaluate social, economic and health
benefits of such an operation. If a person’s values and
sensitivities are dismissed, then they will not engage
with operational processes which can jeopardise the
whole project. The assessment had to focus on obtaining
the opinions of all community members. The feasibility
assessment was conducted by Wildlife Management
International Ltd (WMIL; Bell, 2011a; 2011b)
The ‘vision’ for the project primarily focused on
protecting Manx shearwaters and storm petrels, because rat
eradication was considered the only land-based option that
would feasibly increase the abundance of these species.
MATERIALS AND METHODS
Feasibility phase
The community firstly needed to understand that control
of rats was not an option and that eradication was only a
viable goal if all parties worked together. The feasibility
study therefore set out to ascertain each resident’s opinion
on whether they would support the eradication of rats,
what benefits they would expect for themselves and the
community, and what would motivate them to keep the
islands rat-free. Face-to-face interviews using a standard
questionnaire were conducted with all adults on St Agnes
and Gugh. Controversial topics are often better received
if personnel are open to discuss less positive outcomes,
acknowledging inherent risks and ethical challenges as
it allows questions to be voiced and addressed from the
outset (Crowley, et al., 2017a).The risks stated were (a)
inconvenience (e.g. temporary or long-term changes in
waste disposal, pet and livestock treatment), (b) time away
from other activities due to volunteer participation during
eradication and long term biosecurity and monitoring, (c)
adjustment to new regulations (e.g. undergoing rodenticide
training), (d) that economic benefits may take time and
only apply to some community members, and (e) funding
for eradication may come from grant funding, which
communities may feel reduces the availability of financial
resources for alternative projects.
In order to make their own decisions, community
members each needed to have a full understanding of
the technical aspects of the rat removal operation, and
what their personal role in the project could potentially
be. The feasibility assessment incorporated two general
community workshops; a combined meeting with the
six farmers to discuss the eradication in detail, covering
aspects that were specifically relevant to stock, crops
and farms as well as face-to-face meetings on each farm;
visits to St Agnes School; and face-to-face meetings with
representatives from each household. Every resident was
asked to provide full details of their willingness to support
a potential eradication and any stipulations they had.
Achieving complete rat eradication was only part of the
process, the legacy of the project was to keep the islands
rat-free in perpetuity. The feasibility study therefore also
set out to ascertain the willingness of each community
member to carry out biosecurity measures in the long term.
671
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
Interim phase
While RSPB and AONB continued to deliver education
work to invest in on-the-ground community relationships
between June 2011 and January 2013, a Project Steering
Group and Communications Group were formed.
Start of the project; preparation for ‘rat-removal
ready’ phase
When funding was confirmed, the five-year ‘Isles of
Scilly Seabird Recovery Project’ (IOSSRP) was launched.
Two staff members were employed by RSPB, providing
continuity for the community at each phase. Through a
competitive tender process, WMIL were the successful
contractors for rat eradication. Community conservation
actions at this stage were named ‘rat-removal ready’
actions and were focused on reducing potential rat food and
harbourage to a minimum, so rats could be easily detected
and take bait when the eradication phase commenced.
The IOSSRP recognised the importance of monitoring
the response of other species on St Agnes and Gugh
following the eradication of brown rats and implemented
a monitoring programme for birds, mammals (shrew and
rabbits), invertebrates and vegetation. This work was
completed under contract by Spalding Associates. Most
species benefit from rat eradications on islands, but there
have also been unforeseen and negative impacts recorded
in several projects around the world (Courchamp, et al.,
2003; Towns, et al., 2006; Bell, et al., 2011).
Eradication and short-term monitoring phase
The eradication delivered by WMIL, was a groundbased bait station operation using rodenticide over winter
when natural food was minimal (Bell, 2019). Monitoring
tools were used to detect any rats not taking bait or
avoiding bait stations. Community members were required
to assist WMIL with specific eradication activities such as
checking bait stations in their own homes and reporting
rat sightings. During the eradication phase, WMIL
and IOSSRP personnel built good relationships with
community members to create the best foundation for wellcoordinated actions for on-going biosecurity and potential
incursions in the future.
Post eradication monitoring and final check phase
WMIL produced a Biosecurity Plan and returned for
a six-week ‘final check’ phase in winter 2016. IOSSRP
personnel trained community members to assist with the
monthly checks of the permanent biosecurity stations
and surveillance after a ‘rat on a rat’ (ROAR) call (a 24hour hotline based at IOSWT where anyone can report
rat sign or a suspected rat sighting). A ROAR required a
monitoring grid extending 300 m in all directions from
the sighting spot with daily checks for a month (and was
removed when no evidence of a rat was detected). During
the ‘final check’, questionnaires as part of semi-structured
interviews were carried out. The community questionnaire
consisted of 22 socio-economic evaluation questions, 14
delivery questions and eight biodiversity questions. Semistructured interviews represented feedback from the full
population of St Agnes and Gugh. Qualitative analysis
was deemed the best fit as interviews allow each person
to express themselves, including personal narrative, and
common themes can emerge (Crowley, et al., 2017a). It
is known that successful eradication of rodents has turned
some islands into attractions for visitors, facilitating the
establishment of local tourism businesses (Oppel, et al.,
2010). Therefore, specific questions were asked to ascertain
whether tourism or other businesses had benefited on St
Agnes and Gugh following the eradication of rats. During
672
these interviews, community members who were able to
commit to long-term biosecurity actions were registered
with RSPB as Seabird Heritage Volunteers (SHVs) and
were provided with additional training and support to
complete these actions.
Long-term monitoring phase
The SHVs took ownership of their biosecurity roles to
continue to keep the islands rat-free after the formal end of
the IOSSRP project. SHV Coordinators were recruited in
the community to coordinate these community volunteers
and record data from each biosecurity action. An updated
Biosecurity Plan for St Agnes and Gugh was prepared by
IOSSRP with contributions from SHVs. A Maintenance
Plan was written by the partners and the community aspects
were ‘sense-checked’ by the SHVs.
RESULTS
Feasibility phase
All community members valued seabirds and supported
the eradication of rats for the protection of seabirds.
The collective support for the project was not solely for
seabirds but for the added benefits to people (Bell, 2011a;
Bell, 2011b). Rats were having an impact on the livelihood,
health, enjoyment and lifestyle of the local community as
well as the biodiversity of the island (Bell, 2011a; Bell,
2011b). Farmers reported rats were damaging crops and
taking or damaging stock food, fishermen reported rat
damage to lobster pots and nets and the campsite suffered
damage to tents and customer’s food and belongings. Over
¾ of residents reported rats entering their houses. It was
estimated that rats were costing the St Agnes and Gugh
community approximately £15,000 per year (between £10
and £1,000 per household per year), due to purchasing
bait and damage to property and goods (Bell, 2011a; Bell,
2011b).
While explaining that ‘the decision to carry out the
project is yours’, the eradication methodology and actions
were discussed with the community to ensure that they had
all the information needed to make the decision of whether
to proceed with the project or not. This gave the community
an opportunity to air concerns such as finding adequate
funding (86% of residents), incorrect waste management
causing eradication failure (80% of residents) and
community involvement and support (77% of residents).
These concerns were addressed or actions to mitigate
these concerns were outlined including information on
possible funding streams; bespoke waste training at each
property, eatery and farm; provision of rat-proof garbage
bins and composters; revised process for waste collection
and removal to St Mary’s; and the communication strategy
(including a 24-hour call line).
Interim phase
A number of activities were completed during the interim
phase including putting ‘Frequently Asked Questions’
on project partner websites, addressing community and
wider community questions; delivering two press releases;
education and outreach activities on how to detect rats and
shrews and providing funding updates to the community.
Funding applications were completed and included fullycosted mitigation options for identified issues collected
during the feasibility assessment.
Preparation ‘rat-removal ready’ phase
A five-year ‘activity programme’ was developed for
the community and visitors. A full audit of St Agnes and
Gugh was carried out in June 2013 to prepare the islands
Pearson, et al.: Working with the community on the Isles of Scilly
and provide final ‘rat- removal ready’ instructions to
all residents as requested by the community during the
feasibility assessment.
Eradication and short-term monitoring phase
All community members allowed daily access to
property for WMIL personnel to carry out the ground-based
bait station eradication using rodenticide (either Contrac®,
containing the anticoagulant bromadiolone at 0.005% w/w
or Roban Excel®, containing the anticoagulant difenacoum
at 0.005% w/w) in more than 1,000 stations between
October 2013 to March 2014 (Bell, et al., 2019). There was
no rat-sign after three and a half weeks (Bell, et al., 2019).
There were no instances of non-target species being
affected by the bait (Bell, et al., 2019). Nine rats were
picked up above the surface; six of them were discovered
by community members, and eradication personnel
responded immediately by collecting the carcasses (Bell,
et al., 2019). WMIL trained IOSSRP personnel to gain
expertise in eradication techniques, which enabled them to
further support the community for the later phases. WMIL
and IOSSRP personnel delivered the activity programme
which included two community update talks, weekly
update newsletters and school education sessions.
The eradication methods were reviewed throughout
by the Project Steering Group and adaptations were made
when necessary. On farms a number of baiting tunnels
were dislodged by stock (no bait was consumed) and a
number of monitoring stations (i.e. non-toxic flavoured
wax) were eaten by cows, so farmers and WMIL liaised to
organise a rotation of paddocks where cows would graze,
allowing tunnels and monitoring tools to be moved in and
out of these areas at certain times and remain intact (Bell,
et al., 2019).
Post eradication monitoring and final check phase
Monitoring of the key species showed breeding success
for the first time in living memory post-eradication. There
were eight Manx shearwater chicks recorded in 2015 and
32 in 2015. Storm petrels returned to breed in 2016 with
nine breeding pairs recorded. IOSSRP personnel trained
12 community members to assist checking the permanent
monitoring stations and surveillance from ‘rat on a rat’
(ROAR) calls (Fig. 2).
Fig. 2 IOSSRP personnel train the SHV Coordinators in
biosecurity methods. Credit Nick Tomalin.
There were 28 ROAR reports during this posteradication monitoring phase. Community members
assisted the IOSSRP team establish and maintain the ROAR
surveillance grid. After the final check was completed, it
was deemed appropriate to adapt a ROAR response to the
community checking the permanent biosecurity stations
only instead of establishing and maintaining a 300 m wide
monitoring grid (unless additional evidence of a rat was
identified).
The questionnaire responses showed that the entire
community felt the eradication had a positive effect on
the island and the community (Tables 1–7).When asked
what they liked about the project, 31% of the community
enjoyed having the eradication team on the islands, 15%
liked having no rats on the islands any longer, 10% liked the
eradication team and community working together towards
the successful completion of the eradication and 5% liked
how the project worked closely with the St Agnes School
(Table 1). The community thought the project gathered
the island together and allowed everyone to work together
towards a common goal (Table 2). Half the residents felt
that this project had made a positive change to the history
for the island including raising cultural awareness of the
seabirds and their importance to St Agnes and Gugh and
the Isles of Scilly (Table 3). All of the community felt that
the project had benefited the economy of the island, with
several businesses on the island directly benefitting during
and after the eradication (Tables 4 and 5).
Table 1 Response ‘themes’ from St Agnes and Gugh community members (shown as number of people and percent
of the community) to the question ‘What did you like about the project?’
Theme of reply
No.
%
The team being on the islands – nice people to have around
No rats
Team and community working together for a common goal
Team were unobtrusive and respectful which made the experience
enjoyable
The project worked with the school
Manx shearwaters and storm petrels breeding success
Team helped me learn about wider island biodiversity
Learnt about rats and their ecology
The eradication was professionally delivered
Like to see the bait-take in real time and the speed of operation in
daily updates from the team and in newsletters
I was sceptical at the start but was proved wrong, complete
eradication is possible
18
9
6
6
31%
15%
10%
10%
Social (S)
Biodiversity (B) or
Delivery (D) theme
S
B
S
S
5
4
4
2
2
2
8%
7%
7%
3%
3%
3%
S
B
B
S
D
D
1
2%
D
673
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
Table 2 Response ‘themes’ from St Agnes and Gugh
community members (shown as number of people
and percent of the community) to the question ‘Do you
think there have been any positive or negative impacts
to community by the removal of rats from St Agnes and
Gugh?’.
Theme of reply
Negative or no impact
Positive (no further comment)
Positive, community no longer needs to
worry about damage caused by rats
Positive impacts for farms and visitor
accommodation
Positive, the project generated interest in
the community
Positive as it was nice for the
community to have the team on the
islands in winter
Positive, the community was united
and not divided in any way, it was a
community project
Positive, due to the school and children
being involved throughout
Positive, apart from the increase in
rabbits which is negative for farmers
No.
0
16
14
%
0%
28%
24%
12
21%
7
12%
4
7%
3
5%
1
2%
1
2%
Table 3 Response ‘themes’ from St Agnes and Gugh
community members (shown as number of people and
percent of the community) to the question ‘Do you think
there have been any positive or negative impacts to
culture/history by the removal of rats from St Agnes and
Gugh?’
Theme of reply
No impact
Positive, as we are making history here
on St Agnes
Positive impact (no further comment)
No.
29
10
%
50%
17%
7
12%
Positive, culturally we have all worked
together as a community
Positive, raised cultural awareness of
where birds are in our history, memory,
collective consciousness, part birds
played in our community. Better for
historical buildings
Positive, as part of our history that we
bought rats over and now we are putting
our mistake right
Positive, we have better waste
management and awareness of how to
think carefully about staying rat-free
Positive, as the project will reinstate
historical bird lovers
3
5%
2
3%
Positive, we can look back and feel
proud. I have kept all articles about the
project for a community scrapbook to
help us remember details correctly.
674
2
3%
2
3%
2
3%
1
2%
Table 4 Response ‘themes’ from St Agnes and Gugh
community members (shown as number of people
and percent of the community) to the question ‘Do you
think there have been any positive or negative impacts
to economy by the removal of rats from St Agnes and
Gugh?’
Theme of reply
Positive (no further comment)
Positive in respect to what other
community members have told them,
but not personally to them
Farmers and/or fishermen will not lose
profits from rat damage
The project itself brought extra business
to the islands (using accommodation/
shop/boats)
Don’t have to spend money on rat
control and damage
More boating/bird tours
More visitors due to not having rats in
lets/tents
More visitors in the future if we market
the islands as ‘rat- free’
No.
17
%
29%
12
21%
8
14%
6
10%
5
9%
4
7%
3
5%
3
5%
Table 5 Response ‘themes’ from St Agnes and Gugh
community members (shown as number of people and
percent of the community) to the question ‘Has your
business benefited from this project?
Theme of reply
Not applicable
No longer have to worry about rat
damage to any goods
More tourists in holiday lets and
accommodation as a result of media
exposure
Profit in the shop, accommodation
Business is now more hygienic and safe
for visitors without rats
The project team used the boats more,
visitors on wildlife trips have increased
by 200%, there has been more publicity
through the project
Yes (no further comment)
Composter and bins provided by the
project have benefited business
Tourists have a more positive experience
on the islands
Yes, more visitors camping and buying
ice-cream as they know the campsite is
rat-free
The WMIL team using holiday lets
Possible knock-on effect as more visitors
Guests are actively interested in the
project, improving their stay
Team bought eggs
‘Lifelong learning’ has benefited from
walks and talks
No.
19
12
%
32%
20%
4
7%
3
3
5%
5%
3
5%
3
2
5%
3%
2
3%
2
3%
2
1
1
3%
2%
2%
1
1
2%
2%
Pearson, et al.: Working with the community on the Isles of Scilly
Table 6 Response ‘themes’ from St Agnes and Gugh
community members (shown as number of people and
percent of the community) to the question ‘Do you think
there have been any positive or negative impacts to
tourism by the removal of rats from St Agnes and Gugh?
Theme of reply
Positive (no further comment)
Positive in respect to what other
community members have told them,
but not personally to them
Positive, visitors’ experience of the
islands could be negative due to rats in
tents/lets/on beaches
Positive, more birdwatcher and tourists
will visit to see more seabirds
Positive if we market being rat-free
more to visitors
Positive, already heard good feedback
from visitors
Positive, the project has already
promoted the islands as a travel
destination, tourists told me they were
here as they saw the project/islands on
BBC Countryfile
Positive, seabird boat tours have had far
more visitors onboard due to the project,
my business has a 10% increase in
turnover due to the project
No
14
4
%
24%
7%
21
36%
12
21%
4
7%
2
3%
1
2%
1
2%
Long-term monitoring phase
Legacy workshops held in 2016 confirmed the role
of the SHVs in the on-going biosecurity of St Agnes and
Gugh. Quarterly biosecurity monitoring completed by
the community SHVs to date has not detected any rats (J.
Peacock, St Agnes, pers. comm.).
DISCUSSION
Feasibility phase
Community ‘stipulations’ or requirements to address
concerns were developed following the questionnaire and
face-to-face interviews.
The community requested updates on funding
opportunities, waste training and provision of bins and
composters, a bespoke audit of actions to get the islands
‘rat-removal ready’ and clear communication lines
between the eradication team and the community through
community talks, face-to-face dialogue, newsletters and
school education visits.
As most residents had concerns over the health and
safety of the children, a ‘school education day’ was
delivered whereby school children saw snap traps, bait,
tube and lockable bait stations and received training on
how to stay safe (Fig. 3). Concerns about personnel whom
residents didn’t recognise being on their land were resolved
by WMIL suggesting that all personnel wear an identifiable
uniform (i.e. blaze orange hats with the project logo).
Concerns over where the money for travel and subsistence
for the eradication team would be spent were answered by
WMIL assuring residents that much of it would be spent on
St Agnes and Gugh using local providers (i.e. purchasing
milk and eggs from the local farmers and supplies from
the St Agnes Store). Concerns were also expressed over
the potential poisoning of non-target species, particularly
pet cats (24 were present during operation) and dogs (four
were present). The safety of pets is always a concern to
Table 7 Response ‘themes’ from St Agnes and Gugh
community members (shown as percent of the
community) to the question ‘What support will you offer
the project?’ asked in 2010 (during feasibility phase
questionnaires) and 2017 (long-term monitoring phase).
% of community members
changed from ‘No’ in 2010 to
‘Yes’ in 2017 (descending order)
In-kind logistical support
59%
Other (mainly in-kind support such as lifts 58%
in vehicles)
Volunteering time
55%
Training in rodent detection and
41%
identification
Long-term monitoring for rodents
37%
Training in interview and site inspection
30%
procedures and methods
Assisting with any contingency operation
22%
Check for rodent damage to your own
19%
cargo
Written support to decision makers (e.g.
19%
funders, councillors, MPs).
Listed as a reporting location (where any
17%
rat sighting is reported to you for action)
Transporting food to and between islands
17%
in rodent-proof containers
Installing and maintaining a bait station on 13%
your vessel and/or property
Partner to the project
No change
Financial support
No change
Theme of reply
owners, so the mitigation information was provided
sensitively, including explanation of the unlikelihood of
accidental poisoning due to the design of the bait station
and unlikely access to the rodenticide. Pet owners were
given information that the antidote to the anticoagulant
rodenticide (vitamin K injections and tablets) would
be stored on St Agnes, with WMIL personnel being
contactable 24 hours a day throughout the operational
phase to administer the antidote if necessary. Residents
were asked to alert eradication personnel of any dead rats
found above the surface so the carcasses could be retrieved
immediately.
Several residents raised the issue regarding the possible
impacts of rat eradication on the wider ecology of the
islands; in particular in regards to the endemic ‘Scilly
shrew’ consuming bait; rabbits consuming bait during the
operation as well as potentially increasing rapidly after
the eradication; birds eating the bait; cats prey-switching
from rats to other species such as birds. WMIL explained
the long-term monitoring and mitigation options for these
species such as providing diet information for Scilly
shrews (insectivorous diet as opposed to cereal-based
diet); mitigation methods for rabbits including additional
wires on either side of the bait stations to reduce access,
and rabbit control by the community as necessary after
the eradication; mitigation methods for birds including
bait station design preventing access; and mitigation
methods for all non-target species including daily careful
monitoring of bait blocks for signs of non-target species
consumption, re-sighting of bait stations and the use of
‘crow-clips’ (which further prevent entry by birds such
as gulls and corvids). It was recommended that no new
cats come to the island if previous cats were originally
675
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
kept as ‘ratters’, and collars and bells should be used for
all pet cats. Two residents also struggled with the ethical
dilemma of eradicating a species but decided that the
threat to seabirds was of larger concern, and the complete
eradication of rats was the only viable solution to remove
the threat to seabirds on the islands.
The feasibility report (Bell, 2011a; Bell, 2011b)
detailed the ‘technical conservation actions’ required and
confirmed that the entire community on St Agnes and
Gugh were supportive and willing to carry out general and
bespoke actions.
Start of the project; preparation for ‘rat-removal
ready’ phase
As entrance to St Agnes and Gugh via boats is not
regulated by any authority, this presented the highest risk
pathway for biosecurity. Talks to all community members
and the Harbour Users Group (for all boat users on Scilly)
regarding biosecurity requirements and vigilance were
held throughout the project.
Eradication and short-term monitoring phase
The contractors, team members and community
members worked well together to ensure complete
eradication of rats which would be confirmed after a further
two-year check.
Post eradication monitoring and final check phase
Before the eradication phase, the community helped
complete a number of required actions including the
cessation of any baiting for 12 months prior (snap traps
were supplied for local control). Livestock food and
bedding on the six farms was reduced to minimum levels
and rat-proof feed storage systems were implemented. To
ensure there were no areas without bait, livestock pens,
paddocks fences, windbreaks and stone walls were mapped
using GIS to ensure complete bait station coverage. Where
possible, farmers carried out these necessary actions, but
any work not completed was carried out by WMIL and
IOSSRP personnel the month before the eradication.
Residents’ waste management practices were improved
by the provision of new bins and composters as part of ‘Bin
Friendly Days’. ‘Shed clearance days’, ‘beach clean days’
and ‘wood collection and bonfire night’ reduced rat food
and harbourage around the islands. The St Agnes School
held an ‘Apple Day’ to remove wind-fallen apples from
the ground. Rats were trapped for resistance testing to
confirm final bait choice for the eradication. Index trapping
results estimated the rat population on St Agnes and Gugh
to be between 3,000 and 3,500 rats. Any restrictions
or sensitivities in regard to accessing peoples land and
properties was obtained.
Various themes emerged from the post-eradication
interviews which are summarised in Tables 1–7, including:
Social: The entire community felt the project had
positively affected their day-to-day life. A strong theme
was they no longer needed to worry about rats “They used
to be on my mind, worrying about where they are and what
they do”. Most of the community (86%) felt the removal
of rats had improved health due to the reduction of diseases
spread by rats. When asked ‘what did you like most about
the project?’ eleven themes developed with social-themed
responses being most popular (Table 1). When asked ‘what
did you dislike most about the project?’ the answer ‘nothing’
was overwhelmingly the most popular answer with three
other themes (increase in other nuisance species, ethical
dilemma and concern about accidental pet poisoning)
being mentioned, however they felt that each concern had
been mitigated against (Table 1). When asked if the project
had any positive or negative impacts on the community,
100% answered ‘positive’ (Table 2). One theme that stood
out was that ‘the community was united and not divided
Fig. 3 Bait awareness workshop with St Agnes School
children.
Fig. 4 WMIL training community member to store bait box
safely. Credit Alastair Wilson.
676
Pearson, et al.: Working with the community on the Isles of Scilly
in any way, it was a community project’. When asked if
there had been any impacts to culture and history (Table 3),
one person said, “It has raised cultural awareness of where
birds are in our history, memory, collective consciousness
and the part birds played in our community”.
Economy: Again, the entire population felt the project
had benefited the local economy (Table 4), with most of
this benefit to certain sectors; agricultural, fishing and
particularly tourism and that the benefits had potential to
increase. Over two-thirds of the community (68%) felt
that their businesses had benefited from the project (Table
5). A section of the community (17%) had developed new
products; e.g. one farmer explained that ‘Apple day had
been the catalyst to a new apple juice product and cider
products he developed’. Another community member
explained that ‘visitors on his ‘boating wildlife trips’ had
increased by 200%, as there has been high publicity of
the project, combined with interpretation resources, so
he could offer improved tours”. Publicity was an added
benefit to the project, which was not originally anticipated
by the IOSSRP ‘activity programme’. Shows such as BBC
‘Countryfile’, BBC ‘One Show’, BBC ‘Springwatch’ and a
German wildlife show, were viewed by approximately 20
million viewers in total (pers. comms.) and directly led to
increased tourism with one community member saying ‘A
tourist told me they had visited due to seeing the project on
BBC ‘Countryfile’. Tourism generates the largest income
on the island (Blue Sail, 2011), and 100% of the population
felt the project had a positive impact on tourism (Table 6).
Interestingly, once rats had been eradicated, more
residents (94% in 2016 compared to 76% in 2010)
recognised that they had been having a greater issue with
rats than first thought, regarding damage, and on reflection
the cost rats had caused them was revised as being higher
(Table 7).
Biodiversity: Compared to the 2012 questionnaire,
the number of residents being sympathetic to seabirds had
increased by 47% (Table 7). Regarding the wider species
present on St Agnes and Gugh, none of the community felt
that the eradication of rats had any negative impact on any
non-target species.
Project procedures and delivery: All of the
community were happy with the project procedures and
methods (Table 1). When asked if it was helpful having
WMIL team members assisting ‘rat-removal ready’ action
‘shed clearance’ one person said: ‘it generated goodwill in
the community and got everyone on board with the project’.
When asked whether the different communication methods
were correct, the entire community said yes. Common
themes were, ‘clear explanation of what we needed to do
and when’, ‘involved everyone and engagement with all
children at the school’, ‘the team was passionate about the
cause’, ‘we felt listened to, as things were altered if we
asked for them to be’.
The final questions asked what support the community
could give to future biosecurity to keep the islands rat-free.
More residents were willing to offer support compared to
2010 (Table 7). An additional 20 community members
said they would volunteer to assist biosecurity monitoring,
due to being proud of the project and wanting to play their
part to keep the island rat-free. A total of 32 community
members have registered with RSPB as ‘Seabird Heritage
Volunteers’ (SHVs).
Long-term monitoring phase
The role of the SHVs was confirmed as covering five
tasks; (1) checking permanent monitoring stations once a
month; (2) sustaining biosecurity on boats and freight; (3)
carrying out surveillance for potential incursions (within
24 hours of a ‘ROAR’ call’); (4) assisting with incursion
response baiting; and (5) assisting with the ecological
monitoring of the key species.
Each SHV received LANTRA rodenticide training
as well as bespoke training for incursion response
protocols; a social media (Facebook) group was set up as a
mechanism to send monthly check information to the SHV
coordinator; biosecurity protocols were reaffirmed; and
incursion response methodology was revised (i.e. check all
biosecurity stations within 24 hours of a ‘ROAR’ especially
those with the stations nearest to the report location and
report back to the SHV coordinator) and tested by a ‘mock
incursion response’ exercise.
If rat-sign is found at any time in the future, the SHV
coordinator will inform IOSWT on St Mary’s and the
RSPB Conservation Officer in Penzance. The SHVs will
swap monitoring wax for rodenticide in their biosecurity
stations within 24 hours and report any new rat sign. An
RSPB-coordinated incursion team will arrive to assist
incursion response baiting for one month, with the SHVs
assisting where possible.
In addition to biosecurity monitoring, SHVs assist
IOSSRP personnel and IOSWT contractors to survey
Manx shearwater and storm petrel breeding sites (using
play-back at burrows) and ‘evening chick-check walks’.
The IOSWT has committed to fund the work outlined
in the ‘St Agnes and Gugh Maintenance Plan’, including,
but not limited to, ongoing biosecurity training for the
community, seabird surveys and resources required to keep
the biosecurity shed functional.
CONCLUSION
The success of this project was due to three factors;
the vision for the sustainability and legacy of the project
from concept; robust preparation; and being ‘community
based’. Community members joined decision-making
processes from the offset, and in advance of this, a
decade of preparation activities meant relationships had
started to be built and methods on how to protect seabird
heritage had started to be shared. These relationships then
sustained trust through the ‘rat-removal ready actions’ and
eradication phase, enlisting an excellent contractor and
team whom worked with the community addressing all
stipulations, and having available, adaptive project staff
to accommodate community concerns when required.
Community members therefore felt listened to and valued.
The IOSSRP experience shows that, to ensure that
an island restoration project on an inhabited island
runs successfully, the support and agreement from the
community must be secured. It is vital that access to
all properties is obtained to effectively carry out an
eradication. The community must share the project’s
vision and feel that they are one of the beneficiaries. To do
this, they will need to be included in the decision-making
process and management of the project. In this way the
legacy of the project will be much stronger. The larger the
community, the longer, potentially, the project managers
will need to ensure that the residents are all at the same
position of understanding through the various stages
of the project. Archipelagos or groups of islands bring
additional stakeholders and interested parties that need to
be engaged compared to single islands. Ten years is not an
unreasonable timescale depending upon the starting point,
the value placed upon seabirds by the community, and the
strength of the project partnership.
It is important to recognise the social science
requirements for eradications planned on inhabited islands.
The views and concerns of each and every resident and
stakeholder group are important. Community engagement
and consultation should be completed during every stage
of an operation. Most importantly, all aspects of the
eradication should be debated with the community in the
early stages of the proposal. Unlike eradication operators,
most members of the public do not have any knowledge of
677
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
the principles and techniques of an eradication, particularly
in regard to rodenticide choice and operational procedures.
It is important that each community member understands
these aspects and how they will be affected by the day-today operational requirements. A lack of public awareness
about invasive species impacts and misunderstanding
of eradication techniques from island communities are
thought to have been responsible for the opposition of
proposed eradications on inhabited islands around the
world and investing in greater education and consultation
effort can ensure a suitable environment for eradication
projects to proceed (Bryce, et al., 2011).
The additional, and more unusual, preparations
which were required on St Agnes and Gugh (e.g. clearing
sheds, communicating with school children and taking
precautions to ensure the safety of the community’s
pets) were essential and would have contributed to what
was effectively a three and half week eradication period.
Maybe even more importantly, was that these activities
were a possible turning point for the community, where
they recognised what was involved for the whole project
to be successful. The methods used in this project ensured
the community knew that staff and community were part
of a team striving for the same goal, which would be
challenging, but rewarding for birds and people.
The defining factors underpinning the success of
the IOSSRP were the professional management of the
eradication, dedicated and passionate volunteer team
involvement, efficient and systematic monitoring, adapting
to local conditions and ensuring a community-inclusive
approach. The trust and knowledge the community gained
during the preparation and eradication phase paired with
the positive impacts the eradication of rats had on the
seabirds and socio-economics for the community turned
into ‘pride and ownership’ of their project.
ACKNOWLEDGEMENTS
The Isles of Scilly Seabird Recovery Project is a
partnership between the RSPB, the Isles of Scilly Wildlife
Trust, Isles of Scilly Area of Outstanding Natural Beauty,
Natural England and the Duchy of Cornwall. The project
was funded by EU LIFE (Scilly Isles LIFE Project 11 UK
NAT 387), UK’s Heritage Lottery Fund (HG-11-06880)
and the Defra-funded Isles of Scilly Area of Outstanding
Natural Beauty (AONB) Partnership Sustainable
Development Fund. Thanks to all the IOSSRP Seabird
Taskforce Volunteers for their hard work throughout the
project, the IOSSRP Steering Group members for their
support and advice throughout the project and the entire St
Agnes and Gugh community for their support, enthusiasm
and involvement throughout the project and for their
willingness to implement ongoing biosecurity procedures.
REFERENCES
Bell, M.D., Bullock, I. and Humpridge, R. (2000). The Eradication of
Rats from Ramsey Island, Wales. Unpublished technical report to Royal
Society for the Protection of Birds, South-west Regional Office, Devon,
UK. Blenheim, New Zealand: Wildlife Management International Ltd.
Bell, E. (2004). Lundy Seabird Recovery Project 2002–2004: The
Eradication of Black and Brown Rats from Lundy Island. Unpublished
client report to the Royal Society for the Protection of Birds, SouthWest Regional Office, Devon, UK. Blenheim, New Zealand: Wildlife
Management International Ltd.
Bell, E.A. (2011a). Isles of Scilly Seabird Recovery Project: Summary
Report: Improving Rodent Control on Uninhabited Islands, Assessment
of the Feasibility of Rat Removal across the Isles of Scilly Archipelago
and Feasibility of Rat Removal from St Agnes and Gugh. Unpublished
client report to Isles of Scilly Seabird Recovery Project Partnership.
Blenheim, New Zealand: Wildlife Management International Ltd.
Bell, E.A. (2011b). Isles of Scilly Seabird Recovery Project: Part 3:
Assessing the Feasibility of Removing Brown Rats from St Agnes and
Gugh, Isles of Scilly. Unpublished client report to Isles of Scilly Seabird
Recovery Project Partnership. Blenheim, New Zealand: Wildlife
Management International Ltd.
678
Bell, E.A., Boyle, D., Floyd, K., Garner-Richards, P., Swann, B.,
Luxmoore, R., Patterson, A. and Thomas, R. (2011). ‘The groundbased eradication of Norway rats (Rattus norvegicus) from the Isle
of Canna, Inner Hebrides, Scotland’. In: C.R. Veitch, M.N. Clout and
D.R. Towns (eds.) Island invasives: eradication and management, pp.
269–274. Occassional Paper SSC no. 42. Gland, Switzerland: IUCN
and Auckland, New Zealand: CBB.
Bell, E.A. (2019). ‘It’s Not All Up in the Air: the development and use
of ground-based rat eradication techniques in the UK’. In: C.R. Veitch,
M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) Island
invasives: scaling up to meet the challenge, pp. 79–87. Occasional
Paper SSC no. 62. Gland, Switzerland: IUCN.
Bell, E., Floyd, K., Boyle, D., Pearson, J., St Pierre, P., Lock, L., Mason,
S., McCarthy, R. and Garratt, W. (2019). ‘The Isles of Scilly seabird
restoration project: the eradication of brown rats Rattus norvegicus on
the inhabited islands of St Agnes and Gugh, Isles of Scilly’. In: C.R.
Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) Island
invasives: scaling up to meet the challenge, pp. 88–94. Occasional
Paper SSC no. 62. Gland, Switzerland: IUCN.
Blue Sail. (2011). Future of Tourism on Scilly. UK: Blue Sail.
Bryce, R., Oliver, M.K., Davies, L., Gray, H., Urquhart, J. and Lambin,
X. (2011). ‘Turning back the tide of American mink invasion at an
unprecedented scale through community participation and adaptive
management.’ Biological Conservation 144: 575–583.
Campbell, K.J. Beek, J., Eason, C.T., Glen, A.S., Godwin, G., Gould, F.,
Holmes, N.D. Howald, G.R., Madden, F.M., Ponder, J.B., Threadgill,
D.W., Wegmann, A. and Baxter, G.S. (2015). ‘The next generation
of rodent eradications: Innovative technologies and tools to improve
species specificity and increase their feasibility on islands.’ Biological
Conservation 185: 47–58.
Courchamp, F. Chapius, J-L. and Pascal, M. (2003). ‘Mammal invaders
on islands: impacts, control and control impact.’ Biological Review 78:
347–383.
Crowley. S., Hinchliffe. S. and McDonald. R. (2017a). ‘Invasive species
management will benefit from social impact assessment’. Journal of
Applied Ecology 54(2): 351–357.
Crowley. S., Hinchliffe. S. and McDonald. R. (2017b). ‘Conflict in invasive
species management’. Frontiers in Ecology and the Environment 15(3):
133–141.
Eason. C., Miller. A., Ogilvie. S. and Fairweather. A. (2011). ‘An updated
review of the toxicology and ecotoxicology of sodium fluoroacetate
(1080) in relation to its use as a pest control tool in New Zealand’. New
Zealand Journal of Ecology 35(1): 1–20.
Howald, G., Donlan, C.J., Galvan, J.P., Russell, J.C., Parkes, J.,
Samaniego, A., Wang, Y., Veitch, D., Genovesi, P., Pascal, M., Saunders,
A. and Tershy, B. (2007). ‘Invasive rodent eradication on islands’.
Conservation Biology 21: 1258–1268.
Lock, L., Brown, A., Clitherow, J., Mawer, D. and St. Pierre, P. (2006).
‘Isles of Scilly Seabird Conservation Strategy 2006–2009’. Paper
delivered at the Isles of Scilly Seabird Liaison Group, Isles of Scilly,
UK, 20 October 2006.
Lock, L., Brown, A., Webber, J., Mawer, D. and St. Pierre, P. (2009). ‘Isles
of Scilly Seabird Conservation Strategy 2009–2013’. Paper delivered at
the Isles of Scilly Seabird Liaison Group, Isles of Scilly, UK, 25 October
2009.
McCann, J. (2005). ‘The introduction of the brown rat (Rattus
norvegicus)’. Somerset Archaeology and Natural History Journal 2005:
139–141.
Oppel, S., Beaven, B., Bolton, M., Vickery, J., and Bodey, T. (2010).
‘Eradication of invasive mammals on islands inhabited by humans and
domestic animals.’ Conservation Biology 25: 232–240.
Parslow, R. (2007). The Isles of Scilly. Yorkshire, UK: Collins New
Naturalist.
Stanbury, A., Thomas, S., Aegerter, J., Brown, A., Bullock, D., Easton,
M., Lock, L., Luxmoore, R., Roy, S., Whitaker, S. and Oppel, S. (2017).
‘Prioritising islands in the United Kingdom and crown dependencies
for the eradication of invasive alien vertebrates and rodent biosecurity.’
European Journal of Wildlife Research 63: 31.
St Pierre, P., Mason, S., Sugar, K., Buckley, P., Heaney, V., and Pearson, J.
(2014.) ‘The Isles of Scilly Seabird Conservation Strategy 2014–2018’.
Paper delivered at the Isles of Scilly Seabird Liaison Group, Isles of
Scilly, UK.
Thomas, S., Brown, A., Bullock, D., Lock, L., Luxmoore, R., Roy, S.,
Stanbury, A. and Varnham, K. (2017). ‘Island restoration in the UK –
past present and future’. British Wildlife 28: 231–242.
Towns, D. R. and Broome, K. G. (2003). ‘From small Maria to massive
Campbell: Forty years of rat eradications from New Zealand islands’.
New Zealand Journal of Zoology 30: 377–398.
Towns, R.T., Atkinson, I.A.E. and Daugherty, C.H. (2006). ‘Have
the harmful effects of introduced rats on islands been exaggerated?’
Biological Invasions 8: 863–891.
Walsh, A., Wilson, A., Bower, H., McClelland, P. and Pearson, J. (2019).
‘Winning the hearts and minds – proceeding to implementation of the
Lord Howe Island rodent eradication project: a case study’. In: C.R.
Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) Island
invasives: scaling up to meet the challenge, pp. 522–530. Occasional
Paper SSC no. 62. Gland, Switzerland: IUCN.
Zonfrillo, B. (2001). ‘Wildlife Conservation on Ailsa Craig’. The Thomas
Duncan Memorial Lectures. Friends of the McKechnie Institute 22: 1.
J.K. Reaser, G.R. Howald and S.D. Veatch
Reaser, J.K.; G.R. Howald and S.D. Veatch. A plan for the eradication of invasive alien species from Arctic islands
A plan for the eradication of invasive alien species from Arctic islands
J.K. Reaser1, G.R. Howald2,3 and S.D. Veatch1
National Invasive Species Council Secretariat, U.S. Department of the Interior, 1849 C Street NW, Washington, D.C.
20240, USA. <gregg.howald@islandconservation.org>. 2Island Conservation, 2100 Delaware Avenue, Suite 1, Santa
Cruz, CA 95060, USA. 31531 Appleridge Rd. Kelowna, B.C., Canada V1W 3A5.
1
Abstract Invasive alien species represent one of the most significant threats to Arctic ecosystems and their inhabitants.
Rapidly changing environmental conditions and a growing interest in resource extraction, settlement and tourism make
the Arctic region particularly vulnerable to biological invasion. For this reason, invasive alien species are of substantial
concern to the Arctic Council, a multi-national body comprised of Canada, the Kingdom of Denmark (including Greenland
and the Faroe Islands), Finland, Iceland, Norway, Russia, Sweden, and the United States, as well as six international
organisations that represent Arctic indigenous peoples as Permanent Participants. The Arctic Council’s Arctic Invasive
Alien Species (ARIAS) Strategy and Action Plan includes the priority to: “actively facilitate the eradication of invasive
alien species from island ecosystems throughout the Arctic, as well as the recovery of native island species and habitats
that have been impacted by invasive alien species.” A multi-national team of governmental and non-government partners
is collaborating in the development of an action plan (hereafter ‘islands plan’) for the eradication of invasive alien species
from Arctic island ecosystems. The intent of the plan is to provide a vision and strategy for a region-wide approach to
the eradication of island invasive alien species as a multi-national commitment. The islands plan will set forth a strategy
for prioritising island eradications consistent with the growing pressures on ecological and cultural systems. We have
a unique opportunity in the Arctic to take decisive action to prevent and mitigate the adverse impacts of invasive alien
species that plague much of the rest of the world. The eradication of invasive alien species from islands in other parts of
the world provides useful insights into best practices, including approaches to prioritisation and cost-effectiveness.
Keywords: Arctic Council, invasive alien species, non-native species, policy, prioritisation
INTRODUCTION
Throughout the world, invasive alien species have
driven the endangerment and extinction of a wide range of
plants and animals (Wilcove, et al., 1998; McNeely, et al.,
2001; Bellard, et al., 2016), contributed to the degradation
of freshwater, marine, and terrestrial ecosystems (Howard,
1999; Rahel & Olden, 2008; Pejchar & Mooney, 2009) and
hastened the alteration of ecological cycles (Chapin, et al.,
2000; Towns, et al., 2006; Kurle, et al., 2008; Doherty,
et al., 2015). Invasive alien species place constraints on
a wide range of ecosystem services that underpin human
well-being and economic growth, such as pollination, food
and fibre production, disease prevention, climate resilience,
and recreational opportunities (Mack, et al., 2000; Mooney
& Hobbs, 2000; McNeely, 2001; Ehrenfeld, 2010;
Simberloff, 2011). Invasive alien species are regarded as
a threat to national security; in addition to undermining
food, water, and energy security, they may impede military
readiness or cultural survival of native peoples (White
House, 2016).
native to the Arctic. Highly charismatic species include
the polar bear (Ursus maritimus), narwhal (Monodon
monoceros), caribou/reindeer (Rangifer tarandus), and
snowy owl (Bubo scandiaca). The Arctic is characterised
by extreme seasonality; many species migrate long
distances in order to follow resource productivity, some
species by the millions. Although Arctic ecosystems are
low in species richness, abundance is often high (e.g. sea
birds) (Meltofte, 2013; Fernandez, et al., 2014).
Three primary factors make islands particularly
vulnerable to the impacts of invasive species: geographic
isolation, size and high percentage of global biodiversity per
area (Reaser, et al., 2007; Kier, et al., 2009). While relatively
few invasive alien species have been documented in the
Arctic region (Fig.1) and there is currently no systematic
effort to build a comprehensive dataset and thus provide
species lists, biological invasion is expected to increase in
concert with increasing human activity and climate change
(Walther, et al., 2009; Hall, et al., 2010; Bennett, et al.,
2015). The threat that invasive alien species pose to Arctic
island ecosystems is thus of growing concern (Meltofte,
2013). Fortunately, Arctic governments and their partners
still have the opportunity to act decisively to prevent and
mitigate the adverse impacts of invasive alien species that
plague much of the rest of the world.
ECOLOGICAL CONTEXT
More than 21,000 species of mammals, birds, fish,
amphibians, reptiles, invertebrates, plants, and fungi are
Fig. 1 The Arctic Region. There are varying approaches to
defining the Arctic according to geophysical, ecological,
or political criteria. For the purposes of this paper, the
CAFF delineation of the Arctic is used (including 32
million km2).
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 679–686. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
679
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
Invasion pathways of particular concern in the
Arctic include: commercial shipping (i.e. introductions
via ballast water, hull biofouling); the introduction of
organisms and reproductive material through horticulture
and aquaculture activities; large-scale tree planting for
aesthetics, fuel, windbreaks, and carbon sequestration;
transport of contaminated material and equipment for
energy development and mineral exploration; and tourism,
including recreational hunting and fishing (e.g. through
contaminated boats, equipment, and gear). Examples of
other anthropogenic pathways include translocated piers,
docks and pilings, marine debris and the release or escape
of live animals (e.g. from fur farms or the commensal
rodents (Mus spp., Rattus spp.) inadvertently transported
to the islands) (CAFF, 2017). Table 1 provides examples of
specific pathways of introduction, the species introduced
and the implications for the Arctic. At this time, data are
insufficient to develop a comprehensive list of non-native
species in the Arctic.
SOCIO-ECONOMIC AND POLITICAL CONTEXT
Numerous people, those who reside in the Arctic
and many who do not, benefit from the region’s natural
resources. Approximately four million people live in the
Arctic, including indigenous peoples who depend upon
subsistence gathering and harvesting of native species as
a major source of their daily food intake and as a vital
element of their culture. Each year, commercial fisheries
harvest millions of tons of native marine organisms valued
in the billions of U.S. dollars (Christiansen & Reist, 2013;
Sundet, 2014).
Extractive industries (e.g. oil, gas, and minerals) are
already well-established in the region and are expanding
their activities as melting ice makes access to natural
resources more feasible. The increase in rate and numbers
of commercial investments in the Arctic is expected to
increase the risk of biological invasion into and throughout
the region (Emerson & Lahn, 2012; Miller & Ruiz, 2014;
Eguíluz, et al., 2016).
Invasive alien species do not respect jurisdictional
boundaries. Effective communication and collaboration
with neighbouring countries, stakeholders, and trading
partners is of paramount importance in the prevention,
eradication, and control of invasive alien species in
the Arctic. The Arctic Council—a policy framework
that includes Arctic Council member countries (known
as States), Permanent Participants (Arctic indigenous
communities), and Observers (generally, non-member
States)—recognises the connection between economic
well-being, social stability, and environmental health. The
Council actively promotes cooperation, both within the
Arctic and globally, to address the environmental changes
facing the region (Arctic Council, 2013), ideally through
an ecosystem-based approach which balances conservation
and sustainable use of the environment (PAME, 2011).
The Arctic Biodiversity Assessment’s findings (Meltofte,
2013; Box 1) have served as the impetus for the Arctic
Council’s programme of work on invasive alien species. In
May 2017, the Council adopted the Arctic Invasive Alien
Species (ARIAS) Strategy and Action Plan (CAFF, 2017).
This document is a call to action voiced by Arctic nations;
The Arctic Biodiversity Assessment
The Arctic Biodiversity Assessment (Meltofte, 2013) recognises that there are currently few invasive
alien species in the Arctic, and underscores that more are expected with climate change and increased
human activity. Authors recommended:
“Reducing the threat of invasive alien/non-native species to the Arctic by developing and implementing
common measures for early detection and reporting, identifying and blocking pathways of introduction,
and sharing best practices and techniques for monitoring, eradication and control. This includes supporting
international efforts currently underway, for example those of the International Maritime Organization to
effectively treat ballast water to clean and treat ship hulls and drilling rigs. (Recommendation 9)”
Actions for Arctic Biodiversity: Implementing the Actions of the Arctic Biodiversity Assessment 2013–
2021 (CAFF, 2015) sets forth two actions to address Arctic invasive alien species:
Action 9.1 (2015-2017): Develop a strategy for the prevention and management of invasive species
across the Arctic, including the identification and mitigation of pathways of introduction of invasions.
Include involvement of indigenous observing networks, which include invasive and new species reporting,
to assist with early detection.
Action 9.2 (2017-2019): Incorporate common protocols for early detection and reporting of non-native
invasive species in the Arctic into CAFF’s Circumpolar Biodiversity Monitoring Programme (CBMP).
Table 1 Examples of introduction pathways and impacts.
Pathway
Escape from fur farms
Gardening and land
reclamation
Intentional releases into
the natural environment
for food
Intentional releases into
the natural environment
for hunting
680
Species
Impact(s)
American mink
(Mustela vison)
Nootka lupine
(Lupinus
nootkatensis)
Red king crab
(Paralithodes
camtschaticus)
Raccoon dog
(Nyctereutes
procyonoides)
High predation on native species in Iceland and Scandinavia (Birnbaum,
2013)
Successful competition against native plants that has changed the
ecological structure and function in Iceland (Magnusson, 2010)
Effective predation of a wide range of marine species in some Norwegian
fjords (Oug, et al., 2011)
Effective predation of ground-nesting birds and amphibians, and service
as a vector of rabies and other pathogens and parasites in northern
Scandinavia (Sutor, et al., 2010; Kowalczyk, 2014; Dahl & Åhlén, 2016)
Reaser, et al.: Plan for invasives erdication Arctic islands
Table 2 Arctic Invasive Alien Species Strategy and Action Plan priority actions.
Arctic Invasive Alien Species Strategy and Action Plan
1. Inspire urgent and effective action: Raise awareness of the unique opportunity that the Arctic Council and its
partners have to inspire the urgent and effective action necessary to protect the Arctic from invasive alien species.
1.1
Promote and, as needed, develop targeted communications and outreach initiatives to raise awareness of the urgent
need and unique opportunity to protect the Arctic region from the adverse impacts of invasive alien species;
1.2
Encourage Arctic States and non-Arctic States (including Arctic Council Observer States), working collaboratively
with Permanent Participants, to implement effective programmes for preventing the introduction and controlling the
spread of invasive alien species through domestic actions and/or international agreements and relevant guidelines,
such as the International Convention for the Control and Management of Ships’ Ballast Water and Sediments, and
the IMO Guidelines for the control and management of ships' biofouling to minimise the transfer of invasive aquatic
species (Biofouling Guidelines);
1.3
Promote and coordinate the Arctic Council’s work on invasive alien species with relevant scientific, technical, and
policy-making bodies and instruments; and
1.4
Encourage the integration of the outputs of the Arctic Council’s work on invasive alien species into international
efforts and legal and institutional frameworks, especially planning and coordination mechanisms, including at the
national and sub-national levels, where appropriate.
2. Improve the knowledge base for well-informed decision making: Improve the capacity of the Arctic Council
and its partners to make well-informed decisions on the needs, priorities, and options for preventing, eradicating, and
controlling invasive alien species in the Arctic by improving the knowledge base.
2.1
Identify and assess: a) the invasive alien species and pathways that pose the greatest risk of biological invasion into,
within, and out of Arctic ecosystems; b) the Arctic ecosystems, livelihoods, and cultural resources most vulnerable to
biological invasion; and c) the current and projected patterns and trends of introduction and impacts of invasive alien
species in the Arctic;
2.2
Produce a series of topic-specific assessments of invasive alien species issues in the Arctic considering scientific,
Traditional Local Knowledge (TLK), technical, environmental, economic, socio-cultural, legal, and institutional
perspectives;
2.3
Improve the collection of information on the occurrence and impacts of Arctic invasive alien species, taking
advantage of new technologies for early detection, and integrate this information into circumpolar, regional,
and community-based observing networks, monitoring programmes, (in particular the Circumpolar Biodiversity
Monitoring Programme), and associated information systems such as (the Arctic Biodiversity Data Service); and
2.4
Facilitate full, timely, and open sharing of data and other information relevant to Arctic invasive alien species
prevention and management through the Arctic Biodiversity Data Service and the CAFF Web portal.
3. Undertake prevention and early detection/rapid response (EDRR) initiatives: Protect Arctic ecosystems and
human well-being by instituting prevention and early detection/rapid response programmes for invasive alien species
as a matter of priority.
3.1
Collaborate with industries, such as, tourism, energy, fisheries, mining, and shipping, and other stakeholders, as
relevant, to develop and implement a wide range of biosecurity measures for points of entry and along priority
pathways to reduce the initial transfer of species;
3.2
Encourage the establishment of new, or strengthen existing, surveillance, monitoring, reporting, and rapid response
programmes necessary to ensure EDRR at points of entry. Consideration of TLK and community-based monitoring
programmes should be encouraged;
3.3
Encourage the development and sharing of tools to enable EDRR for invasive alien species that may pose a
substantial threat to the Arctic;
3.4
Actively facilitate the eradication of invasive alien species from island ecosystems throughout the Arctic as well as
the recovery of native island species and habitats that have been impacted by those invasive alien species;
3.5
Develop guidance for the use and transfer of native and alien species to and throughout the Arctic environment, and
identify opportunities to foster ecological resistance and resilience to environmental change;
3.6
Collect information on best practices and assess whether there is a need for the International Maritime Organization
to develop Arctic specific guidance for minimising the threat posed by ballast water and biofouling as vectors for the
transfer of aquatic invasive alien species from shipping; and
3.7
Foster development of the innovative research, tools, and technologies needed to advance invasive alien species
prevention and EDRR capacities in the Arctic region, including through support from funding programmes.
681
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
it establishes near-term priorities for securing the future of
the Arctic. These priority actions (Table 2) span terrestrial,
freshwater, and marine ecosystems and take environmental,
cultural and economic factors into consideration. Some of
the priority actions apply to the Arctic Council as a whole,
while others are best addressed at the working group level
or through national implementation. The Conservation
of Arctic Flora and Fauna (CAFF) and Protection of the
Arctic Marine Environment (PAME) working groups of
the Arctic Council hope that each Arctic State, working
collaboratively with its partners, will integrate the actions
from the ARIAS Strategy and Action Plan into national
plans and employ the priority actions. This would enable
the advancement of relevant decisions made under the
auspices of other multi-lateral fora and instruments (e.g. the
Convention on Biological Diversity and the International
Maritime Organization).
The effective implementation of these priority actions
will, of course, depend upon securing the resources
necessary to implement them as a matter of urgency and
upon collaboration with Permanent Participants, nonArctic States (including Arctic Council Observers),
regional and local authorities, industry and all others who
live, work, and travel in the Arctic. Recognition by States,
authorities and external organisations that collaborating
with the Arctic Council provides a collective and highly
desirable benefit will also be crucial. CAFF and PAME
will coordinate implementation under the overall direction
of the Senior Arctic Officials, drawing on other Arctic
Council working groups and partners as needed. Progress
reports will be submitted by CAFF and PAME to the
Senior Arctic Officials and Arctic Council Ministers every
two years.
Although only one of the priority actions set forth in the
ARIAS Strategy and Action Plan (CAFF, 2017) is explicitly
focused on islands, all of the action items are relevant to
protecting island ecosystems. Invasive alien species issues
are inherently context-specific; they change through time
and across landscapes. These particular measures will need
to be tailored to particular pathways, populations of nonnative species, localities, type and scale of impact, and the
available resources.
IMPLEMENTING PRIORITY ACTION
ARIAS Strategy and Action Plan priority action 3.4 calls
for the Arctic Council and its partners to “actively facilitate
the eradication of invasive alien species from island
ecosystems throughout the Arctic as well as the recovery of
native island species and habitats that have been impacted
by those invasive alien species”. The ARIAS Strategy and
Action Plan Steering Committee identified this item as a
priority because:
1. Island species and ecosystems are well documented
as being particularly vulnerable to the impacts of
invasive alien species (per previous discussion in this
paper). Of particular concern are seabird species that
have evolved in the absence of persistent, successful
nest-site predators such as the commensal rodents.
2. The level of biological invasion on Arctic islands is
relatively low. Due to a lack of other confounding
variables, the likelihood for native species/ecosystem
recovery following the eradication of invasive
vertebrates is high.
3. There are already several examples of successful
invasive vertebrate eradications from Arctic islands
(Croll, et al., 2015; Jones, et al., 2016; Brooke, et al.,
2017). Lessons learnt from these initiatives can be
readily applied to future efforts.
682
To date, efforts to eradicate invasive alien species in
the Arctic have been undertaken domestically by the
jurisdictional governing body. Priority action 3.4 sets a
new precedent for invasive alien species management and
creates new opportunities for collaboration, funding, and
planning across the region.
The United States Arctic Invasive Species Working
Group (coordinated by the National Invasive Species
Council (NISC) Secretariat: <www.invasivespecies.gov>)
is exploring opportunities to collaborate with domestic
and international partners to develop and begin to enact an
implementation plan for priority action 3.4. As a minimum,
this will include measures to:
1. Identify relevant data available in the Arctic island
context and make the data available through openaccess information systems, including the Threatened
Island Database (TIB) and Biodiversity Information
Serving Our Nation (BISON) information system.
2. Summarise the available data to generate information
on current knowledge and identify gaps in key
information (data gaps).
3. Develop and execute a strategy for filling data gaps.
4. Create a prioritisation schema for determining which
island eradications will take precedence and why.
5. Using the schema, determine priorities for the
eradication of invasive vertebrates from Arctic
islands based on available information and with input
from the Arctic Council members and other relevant
stakeholders.
6. Based on these priorities, develop an implementation
plan, including a co-financing strategy, and secure
the additional resources necessary to address these
priorities.
7. Implement the eradication plan for the priority
island(s) identified in step 5.
8. As appropriate, develop and implement a recovery
plan for native island species and habitats of concern.
The recovery plan should include a monitoring
programme to enable early detection and rapid
response to any future invasions.
Invasive alien species have only recently become an
issue of concern in the Arctic. Relatively few baseline
data on species presence and impacts are available in
either the continental or island context. In implementing
priority action 3.4, there is a need to start with the basics:
assembling/collecting baseline data and evaluating the
current status and trends of invasive alien species according
to island, species and pathway specific parameters. These
assessments are necessary to enable governments to
set priorities: which islands, where, why, and how? The
findings generated by these assessments can be coupled with
data on changes in human activity patterns and climate to
generate projections of potential future conditions and thus
strengthen and expand the programmes of work necessary
to minimise the risk of impending impacts to Arctic island
ecosystems (see Hendrichsen, et al., 2014; McGeoch, et
al., 2016, for general discussion on assessment needs).
Unfortunately, data collection, sharing, and
standardisation is a substantial challenge to filling
information gaps in the Arctic. To the best of our knowledge,
no one has previously assembled data on invasive alien
species occurrence on Arctic islands, although some
relevant data can be accessed as subsets of data contributed
to national and regional biodiversity information systems
[e.g. Global Biodiversity Information Facility (GBIF)].
Where information is unavailable via publicly accessible
databases or published literature, information will need to
be actively solicited from other available sources, including
Reaser, et al.: Plan for invasives erdication Arctic islands
experts in the field, institutional and/or scientific networks,
and traditional local knowledge.
Islands, in general, offer stronger benefits to eradication
projects given their high biodiversity, high vulnerability and
generally lower risks of reinvasion (compared to non-island
ecosystems) that tend towards lasting eradication success
(Helmstedt, et al., 2016). However, eradication projects and
similar conservation initiatives are proportionately more
expensive on islands than other geographical areas due to
their typically restricted access and lack of infrastructure,
a reality exacerbated in the Arctic (Martins, et al., 2006;
Donlan, et al., 2014). Limited resources, cross-jurisdictional
collaboration, and evolving techniques/technologies define
our capacities to carry out eradication projects. This makes
it very important to strike the right balance between the
biological need for eradication and the feasibility and
sustainability of operations when prioritising locations
(Saunders, et al., 2011; Martinez-Abrain & Oro, 2013).
Defining clear objectives and measures of performance
will be vital in order to effectively and efficiently maximise
the limited available funding. Consequent restoration
efforts, the second half of priority action 3.4, contribute to
the need for an innovative, flexible and integrated portfolio
of eradication actions and strategic planning tools. Both
restoration capabilities and eradication technical abilities
have made exponential progress over the last decades, and
yet accurate inclusion of economic costs when prioritising
project scope remains a challenge due to its complexities
and data gaps that require assumptions and estimates
(Donlan & Wilcox, 2007; Carrion, et al., 2011; Veitch, et
al., 2011).
To date, no comprehensive invasive alien species
eradication prioritisation scheme has been developed
for Arctic islands. Recent studies on the prioritisation of
islands for invasive alien species eradication projects have
highlighted and critiqued approaches to the removal of
invasive alien species on a given island from multi-taxa
and single-species perspectives. Helmstedt, et al. (2016)
highlight the importance of including cost analyses and
consideration of high-risk options or targeted, logistical
options when weighing the risks and benefits of eradication
(Game, et al., 2013; Joseph, et al., 2009). Helmstedt, et al.
(2016) point to the value of learning from successes and
failures, as well as targeting combinations of invasive
alien species, and emphasise three main factors when
determining the conservation benefit of various portfolios
of action: ecological benefit, economic cost and feasibility
of each eradication action. In addition, the study outlines the
importance of cost calculations across combined portfolios
of action in order to determine cost-sharing opportunities.
In the context of the Arctic islands project outlined
above, detailed assessments of invasive alien species
eradication options, cost-sharing opportunities and
logistical feasibility will need to be conducted once the
choice of candidate islands has been narrowed down
with the view of maximising potential ecological and
social benefits. Table 3 provides an overview of relevant
prioritisation criteria to be considered during project
planning and implementation. These criteria are not listed
according to priority. The level of importance will be
assigned during the schema development process.
Translating priorities into action on the ground can
be challenging, but it is a reasonable goal when local
communities, national and local government agencies, and
landowners value the benefits that can be realised from
the eradication of invasive alien species from islands. A
key strategy to successful implementation will be the
development of a “top down/bottom up” approach, where
policy, regulatory, and financial support is in place, and the
local island communities, landowners and agencies begin
investing in the work on the ground. Implementation can
be realised when the “demand” finds the resources, support
and policies to move forward.
Restoration of island ecosystems is only achievable
if adequate and robust funding mechanisms are in place.
Projects and programmes tend to be expensive with a
large upfront investment required, but the financial return
on investment can be high (see Walsh, et al., this 2019).
With greater demands and competition for government
resources, projects tend to be funded one island at a
time. Managers typically rely on blending funding from
multiple grant programmes and through partnerships with
non-governmental organisations, private foundations
and/or philanthropy. This partnership approach to
funding projects can be inefficient, and the opportunity
to investigate partnerships to co-finance and implement
programmatic portfolios is being considered (see Stringer,
et al., 2019). Adequate financing is critical to ensure
long-term sustainability and protection of the investment
to respond to new introductions and facilitate active and
passive restoration.
CONCLUSION
Invasive alien species impacts in the Arctic region have
global implications. Arctic biodiversity is an irreplaceable
asset. To envision the Arctic as ecologically, culturally
and economically sustainable necessitates a focus on the
factors that threaten the region’s environment and human
well-being. Thus, eradicating invasive alien species from
Arctic island ecosystems will have cumulative benefits.
If these islands are protected from invasive alien species,
they may have a greater ability to resist and be resilient to
other potential stressors. The achievements made through
the adoption of the ARIAS Strategy and Action Plan
present a unique opportunity for collaboration, innovation
and collective action across the Arctic at all levels of
governance, from regional to local community scales.
Governments and their partners need to work together to
make the eradication of invasive alien species from Arctic
islands feasible, reduce the risks of future island invasions
through commerce and other pathways by cooperating
in prevention and management efforts across all shared
ecosystems, and address the various factors that make
island ecosystems particularly vulnerable to the adverse
impacts of invasive alien species.
ACKNOWLEDGEMENTS
The authors extend their gratitude to the members of
the ARIAS Strategy and Action Plan Steering Committee,
the United States Arctic Invasive Species Working Group,
and the CAFF and PAME Secretariats for the multi-year
discussions that have facilitated the work proposed here.
This paper draws heavily on the ARIAS Strategy and
Action Plan which is a product of the Arctic Council’s
CAFF and PAME Working Groups. The views expressed
in this publication are solely those of the authors and
do not necessarily reflect the views of the United States
Government, the U.S. Department of the Interior, or the
National Invasive Species Council.
Post Script Since completion of this paper, the
National Invasive Species Council Secretariat and Island
Conservation collaborated in the production of an analysis
of available non-native species data and developed a
preliminary prioritisation schema for Arctic islands. That
report, Data Matters: informing the eradication of invasive
species on islands, is available on the Council’s website
<https://www.doi.gov/sites/doi.gov/files/uploads/data_
matters_island_conservation_report.pdf> and through
Island Conservation (Gregg.howald@islandconservation.
org).
683
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
Table 3 Preliminary factors for consideration in any prioritisation scheme for the Arctic.
Factor
IUCN Red
Listed species
Considerations
This includes migratory bird species and should consider the current and trend status of the IUCN
Red Listed species, the threat level by the target invasive, and the IUCN Red Listed species’
historical recovery status.
Direct and
This is particularly important in understanding how to maximise project-wide benefits that may
indirect
span varying islands and island systems, species, or stages in the invasion process. Direct benefits
benefits
include eliminating the threat or degradation posed by the invasive alien species to targeted native
ecosystems or species. Indirect benefits may include eliminating the threat or degradation posed
to non-targeted ecosystems and species such as those not listed on the IUCN Red List or in other
policy.
Direct and
Eradication projects can have significant negative and unintended impacts to native species from
indirect
the techniques or technologies used, failure of control measures, or greater disruptions to ecosystem
consequences
equilibriums from the removal of an established invasive alien species. It is important to assess the
possibility and probability of potential consequences specific to the prioritisation scheme’s target
goals. Where other factors outweigh foreseen consequences, mitigation or prevention activities will
need to be considered in overall cost and feasibility planning.
Reinvasion
The risks of anthropogenic reinvasion vary between islands depending on which pathways they
potential
connect to, their geographical proximity to other land masses such as those within swimming
distance, the extent of environmental degradation or negative impacts post eradication that affect
the feasibility of reestablishment, among others. This component has significant impacts on the
sustainability and projected costs of a project.
Biological
Biological and ecological vulnerabilities serve as high conservation value components and
and ecological contribute to project feasibility. Vulnerabilities include islands that come in contact with pathways
vulnerability
and the islands’ ecological resiliency capacities to biological invasion and reinvasion which impact
and resiliency
additional prevention and restoration initiatives.
Impacts
This consists of not only the direct and indirect economic impacts that disrupt or limit subsistence
on Arctic
living and local economies, but also the cultural/spiritual aspects of Arctic life that depend on
inhabitants
natural resource identity and use. These considerations in a prioritisation scheme should make use of
Traditional Local Knowledge.
Opportunities
Utilising community management opportunities has the potential to not only cut costs and fill
for community knowledge gaps, but also engage local managers and community members in complementary
management
conservation practices such as early detection and rapid response efforts and restoration projects.
Costs and
This consideration needs to extend beyond the direct monetary losses to include the indirect impacts
impacts on
on economies and labour resources (e.g. reduced yields from natural resources, prevention of
economies
future yields, alterations and reductions in ecosystem services, and market/non-market value losses
(Colautti, et al., 2006).
Feasibility and Feasibility needs to include both the probability of successful eradication and the sustainability of
technology
that success. Technology feasibility/availability will differ between islands, species, and ecosystems
and need to be assessed and prioritised per project proposal.
Political will of Sustained political will plays a significant role in the success of any government funded project.
jurisdiction
When considering a potential site location, island system, or species, it will be important to assess
the political will at each level surrounding the project’s target and objectives.
Gaps in
The Arctic has relatively fewer studies regarding native species, invasive alien species, island
knowledge
vulnerabilities, and future risks of biological invasion that go beyond generalisations on warming
climates and increasing pathways. It is important that these data gaps are recognised throughout the
prioritisation process and adjusted for, where possible.
Climate change Climate change impacts the vulnerability and susceptibility for biological invasion, reinvasion, and
impacts
establishment and should be taken into consideration for the long-term feasibility of an eradication
project. Together, these two issues can result in exacerbated impacts to ecosystem function and
biodiversity (Mooney & Cleland, 2001; Hellman, et al., 2008; Rahel & Olden, 2008).
684
Reaser, et al.: Plan for invasives erdication Arctic islands
REFERENCES
Arctic Council (2013). Vision for the Arctic. Kiruna, Sweden: Arctic
Council.
Bellard, C., Cassey, P. and Blackburn, T.M. (2016). ‘Alien species as a
driver of recent extinctions’. Biology Letters 12: 20150623.
Bennett, J.R., Shaw, J.D., Terauds, A., Smol, J.P., Aerts, R., Bergström,
D.M., Blais, J.M., Cheung, WL., Chown, S.L., Lea, M-A., Nielsen,
U.N., Pauly, D., Reimer, K.J., Riddles., M.J., Snape, I., Stark., J.S.,
Tulloch, V.J. and Possingham, H.P. (2015). ‘Polar lessons learned:
long-term management based on shared threats in Arctic and Antarctic
environments’. Frontiers in Ecology and the Environment 13: 316–324.
Birnbaum, C. (2013). ‘NOBANIS – Invasive Alien Species Fact Sheet:
Neovison vison’. European Network on Invasive Alien Species:
NOBANIS. <http://www.nobanis.org>. Accessed 5 November 2016.
Brooke, M.d.L., Bonnaud, E., Dilley, B.J., Flint, E.N., Holmes, N.D.,
Jones, H.P., Provost, P., Rocamora, G., Ryan, P.G., Surman, C. and
Buxton, R.T. (2017). ‘Seabird population changes following mammal
eradications on islands’. Animal Conservation. DOI: 10.1111/acv.12344.
Helmstedt, K., Shaw, J., Bode, M., Terauds, A., Springer, K., Robinson, S.
and Possingham, H. (2016). ‘Prioritizing eradication actions on islands:
it’s not all or nothing’. Journal of Applied Ecology 53: 733–741.
Hendrichsen, D.K., Åström, J., Forsgren, E. and Skarpaas, O. (2014).
Dispersal pathways for alien species in Norway. NINA Report 1091.
Trondheim, Norway: Norwegian Institute for Nature Research. [In
Norwegian with English summary]
Howard, G.W. (1999). ‘Invasive species and wetlands. Outline of a
keynote presentation to the 7th Conference of the Contracting Parties to
the Convention on Wetlands’, Paper delivered at the 7th Meeting of the
Conference of the Contracting Parties to the Convention on Wetlands,
San José, 10–18 May 1999.
Jones, H.P., Holmes, N.D., Butchart, S.H.M., Tershy, B.R., Kappes,
P.J., Corkery, I., Aguirre-Munoz, A., Armstrong, D.P., Bonnaud, E.,
Burbidge, A.A., Campbell, K., Courchamp, F., Cowan, P.E., Cuthbert,
R.J., Ebbert, S., Genovesi, P., Howald, G.R., Keitt, B.S., Kress, S.W.,
Miskelly, C.M., Oppel, S., Poncet, S., Rauzon, M.J., Rocamora, G.,
Russell, J.C., Samaniego-Herrera, A., Seddon, P.J., Spatz, D.R., Towns,
D.R. and Croll, D.A. (2016). ‘Invasive mammal eradication on islands
results in substantial conservation gains’. Proceedings of the National
Academy of Sciences 113: 4033–4038.
Carrion, V., Donlan, C.J., Campbell, K.J., Lavoie, C. and Cruz, F. (2011).
‘Archipelago-wide island restoration in the Galapagos Islands: reducing
costs of invasive mammal eradication programs and reinvasion risk’.
PLOS One 6: e18835.
Joseph, L.N., Maloney, R.F. and Possingham, H.P. (2009). ‘Optimal
allocation of resources among threatened species: a project prioritization
protocol’. Conservation Biology 23: 328–338.
Chapin, III, F.S., Zavaleta, E.S., Eviner, V.T., Naylor, R.L., Vitousek,
P.M., Reynolds, H.L., Hooper, D.U., Lavorel, S., Sala, O.E., Hobbie,
S.E., Mack, M.C. and Díaz, S. (2000). ‘Consequences of changing
biodiversity’. Nature 405: 234–242.
Kier, G., Kreft, H., Lee, T.M., Jetz, W., Ibisch, P.L., Nowicki, C., Mutke,
J. and Barthlott, W. (2009). ‘A global assessment of endemism and
species richness across island and mainland regions.’ Proceedings of
the National Academy of Sciences 106: 9322–9327.
Christiansen, J.S. and Reist, J. (2013). ‘Fishes’. In: H. Meltofte (ed.)
Arctic Biodiversity Assessment: Status and Trends in Arctic Biodiversity,
pp.192–245. Akureyri, Iceland: Conservation of Arctic Flora and Fauna.
Kowalczyk, R. (2014). ‘NOBANIS – Invasive Alien Species Fact Sheet
– Nyctereutes procyonoides’. European Network on Invasive Alien
Species: NOBANIS. <www.nobanis.org>. Accessed 7 May 2015.
Colautti, R.I., Bailey, S.A., van Overdijk, C.D.A., Amundsen, K.
and MacIsaac, H.J. (2006). ‘Characterised and projected costs of
nonindigenous species in Canada’. Biological Invasions 8: 45–59.
Kurle, C.M., Croll, D.A. and Tershy, B.R. (2008). ‘Introduced rats
indirectly change marine rocky intertidal communities from algae- to
invertebrate-dominated’. Proceedings of the National Academy of
Sciences 105: 3800–3804.
Conservation of Arctic Flora and Fauna (CAFF) (2015). Actions for
Arctic Biodiversity, 2013–2021: Implementing the recommendations of
the Arctic Biodiversity Assessment. Akureyri, Iceland: Conservation of
Arctic Flora and Fauna.
Conservation of Arctic Flora and Fauna (CAFF) (2017). Arctic Invasive
Alien Species Strategy and Action Plan. Akureyri, Iceland: Conservation
of Arctic Flora and Fauna.
Croll, D.A., Newton, K.M., McKown, M., Holmes, N., Williams, J.C.,
Young, H.S., Buckelew, S., Wolf, C.A., Howald, G., Bock, M.F.,
Curl, J.A. and Tershy, B.R. (2015). ‘Passive recovery of an island bird
community after rodent eradication’. Biological Invasions 18: 703–715.
Dahl, F. and Åhlén, P-A. (2016). ‘Abstract: Egg predation by raccoon
dog Nyctereutes procyonoides in the archipelago of northern Sweden’.
Natural Resources and Bioeconomy Studies 21: 25.
Doherty, T.S., Dickman, C.R., Nimmo, D.G. and Ritchie, E.G. (2015).
‘Multiple threats, or multiplying the threats? Interactions between
invasive predators and other ecological disturbances’. Biological
Conservation 190: 60–68.
Donlan, C.J. and Wilcox, C. (2007). ‘Complexities of costing
eradications’. Animal Conservation 10: 154–156.
Donlan, C.J., Luque, G. and Wilcox, C. (2014). ‘Maximizing return
on investment for island restoration and species conservation’.
Conservation Letters 8: 171–179.
Eguíluz, V.M., Fernández-Gracia, J., Irigoien, X. and Duarte, C.M.
(2016). ‘A quantitative assessment of Arctic shipping in 2010–2014’.
Scientific Reports 6: 30682.
Mack, R.N., Simberloff, D., Londsdale, W.M., Evans, H., Clout, M. and
Bazzaz, F.A. (2000). ‘Biotic invasions: causes, epidemiology, global
consequences and control’. Ecological Applications 10: 689–710.
Magnusson, B. (2010). ‘NOBANIS – Invasive alien species fact sheet:
Lupinus nootkatensis’. European Network on Invasive Alien Species:
NOBANIS.
<https://www.nobanis.org/globalassets/speciesinfo/l/
lupinus-nootkatensis/lupinus_nootkatensis.pdf>.
Accessed
19
December 2016.
Martinez-Abrain, A. and Oro, D. (2013). ‘Preventing the development
of dogmatic approaches in conservation biology: A review’. Biological
Conservation 159: 539–547.
Martins, T.L.F., Brooke, M.D.L., Hilton, G.M., Farnsworth, S., Gould,
J. and Pain, D.J. (2006). ‘Costing eradications of alien mammals from
islands’. Animal Conservation 9: 439–444.
McGeoch, M.A., Genovesi, P., Bellingham, P.J., Costello, M.J.,
McGrannachan, C. and Sheppard, A. (2016). ‘Prioritizing species,
pathways, and sites to achieve conservation targets for biological
invasion’. Biological Invasion 18: 299–314.
McNeely, J.A. (ed.). (2001). The Great Reshuffling: Human Dimensions
of Invasive Alien Species. Gland, Switzerland and Cambridge, UK:
IUCN.
McNeely, J.A., Mooney, H.A., Neville, L.E., Schei, P. and Waage, J.K.
(eds.). (2001). A Global Strategy for Invasive Alien Species. Gland,
Switzerland and Cambridge, UK: IUCN.
Ehrenfeld, J.G. (2010). ‘Ecosystem consequences of biological invasions’.
Annual Review of Ecology, Evolution and Systematics 41: 59–80.
Meltofte, H. (ed.). (2013). The Arctic Biodiversity Assessment: Status and
Trends in Arctic Biodiversity. Akureyri, Iceland: Conservation of Arctic
Flora and Fauna.
Emerson, C. and Lahn, G. (2012). Arctic Opening: Opportunity and Risk
in the High North. London, UK: Lloyd’s and Chatham House.
Miller, A.W. and Ruiz, G.M. (2014). ‘Arctic shipping and marine
invaders’. Nature Climate Change 4: 413–416.
Fernandez, L., Kaiser, B. and Vestergaard, N. (eds.) (2014). Marine
Invasive Species in the Arctic. TemaNord 2014: 547. Copenhagen,
Denmark: Nordic Council of Ministers.
Mooney, H.A. and Hobbs, R.J. (eds.). (2000). Invasive Species in a
Changing World. Washington, DC: Island Press.
Game, E.T., Kareiva, P. and Possingham, H.P. (2013). ‘Six common
mistakes in conservation priority setting’. Conservation Biology 27:
480–485.
Mooney, H.A. and Cleland, E.E. (2001). ‘The evolutionary impacts of
invasive species’. Proceeding of the National Academy of Sciences 98:
5446–5451.
Hall, C.M., James, M. and Wilson, S. (2010). ‘Biodiversity, biosecurity
and cruising in the Arctic and sub-Arctic’. Journal of Heritage Tourism
5: 351–364.
Oug, E., Cochrane, S., Sundet, J.H., Norling, K. and Nilsson, H.C.
(2011). ‘Effects of invasive red king crab (Paralithodes camtschaticus)
on soft-bottom fauna in Varangerfjorden, northern Norway’. Marine
Biodiversity 41: 467–479.
Hellman, J.J., Byers, J.E., Bierwagen, B.G. and Dukes, J.S. (2008).
‘Five potential consequences of climate change for invasive species’.
Conservation Biology 22: 534–543.
Pejchar, L. and Mooney, H.A. (2009). ‘Invasive species, ecosystem
services, and human livelihoods’. Trends in Ecology and Evolution 24:
497–504.
685
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
Protection of the Arctic Marine Environment (PAME) (2011). EcosystemBased Management in the Arctic. Akureyri, Iceland: Arctic Council.
Rahel, F. and Olden, J. (2008). ‘Assessing the effects of climate change on
aquatic invasive species’. Conservation Biology 22: 521–533.
Reaser, J.K., Meyerson, L.A., Cronk, Q., De Poorter, M., Eldrege, L.G.,
Green, E., Kairo, M., Latasi, P., Mack, R.N., Mauremootoo, J., O'Dowd,
D., Orapa, W., Sastroutomo, S., Saunders, A., Shine, C., Thrainsson,
S. and Vaiutu, L. (2007). ‘Ecological and socioeconomic impacts of
invasive species in island ecosystems’. Environmental Conservation
34: 98–111.
Saunders, A., Parkes, J., Aguirre-Munoz, A. and Morrison, S. (2011).
‘Increasing the return on investments in island restoration’. In: C.R.
Veitch, M.N. Clout and D.R. Towns. (eds.) Island invasives: eradication
and management, pp. 492–495. Occassional Paper SSC no. 42. Gland,
Switzerland: IUCN and Auckland, New Zealand: CBB.
Simberloff, D. (2011). ‘How common are invasion-induced ecosystem
impacts?’ Biological Invasions 13: 1255–1268.
Stringer, C., Boudjelas, S., Broome, K., Cranwell, S., Hagen, E.,
Howald, G., Kelly, J., Millett, J., Springer, K. and Varnham, K. (2019).
‘Married bliss and shotgun weddings: effective partnerships for island
restoration’. In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and
C.J. West (eds.) Island invasives: scaling up to meet the challenge, pp.
517–521. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Sundet, J.H. (2014). ‘The Red King Crab (Paralithodes camschaticus)
in the Barents Sea’. In: L. Fernandez, B. Kaiser and N. Vestergaard
(eds.) Marine Invasive Species in the Arctic, pp. 71–82. Copenhagen,
Denmark: Nordic Council of Ministers.
Sutor, A., Kauhala, K. and Ansorge, H. (2010). ‘Diet of the raccoon
dog Nyctereutes procyonoides: A canid with an opportunistic foraging
strategy’. Acta Theriologica 55: 165–176.
Towns, D.R., Atkinson, I.A.E. and Daugherty, C.H. (2006). ‘Have
the harmful effects of introduced rats on islands been exaggerated?’
Biological Invasions 8: 863–891.
Veitch, C.R., Clout, M.N. and Towns, D.R. (eds.) (2011). ‘Island
invasives: eradication and management’. Gland, Switzerland: IUCN
and Auckland, New Zealand: CBB.
Walsh, A., Wilson, A. and McClelland, P. (2019). ‘Winning the hearts
and minds – proceeding to implementation of the Lord Howe Island
rodent eradication project: a case study’. In: C.R. Veitch, M.N. Clout,
A.R. Martin, J.C. Russell and C.J. West (eds.) Island invasives: scaling
up to meet the challenge, pp. 522–530. Occasional Paper SSC no. 62.
Gland, Switzerland: IUCN.
Walther, G-R., Roques, A., Hulme, P.E., Sykes, M.T., Pysek., P. Kühn,
I. and Zobel, M. (2009). ‘Alien species in a warmer world: risks and
opportunities’. Trends in Ecology and Evolution 24: 686–693.
White House (2016). Executive Order 13751: Safeguarding the Nation
from the Impacts of Invasive Species. Washington, USA: The White
House.
Wilcove, D., Rothstein, D., Dublow, J., Phillips, A. and Losos, E.
(1998). ‘Quantifying threats to imperiled species in the United States’.
BioScience 48: 607–615.
686
P.A. Robertson, S. Roy, A.C. Mill, M. Shirley, T. Adriaens, A.I. Ward, V. Tatayah and O. Booy
Robertson, P.A.; S. Roy, A.C. Mill, M. Shirley, T. Adriaens, A.I. Ward, V. Tatayah and O. Booy. Invasive species removals and scale – contrasting island and mainland experience
Invasive species removals and scale – contrasting island and
mainland experience
P.A. Robertson1, S. Roy2, A.C. Mill1, M. Shirley1, T. Adriaens3, A.I. Ward4, V. Tatayah5 and O. Booy1,6
Centre for Wildlife Management, Newcastle University, Newcastle upon Tyne, NE1 4DD, UK. < peter.robertson@
ncl.ac.uk> 2International Union for the Conservation of Nature, Gland, Switzerland. 3Research Inst. for Nature and
Forest (INBO), Wildlife Management and Invasive Species, Havenlaan 88 bus 73, B-1000 Brussels, Belgium. 4School
of Biological, Biomedical and Environmental Sciences, The University of Hull, HU6 7RX, UK. 5Mauritian Wildlife
Foundation, Grannum Road, Vacoas 73418, Mauritius. 6Animal and Plant Health Agency, Sand Hutton,
York, YO41 1JW, UK.
1
Abstract Recent years have seen large increases in the number and size of successful invasive species eradications from
islands. There is also a long history of large scale removals on larger land-masses. These programmes for mammals and
terrestrial plants follow the same cost-area relationship although spanning 10 orders of magnitude in scale. Eradication
can be readily defined in island situations but can be more complex on larger land-masses where uncertainties defining
the extent of a population, multiple population centres on the same land-mass and ongoing risks of immigration are
commonplace. The term ‘complete removal’ is proposed to describe removal from an area with ongoing effort to maintain
the area as clear, as features in many larger scale mainland programmes. Examples of complete removal to a boundary, in
patches and in habitat islands are discussed. While island eradications continue to grow in scale, new legislation such as
the lists of Species of European Union Concern will also drive increasing management on larger land-masses. However,
these lists include large numbers of species that are already widespread. Methods are needed to prioritise species to reflect
both the risks posed and the feasibility of management, including the effects of scale on cost and effectiveness.
Keywords: control, eradication, invasive alien species, non-native species
INTRODUCTION
The removal or eradication of invasive alien species is
increasingly used as a conservation tool. New legislation,
for example the European Union’s Invasive Alien Species
Regulation, will also place increasing responsibilities on
states to remove or eradicate high risk species. Both of
these considerations are driving an increased number of
management programmes at increasing scales and there
is a need to understand how the costs and constraints
change in relation to scale. A large number of published
eradications have been based on islands, often at relatively
small scales, while a small number of larger programmes
have been based on mainland experience. There is a need to
pull together these different sources of evidence, to support
an assessment across a wider range of scales than can be
achieved by considering islands or mainland eradications
in isolation.
REMOVAL AT SCALE – ISLANDS AND
MAINLAND EXPERIENCE
of Britain and Ireland in the 1930s; the eradication of
the Himalayan porcupine (Hystrix brachyura) (1970s)
and coypu (Myocaster coypus) (1980s) from the British
mainland; a variety of American mink (Neovison vison)
and grey squirrel (Sciurus carolinensis) removals from the
larger British islands together with the removal of Pallas’
squirrel (Callosciurus erythraeus) from Flanders on the
European mainland (since 2000). Few of the programmes
covered more than a fraction of the total land mass, so
size was defined as the area over which species sightings
occurred and trapping took place. The larger of these
species programmes have covered areas of 3,411 km2 (the
two phases of the Hebridean mink programme), 5,219 km2
(the five separate muskrat eradications) and 19,210 km2
(coypu) (details and full references given in Robertson, et
al. 2017). The ongoing ruddy duck (Oxyura jamaicensis)
eradication from Europe (Robertson, et al., 2015) covers
six states totalling 1,535,509 km2.
Recent years have seen a large increase in successful
invasive species eradications from islands, as well as
significant increases in the size of islands involved.
The number of successful eradications continues to
increase, and in 2012 the Database of Invasive Species
Eradications
(<http://diise.islandconservation.org>)
recorded 1,182 whole-island introduced invasive animal
species eradication projects either completed or underway
on 762 individual islands. In terms of scale, recent years
have seen a number of large island eradications. Cruz, et
al. (2009) describe the eradication of goats from the 584
km2 Santiago Island in Galapagos; Parkes, et al. (2014)
predicted the effort required to remove cats from the 1,680
km2 Stewart Island in New Zealand, while the current rat
removal on South Georgia will cover 3,538 km2 (Piertney,
et al., 2016).
Data on the costs of eradications are available for
projects covering ten orders of magnitude of scale.
Studies have described the costs of successful mammal
eradications from islands (Martins, et al., 2006; Howald,
et al., 2007) and larger land-masses (Robertson, et al.,
2017), while Rejmánek & Pitcairn (2002) describe costed
plant eradications in California. For mammal eradications,
those on large land-masses covered significantly larger
areas than those reported from islands while successful
plant eradications were confined to smaller areas. Data
from these different sources, appear to follow the same
relationship (Fig. 1) whereby the cost per unit area is
reduced by approximately 10% as the area involved
doubles (Robertson, et al., 2017). As experience of
eradications on larger islands grows, the overlap between
island and mainland experiences is increasing (Cruz, et al.,
2009; Parkes, et al., 2014; Piertney, et al., 2016).
Although the point at which an island becomes a
mainland is arbitrary, there is also a long history of invasive
mammal removals from larger land masses in Northern
Europe (Robertson, et al., 2017). These include muskrat
(Ondatra zibethicus) eradications from the mainlands
It is worth recording that two small datasets describe
programmes that fall outside this relationship. Rejmánek &
Pitcairn (2002) also record three aquatic plant eradications
which appeared more expensive than comparably sized
terrestrial plant programmes, while the ruddy duck
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
up to meet the challenge, pp. 687–691. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
687
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
eradication (Robertson, et al., 2015) has been significantly
less expensive compared to similarly scaled mammal
programmes (Robertson, et al., 2017). More data are needed
on the management of other taxa in different environments
before firm conclusions can be drawn. These results are
based upon currently available methods of eradication.
As new technologies, such as gene-drives (Webber, et al.,
2015), e-DNA self-resetting (Carter, et al., 2016) and selfreporting traps (Jones, et al., 2015) become available it is
likely that these costs will decrease.
Eradication and complete removal
In their classic paper, Bomford & O’Brien (1995)
make a clear distinction between eradication and ongoing control, presenting these as alternative objectives
for management. They also identify three key criteria for
successful eradication; that the rate of removal exceeds the
rate of increase at all densities; there is no immigration;
and all reproductive animals are at risk.
These definitions and criteria have guided many
successful eradications and are particularly applicable to
islands where the population extent and risks of immigration
can be readily assessed. However, at the scales found on
larger land masses, these criteria may be more difficult
to apply or achieve, for example where the boundaries
of a population remain poorly defined, where multiple
population centres may occur on the same land mass, or
where immigration remains a risk. Despite this, large scale
programmes frequently lead to the removal of species
from large areas of land. Although not meeting Bomford &
O’Brien’s (1995) definition of eradication, these situations
are also not well described as on-going control as no active
management is required across the majority of the area. In
these circumstances ‘complete removal’ may be a better
definition of the objectives, sitting between Bomford &
O’Brien’s (1995) definitions of eradication and on-going
control.
Fig. 1 The relationship between the area (km2) of a
successful removal and the total cost (US$). The square
symbols represent island mammal eradications reported
by Martins, et al., (2006). The circles are for removals of
mammals from larger land masses in Northern Europe
(Robertson, et al., 2016). The three diamond symbols
are recent examples or predictions of large-scale
mammal eradications from islands: (Cruz, et al., 2009;
Parkes, et al., 2014; Piertney, 2016). Plant eradications
from California are triangles (Rejmánek & Pitcairn, 2002).
Where the study recorded effort as man-years or mandays, total cost is estimated based on US$50k per manyear (Rejmánek & Pitcairn, 2002; Parkes, et al., 2014;
Robertson, et al., 2016).
688
Eradication, the complete removal from an area, with
no immediate prospect of recolonisation from neighbouring
areas.
Complete Removal from an area but with ongoing
effort to maintain the area as clear.
On-going Control within an area to reduce abundance,
associated damage and the risk of spread.
Based on this definition, complete removal has been
applied in a number of forms.
1 - Complete removal to a boundary
One objective of large scale programmes can include
complete removal of a species up to a boundary across
which the risk of reinvasion remains. Control along the
boundary, or in a neighbouring buffer zone, can reduce
the risk of reinvasion and help keep the main area clear.
The nature of the boundary may vary, including fences
(Saunders & Norton, 2001), landscape barriers such as
water bodies or mountains (Schuchert, et al., 2014), or
bottlenecks through which invading animals must move
(Roy, et al., 2015). These boundaries can be permanent
features of the management, requiring ongoing inputs
(Saunders & Norton, 2001), or may be part of a phased
programme to clear a larger area (Yamada & Sugimura,
2004; Bryce, et al., 2011; Robertson, et al., 2015; Russell,
et al., 2015). If the aim is the removal of the species from a
large area, but the funds or resources are insufficient for the
simultaneous management of the entire population, then
removal to a boundary is likely to feature.
The North American ruddy duck was introduced
to the UK in the late 1940s, and its subsequent spread
into Europe threatens the native white-headed duck
(Oxyura leucocephalus) through hybridisation. The
plan to eradicate the ruddy duck from Europe involves
coordinated management across the continent. As the UK
was the original source of this population and contained
the majority of the birds, it was the focus of initial control
(Robertson, et al., 2015). However, once the UK no longer
contained breeding birds (currently it is thought only a few
males remain), the English Channel became a boundary
between a cleared area and the remaining continental
populations. Control of the remaining European birds is
ongoing, in the meantime the UK maintains surveillance
and, if required, control along this boundary to maintain
its cleared status.
In the UK, the native red squirrel (Sciurus vulgaris)
is threatened by the ongoing spread of the invasive grey
squirrel (S. carolinensis). This is mediated by the spread
of a poxvirus by the asymptomatic greys which is typically
fatal to the reds (Rushton, et al., 2000). The island of
Anglesey on the north coast of Wales contained a small
relict population of the native reds although greys were
spreading onto the island. A control programme removed the
greys (Schuchert, et al., 2014), allowing the reds to spread
and recolonise the entire island. Anglesey is separated
from mainland Wales by a narrow tidal channel, crossed
by two bridges. There is evidence that grey squirrels can
cross this boundary and the risk of recolonisation remains.
To reduce this risk and maintain the island as grey squirrelfree, management has included a surveillance and rapid
response programme to pick up incursions (Shuttleworth,
et al., 2016), trapping to reduce the density of greys on the
mainland side of the boundary, and a plan to extend the
area of complete removal to clear greys from the North
Wales coast up to a more distant boundary formed by a
geographic bottleneck where the mountains meet the coast.
The American mink (Neovison vison) spread through
the Western Isles of Scotland following its escape from fur
Robertson, et al.: Contrasting island and mainland experience
farms in the 1950s. Its spread threatened internationally
important populations of ground nesting birds as well
as local economic activities such as salmon fishing. The
decision was taken to aim for the eradication of this species
from the archipelago but logistic and funding constraints,
combined with the need to gain experience, led to a phased
programme. In the first phase, mink were completely
removed from the Uists, the southernmost islands of
the chain (Roy, et al., 2015; Faulkner, et al., 2017). A
buffer zone was maintained (South Harris) between this
cleared area and the remaining mink population on the
main island (Lewis) to the north. This buffer included a
narrow, island strewn channel between the Uists and South
Harris. Trapping on these ‘stepping stone’ islands together
with South Harris itself provided an effective barrier
to recolonisation. Once the Uists’ work had provided
confidence that eradication was feasible, a second phase
extended mink control north to cover the remainder of the
archipelago (Lambin, et al., 2014).
2 - Complete removal from patches
In some cases the primary objective of management
may be the reduction of the impact of an invasive species
with no prospect to eradicate. In many cases this constitutes
ongoing control rather than complete removal (Bomford &
O’Brien, 1995), although in some circumstances it can lead
to complete removal. For this to occur, two criteria must be
met, the species must be controlled at a rate sufficient to
remove all of the resident animals in an area, and the scale
of control should be such that the risk of recolonisation is
so low in the centre of the controlled area that the central
area is effectively maintained clear. The prospects of
this occurring are scale-dependent, with the cleared area
forming a larger proportion of the total as scale increases.
This approach has been used in New Zealand with the
creation of ‘mainland islands’, areas maintained predatorfree through the use of fencing combined with continuing
control (Saunders & Norton 2001; Gillies, et al., 2003).
The same results can be achieved without fencing, for
example in Mauritius where the introduced small Asian
mongoose (Urva auropunctata) (Patou, et al., 2009)
is a major threat to the continued survival of a range of
native bird species (Bunbury, et al., 2008). The mongoose
is widely spread across the island, inhabiting a range of
habitats, while the native birds are largely confined to
remaining patches of good quality native forest. Control
of the mongoose has been carried out in a number of these
forest areas to create ‘mongoose free’ patches within the
wider mongoose distribution. A network of box traps
has been in place since 1989 and maintains a year-round
effort to remove mongoose. As the size of the trapped
area increases, the number of animals captured per unit
area decreases (Fig. 2). Areas less than 5 km in extent
continue to catch high numbers of mongoose per unit area,
presumably because they face constant recolonisation
pressure from neighbouring habitats. However, in larger
areas, particularly those over 10 km2 in area, mongoose
catch per unit area drops dramatically. This is consistent
with catching animals in a boundary area, with the
proportion of the area maintained as mongoose-free
increasing as the total area trapped increases. Achieving
this requires ongoing effort, but complete removal provides
many of the benefits of eradication, and has been a key
element of efforts to conserve a suite of species endemic
to the island. These include the Mauritius kestrel (Falco
punctatus), the pink pigeon (Nesoenas mayeri), the echo
parakeet (Psittacula eques) and a number of passerines
such as the Mauritius black bulbul (Hypsipetes olivaceus),
and Mauritius fody (Foudia rubra). Only through intensive
trapping to maintain these predator-free patches, combined
with a captive breeding and release programme, disease
management and supplementary feeding, have these
species managed to persist.
3 – Complete removal from habitat islands
Islands as blocks of land surrounded by water are
widely recognised, but isolated blocks of habitat within
a matrix of other land uses share many of the same
characteristics. When invasive alien species are confined to
discrete habitats within this matrix, they can be considered
as inhabiting ‘habitat islands’. In these cases, limited rates
of species movement or colonisation between habitat
islands may produce isolated populations, with particular
opportunities for management within large land masses.
The monk parakeet (Myiopsitta monachus) has
established a number of discrete populations in different
European cities (Munoz & Real, 2006; Rodríguez-Pastor,
et al., 2012). Although an attractive species widely kept as
a pet, in the wild this species builds large communal nests
on tall trees or man-made structures such as electricity
infrastructure or radio masts. The large size and volume
of nest material can lead to electrical short-outs and fire
risks, with consequent economic costs (Avery, et al., 2002).
The discrete nature of its current distribution, with isolated
populations including London, Amsterdam and a variety
of Spanish cities suggests that different populations have
resulted from separate releases rather than natural spread
from a single point of release. The management of this
species reflects this, with some regions attempting the
complete removal of isolated populations (Parrott, 2013).
The introduced Pallas’ squirrel (Callosciurus
erythraeus) also has a highly fragmented distribution within
Europe, suggesting a number of separate introductions
rather than spread from a single point of release. A rapid
response in Flanders, Belgium removed a population
whose distribution was constrained to a suburban setting
in a small community surrounded by farmland (Adriaens,
et al., 2015). In effect this species was present on a habitat
island which aided its removal.
The current removal of rats from South Georgia
(Piertney, et al., 2016) uses a similar approach. Glaciers
on the island separate a number of discrete rat populations,
which appear to be genetically isolated (Robertson &
Gemmell, 2004). This allows the complete removal of
discrete populations as steps to achieve the larger goal of
island wide eradication.
These examples illustrate the potential for effective
removal of isolated populations to be undertaken within
larger land masses, using the principles applied to island
Fig. 2 The density of mongoose removed by trapping in
five conservation areas in Mauritius. The control areas
were surrounded by habitat containing mongoose
populations.
689
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
eradications. However, as species establish and spread
these discrete populations will become less pronounced.
Identifying whether the distribution of a species represents
a number of discrete clusters will have important
implications for management, for example the decision
to consider complete removal or on-going control. Spatial
analysis of distributional data can be used to indicate the
presence of discrete populations of a species. A range
of spatial and spatio-temporal clustering algorithms
(Velázquez, et al., 2016) can detect spatial point patterns
and may be useful to differentiate clusters as they form.
EFFECTIVENESS AND SCALE
We used published accounts to assess the costs of
removal at different scales. Doing so requires dealing with
a number of biases. Firstly, it is commonly recognised
that the published literature preferentially records success
(Dwan, et al., 2008). For example, the successful coypu
eradication in the UK is well documented (Gosling &
Baker, 1987; Baker & Clarke, 1988; Gosling, et al.,
1988; Gosling & Baker, 1989; Baker, 2006); the failed
UK attempt to eradicate the American mink is barely
recorded (Sheail, 2004) although it took place on a similar
scale. Other failures are likely to have gone unrecorded.
A publicly available database of island eradications is
available (Keitt, et al., 2011), it would be useful to extend
this to also include details of eradications on larger land
masses. More importantly, the literature only records
attempts, there is very little information on those situations
where no action was taken, either through inaction or a
judgement that it was not worthwhile. Inaction remains the
most common response to invasive species. The successful
island eradications are based on only a tiny proportion
of the world islands, while the number of attempted
eradications of alien species in Europe (Genovesi, 2005) is
a similarly small proportion of the 20,000 species thought
to have established.
If we are to make more objective decisions, we need
to decide if, and when, management is appropriate in both
island and mainland situations. Prioritisation methods
have been applied to islands to identify those where
management may be most beneficial (Harris, et al., 2012;
Dawson, et al., 2015). Booy, et al. 2017 describe a method
to assess the feasibility of eradication which incorporates
the consideration of scale. If, as seems likely but has yet to
be convincingly demonstrated, the prospects for successful
eradication or complete removal decrease as a species
spreads, then these methods offer a route to assess at what
scale eradication or complete removal may no longer be a
realistic outcome.
The application of methods to assess the feasibility of
management is a critical need. The current EU invasive
alien species regulations include the listing of species
considered to be of ‘Union Concern’ and place reporting
and management obligations on member states in which
they occur. The selection of species for listing is largely
based on established methods of risk assessment (Roy, et
al., 2014), identifying species which pose a risk without
similarly considering the feasibility of management.
This focus on risk can result in the listing of species for
which there are few realistic prospects for management.
For example, of the 79 species currently listed or under
consideration as Species of Union Concern, over half are
already present in at least five member states. To date
there are no successful examples of species eradication or
complete removal in Europe when a species has already
spread to this number of countries, although these may
occur in future. Listing species based on risk assessment
alone, without considering the scale and feasibility of
management, risks committing resources into the on-going
690
management of already widespread species, rather than the
more productive routes of prevention and rapid response.
CONCLUSIONS
The experience of island eradications continues to
grow, and to be applied at increasing scales. Alongside this,
new legislation will drive increasing management on larger
land masses. As island eradications grow in scale they will
face many of the challenges experienced on larger landmasses, such as problems defining populations, multiple
population centres on the same land mass, ongoing risks
of immigration and the need for interim objectives. We
suggest the term ‘complete removal’ to reflect the situation
regularly encountered on larger land masses where a
species may be removed from an area but with the need for
an ongoing effort to maintain the area clear given the risk of
reinvasion. The literature contains examples of successful
eradications or complete removals in island and mainland
situations covering 10 orders of magnitude. These island
and mainland programmes appear to follow the same costarea relationship. They also demonstrate an advantage of
scale, with the costs per unit area of control reduced as the
area of control increases. On larger land masses, such as
the EU, care is needed to focus species listing on species
where prevention, eradication or complete removal are
realistic outcomes rather than committing member states
to the on-going control of already widespread species.
Methods of prioritisation which balance both risk and the
feasibility of management, including the effects of scale on
cost and effectiveness, are needed to guide future actions.
REFERENCES
Adriaens, T., Baert, K., Breyne, P., Casaer, J., Devisscher, S., Onkelinx, T.,
Pieters, S. and Stuyck, J. (2015). ‘Successful eradication of a suburban
Pallas’s squirrel Callosciurus erythraeus (Pallas 1779) (Rodentia,
Sciuridae) population in Flanders (northern Belgium)’. Biological
Invasions 17(9): 2517–2526.
Avery, M.L., Greiner, E.C., Lindsay, J.R., Newman, J.R. and Pruett-Jones,
S. (2002). ‘Monk parakeet management at electric utility facilities in
south Florida’. Proceedings of the 20th Vertebrate Pest Conference, pp.
140–145. University of California, Davis.
Baker, S.J. and Clarke, C.N. (1988). ‘Cage trapping coypus (Myocastor
coypus) on baited rafts’. Journal of Applied Ecology 25: 41–48.
Baker, S. (2006). ‘The eradication of coypus (Myocastor coypus) from
Britain: the elements required for a successful campaign’. Assessment
and Control of Biological Invasion Risks, pp.142–147. Kyoto, Japan:
Shoukadoh Book Sellers andOccasional Paper SSC no. 62. Gland,
Switzerland: IUCN.
Bomford, M. and O'Brien, P. (1995). ‘Eradication or control for vertebrate
pests?’ Wildlife Society Bulletin 23: 249–255.
Booy, O., Mill, A.C., Roy, H.E., Hiley, A., Moore, N., Robertson, P.,
Baker, S., Brazier, M., Bue, M., Bullock, R. and Campbell, S. (2017)
‘Risk management to prioritise the eradication of new and emerging
invasive non-native species’. Biological Invasions, 19(8): 2401–2417.
Bunbury, N., Stidworthy, M.F., Greenwood, A.G., Jones, C.G., Sawmy,
S., Cole, R.E., Edmunds, K. and Bell, D.J., (2008). ‘Causes of
mortality in free-living Mauritian pink pigeons Columba mayeri, 2002–
2006’. Endangered Species Research 9(3): 213–220.
Bryce, R., Oliver, M.K., Davies, L., Gray, H., Urquhart, J. and Lambin,
X., (2011). ‘Turning back the tide of American mink invasion at an
unprecedented scale through community participation and adaptive
management’. Biological Conservation 144(1): 575–583.
Carter, A., Barr, S., Bond, C., Paske, G., Peters, D. and van Dam, R.
(2016). ‘Controlling sympatric pest mammal populations in New
Zealand with self-resetting, toxicant-free traps: a promising tool for
invasive species management’. Biological Invasions 18(6): 1723–1736.
Cruz, F., Carrion, V., Campbell, K.J., Lavoie, C. and Donlan, C.J. (2009).
‘Bio-economics of large-scale eradication of feral goats from Santiago
Island, Galápagos’. Journal of Wildlife Management 73(2): 191–200.
Dawson, J., Oppel, S., Cuthbert, R.J., Holmes, N., Bird, J.P., Butchart,
S.H., Spatz, D.R. and Tershy, B. (2015). ‘Prioritizing islands for the
eradication of invasive vertebrates in the United Kingdom overseas
territories’. Conservation Biology 29(1): 143–153.
Robertson, et al.: Contrasting island and mainland experience
Dwan, K., Altman, D.G., Arnaiz, J.A., Bloom, J., Chan, A.W., Cronin, E.,
Decullier, E., Easterbrook, P.J., Von Elm, E., Gamble, C. and Ghersi,
D. (2008). ‘Systematic review of the empirical evidence of study
publication bias and outcome reporting bias’. PLOS One, 3(8): e3081.
Robertson, P.A., Adriaens, T., Lambin, X., Mill, A., Roy, S., Shuttleworth,
C.M. and Sutton‐Croft, M. (2017). ‘The large‐scale removal of
mammalian invasive alien species in Northern Europe’. Pest
Management Science 73(2): 273–279.
Faulkner, S.C., Verity, R., Roberts, D., Roy, S.S., Robertson, P.A.,
Stevenson, M.D. and Le Comber, S.C. (2017). ‘Using geographic
profiling to compare the value of sightings vs trap data in a biological
invasion.’ Diversity and Distributions 23(1): 104–112.
Rodríguez-Pastor, R., Senar, J.C., Ortega, A., Faus, J., Uribe, F. and
Montalvo, T. (2012). ‘Distribution patterns of invasive Monk parakeets
(Myiopsitta monachus) in an urban habitat.’ Animal Biodiversity and
Conservation 35(1): 107–117.
Genovesi, P., (2005). ‘Eradications of invasive alien species in Europe: a
review’. Issues in Bioinvasion Science 7: 127–133.
Roy, H., Schonrogge, K., Dean, H., Peyton, J., Branquart, E.,
Vanderhoeven, S., Copp, G., Stebbing, P., Kenis, M., Rabitsch, W. and
Essl, F. (2014). Invasive Alien Species–Framework for the Identification
of Invasive Alien Species of EU Concern. Report to the European
Commission ENV.B.2/ETU/2013/0026.
Gillies, C.A., Leach, M.R., Coad, N.B., Theobald, S.W., Campbell, J.,
Herbert, T., Graham, P.J. and Pierce, R.J. (2003). ‘Six years of intensive
pest mammal control at Trounson Kauri Park, a Department of
Conservation “mainland island”, June 1996–July 2002’. New Zealand
Journal of Zoology 30(4): 399–420.
Gosling, L.M. and Baker, S.J. (1987). ‘Planning and monitoring an
attempt to eradicate coypus from Britain’. Symposia of the Zoological
Society of London 58: 99–113.
Gosling, L.M., Baker, S.J. and Clarke, C.N. (1988). ‘An attempt to
remove coypus (Myocastor coypus) from a wetland habitat in East
Anglia’. Journal of Applied Ecology 25: 49–62.
Gosling, L.M. and Baker, S.J. (1989). ‘The eradication of muskrats and
coypus from Britain.’ Biological Journal of the Linnean Society 38(1):
39–51.
Harris, D.B., Gregory, S.D., Bull, L.S. and Courchamp, F. (2012). ‘Island
prioritization for invasive rodent eradications with an emphasis on
reinvasion risk’. Biological Invasions 14(6): 1251–1263.
Howald, G., Donlan, C., Galván, J.P., Russell, J.C., Parkes, J., Samaniego,
A., Wang, Y., Veitch, D., Genovesi, P., Pascal, M. and Saunders,
A. (2007). ‘Invasive rodent eradication on islands’. Conservation
Biology 21(5): 1258–1268.
Jones, C., Warburton, B., Carver, J. and Carver, D. (2015). ‘Potential
applications of wireless sensor networks for wildlife trapping and
monitoring programs’. Wildlife Society Bulletin 39(2): 341–348.
Keitt, B., Campbell, K., Saunders, A., Clout, M., Wang, Y., Heinz, R.,
Newton, K. and Tershy, B., 2011. The global islands invasive vertebrate
eradication database: a tool to improve and facilitate restoration of
island ecosystems. In: C.R. Veitch, M.N. Clout and D.R. Towns (eds.)
Island invasives: eradication and management, pp. 74–77. Occassional
Paper SSC no. 42. Gland, Switzerland: IUCN and Auckland, New
Zealand: CBB.
Lambin, X.L., Cornulier, T., Oliver, M.K. and Fraser, E.J. (2014) Analysis
and Future Application of Hebridean Mink Project Data. Scottish
Natural Heritage Commissioned Report No. 522.
Martins, T.L.F., Brooke, M.D.L., Hilton, G.M., Farnsworth, S., Gould,
J. and Pain, D.J. (2006). ‘Costing eradications of alien mammals from
islands’. Animal Conservation 9(4): 439–444.
Muñoz, A.R. and Real, R. (2006). ‘Assessing the potential range
expansion of the exotic monk parakeet in Spain’. Diversity and
Distributions 12(6): 656–665.
Roy, S.S., Chauvenet, A.L.M. and Robertson, P.A., (2015). ‘Removal
of American mink (Neovison vison) from the Uists, Outer Hebrides,
Scotland’. Biological Invasions 17(10): 2811–2820.
Rushton, S.P., Lurz, P.W.W., Gurnell, J. and Fuller, R. (2000). ‘Modelling
the spatial dynamics of parapoxvirus disease in red and grey squirrels:
a possible cause of the decline in the red squirrel in the UK?’ Journal of
Applied Ecology 37(6): 997–1012.
Russell, J.C., Innes, J.G., Brown, P.H. and Byrom, A.E. (2015). ‘Predatorfree New Zealand: conservation country’. BioScience 65(5): 520–525.
Saunders, A. and Norton, D.A. (2001). ‘Ecological restoration at mainland
islands in New Zealand’. Biological Conservation 99(1): 109–119.
Schuchert, P., Shuttleworth, C.M., McInnes, C.J., Everest, D.J. and
Rushton, S.P. (2014). ‘Landscape scale impacts of culling upon a
European grey squirrel population: can trapping reduce population size
and decrease the threat of squirrelpox virus infection for the native red
squirrel?’ Biological Invasions 16(11): 2381–2391.
Shuttleworth, C, Robertson, P. and Halliwell, L. (2016). ‘Identifying
incursion pathways, early detection responses and management actions
to prevent grey squirrel range expansion: an island case study in Wales.’
In: C.M. Shuttleworth, P.W.W. Lurz and J. Gurnell (eds.) The Grey
Squirrel: Ecology and Management of an Invasive Species in Europe,
pp. 475–495. UK: European Squirrel Initiative.
Sheail, J. (2004). ‘The mink menace: the politics of vertebrate pest
control’. Rural History 15(02): 207–222.
Velázquez, E., Martínez, I., Getzin, S., Moloney, K.A. and Wiegand, T.
(2016). ‘An evaluation of the state of spatial point pattern analysis in
ecology’. Ecography 39(11): 1042–1055.
Webber, B.L., Raghu, S. and Edwards, O.R. (2015). ‘Opinion: Is CRISPRbased gene drive a biocontrol silver bullet or global conservation
threat?’ Proceedings of the National Academy of Sciences 112(34):
10565–10567.
Yamada, F. and Sugimura, K. (2004). ‘Negative impact of an invasive
small Indian mongoose Herpestes javanicus on native wildlife species
and evaluation of a control project in Amami-Ohshima and Okinawa
Islands, Japan’. Global Environmental Research 8(2): 117–124.
Parkes, J., Fisher, P., Robinson, S. and Aguirre-Muñoz, A., (2014).
‘Eradication of feral cats from large islands: an assessment of the effort
required for success’. New Zealand Journal of Ecology 38: 307–314.
Parrott, D. (2013) ‘Monk parakeet control in London’. In: C. van Ham,
P. Genovesi and R. Scalera (eds.) Invasive alien species: the urban
dimension, pp 83–85. Gland, Switzerland: IUCN.
Patou, M. L., Mclenachan, P. A., Morley, C. G., Couloux, A., Jennings,
A. P. and Veron, G. (2009). ‘Molecular phylogeny of the Herpestidae
(Mammalia, Carnivora) with a special emphasis on the Asian
Herpestes’. Molecular Phylogenetics and Evolution 53(1): 69–80.
Piertney, S.B., Black, A., Watt, L., Christie, D., Poncet, S. and
Collins, M.A. (2016). ‘Resolving patterns of population genetic and
phylogeographic structure to inform control and eradication initiatives
for brown rats Rattus norvegicus on South Georgia’. Journal of Applied
Ecology 53: 332–339.
Rejmánek, M. and Pitcairn, M.J. (2002). ‘When is eradication of exotic
pest plants a realistic goal.’ In: C.R. Veitch and M.N. Clout (eds.) Turning
the tide: the eradication of invasive species, pp. 249–253. IUCN, SSC
Invasive Species Specialist Group. IUCN, Gland, Switzerland and
Cambridge, UK.
Robertson, B.C. and Gemmell, N.J. (2004). ‘Defining eradication units to
control invasive pests’. Journal of Applied Ecology 41(6): 1042–1048.
Robertson, P.A., Adriaens, T., Caizergues, A., Cranswick, P.A., Devos,
K., Gutiérrez-Expósito, C., Henderson, I., Hughes, B., Mill, A.C. and
Smith, G.C. (2015). ‘Towards the European eradication of the North
American ruddy duck.’ Biological Invasions 17(1): 9–12.
691
J.C. Russell and C.N. Taylor
Russell, J.C. and C.N. Taylor. Strategic environmental assessment for invasive species management on inhabited islands
Strategic environmental assessment for invasive species management
on inhabited islands
J.C. Russell1 and C.N. Taylor2
School of Biological Sciences and Department of Statistics, University of Auckland, Private Bag 92019, Auckland,
New Zealand. <j.russell@auckland.ac.nz>. 2Taylor Baines & Associates, PO Box 8620, Riccarton, Christchurch,
New Zealand.
1
Abstract Over the past decade the challenges of managing invasive species on inhabited islands have clearly become
limiting factors to scaling-up the area of invasive species eradications. Step-change is required to unleash the conservation
and restoration potential of biodiversity on inhabited islands around the globe and avoid the pitfalls previous attempts to
eradicate invasive species on inhabited islands have fallen into. Strategic environmental assessment (SEA) is a systematic
decision support process, aiming to ensure that environmental and other sustainability aspects are considered effectively
throughout policy, plan and programme making. Within the framework of SEAs, on target islands eradication planners
could implement a number of tools including stakeholder engagement, social impact assessment and economic costbenefit analysis alongside existing environmental impact assessment. Such a suite of assessments captures the immediate
impacts of an eradication operation on a range of values, alongside predicted long-term changes in these tightly coupled
socio-ecological systems. In this paper we outline what SEA is, and then contrast invasive species management attempts
occurring outside an SEA framework on two similar but also contrasting UNESCO World Heritage islands; Lord Howe
Island, Australia and Fernando de Noronha, Brazil. We then demonstrate how an SEA approach to invasive species
management would assist planning in New Zealand to eradicate introduced mammalian predators from two large offshore
islands in New Zealand; Aotea (Great Barrier Island) and Rakiura (Stewart Island). We conclude with future prospects for
applying SEA to invasive species management on inhabited islands.
Keywords: eradication, mammals, New Zealand, social impact assessment, World Heritage
INTRODUCTION
Over the past decade the challenges of managing
invasive species on inhabited islands have clearly become
limiting factors to scaling-up the area of invasive species
eradications on islands (Oppel, et al., 2011; Glen, et al.,
2013). This is particularly the case for eradication of small
mammalian predators, where step-change in technology
(e.g. use of helicopters for aerial delivery of toxin; Howald,
et al., 2007) coupled with ongoing incremental advances
(e.g. non-target mitigation; Hanson, et al., 2015) mean
very large islands are now potential targets of wholeisland small mammal eradication, but there has not been a
commensurate increase in the knowledge around engaging
with resident communities (Russell, et al., 2018). Scalingup eradications to larger islands is also confounded by
additional complexities such as mixed land-tenure and
land-use (Holmes, et al., 2015) on larger islands, further
complicating the suite of appropriate methods for social
engagement and technical implementation.
There are many reasons why there should be an
increased emphasis on inhabited islands as targets for
biodiversity conservation. Most uninhabited islands are
small and, although the number of islands from which
invasive species have been eradicated is impressive (e.g.
Jones, et al., 2016), as a percentage island land area the
total is still low (Russell, et al., 2016a). Some endangered
species can only be conserved on large islands (PCE,
2017), while climate change increases the long-term risk
profile for small islands as resilient conservation sites
(Courchamp, et al., 2014). In the tropics, even small islands
can be inhabited (Russell & Holmes, 2015), and small
island developing states (SIDS) are particularly poorly
represented in invasive mammal eradication statistics
(Russell, et al., 2017a). Eradication of invasive mammals
on inhabited islands also brings about many other benefits
beyond biodiversity conservation, including benefits to
agriculture, economies, public health and culture (Russell,
et al., 2017a).
To date approaches to community engagement in
anticipation of mammal eradication on inhabited islands
have been designed and led mainly by biologists with a
particular set of values and priorities (e.g. Bell, 2019). They
have tended to be ad hoc and have not always drawn upon
existing scholarship in community engagement. A new
step-change is required to unleash the conservation and
restoration potential of biodiversity on inhabited islands
around the globe and avoid the pitfalls previous attempts to
eradicate invasive species on inhabited islands have fallen
in to. In this paper we outline the potential for strategic
environmental assessment to enable more consistent
assessment of options and engagement with island
communities in the context of invasive mammal eradication.
We then provide two contrasting illustrative examples
of approaches to invasive predator management on two
similar UNESCO World Heritage island sites, followed by
examples from the two largest inhabited offshore islands
of New Zealand. We conclude with recommendations
for implementing strategic environmental assessment
during planning for invasive species management. We
emphasise that much of the scholarship we present here
is built upon reflection over the past decade on attempts to
eradicate invasive mammals from inhabited islands. These
lessons come from the benefit of hindsight and could not
be anticipated in advance, so they should not be taken as
reflecting poorly on those who initially invested themselves
in advocating for invasive mammal eradication. Our
purpose is to suggest a way towards better processes and
improved outcomes from eradications on inhabited islands.
Strategic Environmental Assessment
Strategic environmental assessment (SEA) is a
widely accepted approach to applying impact assessment
to policies, plans and programmes, contributing to the
planning processes, decision making and the ongoing
management of change (Tetlow & Hanusch, 2012).
Sustainability assessment is another approach often
linked to SEA (Morgan 2012). SEA has been described
as “analytical and participatory approaches that aim to
integrate environmental considerations into policies,
In:
C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
692
up to meet the challenge, pp. 692–697. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Russell & Taylor: Environmental assessment for inhabited islands
plans and programmes and evaluate the inter linkages
with economic and social considerations” (OECD, 2006).
Applications of SEA include spatial planning, sector
planning (e.g. fisheries, energy) and catchment planning
(Tetlow & Hanusch, 2012; Taylor & Mackay, 2016).
Importantly, SEA provides an over-arching framework
of a collection of tools rather than a single, fixed and
prescriptive approach. Such an approach is therefore
analogous to best practice in technical implementation
of eradications on islands (Keitt, et al., 2015), where just
as islands differ ecologically, it is also recognised they
differ socially. Thus, in any particular case variations with
regards to best, or complete, practice will still take place.
Its application is an ongoing adaptive and iterative process
which adds value to and builds capacity in existing systems
(e.g. island human communities). The sorts of tools that
can be considered as contributing to the SEA toolbox for
island eradications include:
● Community and stakeholder engagement techniques
● Social profiles/baselines
assessments (SIA)
and
social
impact
● Health impact assessments
● Cost benefit analyses
● Ecological baselines and impact assessments (EIA)
● Technical feasibility studies
● Livelihoods analyses
● Social marketing/environmental education
● Environmental and social monitoring
● Institutional analyses and change management
(includes ongoing biosecurity planning).
As a toolbox, SEA has been around since about the early
1990s when it developed from a growing realisation that
local and project-specific applications of environmental
impact assessment are insufficient when environmentally
damaging decisions are being made at a more strategic level.
SEA has not been widely applied in the context of wildlife
management (Taylor, et al., 2004). However, in some
countries SEA-like frameworks have been implemented
in all but name (e.g. the Resource Management Act in
New Zealand provides for the application of SEA and
the development of policies and plans for the purposes of
natural resource management). Strategic environmental
assessment is widely accepted internationally as a critical
tool in development planning (e.g. by the World Bank and
OECD), where the focus is on impact analysis through
to institutional assessment. Strategic environmental
assessment is accepted in international development
as a way to incorporate environmental considerations
across all levels of strategic decision-making including
plans, programmes, and policies, setting the context
for environmental and social impacts assessments of
development projects.
In the context of wildlife management on inhabited
islands, we adopt the definition of Russell, et al. (2018)
for an inhabited island. Namely that “inhabitation on an
island incorporates the basic infrastructure to enable a
community to function socially and economically, such as
any of schools, churches, community buildings or general
shared spaces, alongside enterprises delivering goods
and services, and opportunities for residents to pursue a
range of livelihood opportunities in the public and private
sectors”. However, we hasten to add that even when an
island is uninhabited, a social framework process may still
be required during wildlife management planning where
stakeholders and others with vested interests in the island
can be identified.
Poor or inconsistent planning is well known in other
sectors to delay project completion (Flyvbjerg, 2014). To
avoid this problem, we consider wildlife management on
islands, and particularly eradication of invasive species,
should be treated in the same way as any large-scale,
multi-component development project, whereby SEA
is a valuable unifying framework that draws together a
collection of tools. Many of the tools under SEA are already
becoming increasingly applied when planning invasive
species management, such as social profiling (Russell,
et al., 2018), social impact assessment (Crowley, et al.,
2017b), and participatory processes (McEntee & Johnson,
2016). Other tools, such as EIA and economic cost-benefit
analyses, can work under the umbrella of SEA for specific
eradication projects, once the strategic framework is in
place. In particular, eradication practitioners globally
should adopt a best practice approach when working with
communities on inhabited islands, as they already do for
technical best practice when planning the operational
elements of eradications on islands (Keitt, et al., 2015).
Most importantly, SEA provides the policy tool
by which the role of invasive species eradication as a
conservation intervention can cascade throughout all
levels of the decision-making process on islands, including
deliberative and more participatory approaches (Sims,
2012). This more comprehensive approach applies not
just to decisions about wildlife management, but around
sustainability of the environment and the livelihoods of
human communities on islands. We see this as critical to
avoid the pitfalls that previous eradication propositions
on inhabited islands haven fallen into – namely where
invasive species eradication is considered only as a
technical solution to a wildlife management problem on a
project by project basis (isolated from other island issues
and strategies), and where the support for eradication is
seen as merely needing to gain a public consensus through
democratic process.
UNESCO WORLD HERITAGE ISLANDS
Many island groups are listed as UNESCO World
Heritage sites based on their cultural and natural heritage
values. A subset of the islands listed for natural heritage
values are also inhabited. In this section we explore the
contrasting experiences of enabling introduced small
mammal predator management on two similar inhabited
UNESCO World Heritage islands where such predator
management has been proposed; Fernando de Noronha,
Brazil and Lord Howe Island, Australia. These are not the
only UNESCO World Heritage islands where predator
management takes place. Predator management is also
undertaken on Fraser Island, Australia but within the
context of a suite of different social and environmental
issues related to dingo management (Allen, et al., 2018),
and has also been considered on Gough Island (Varnham,
et al., 2011), and undertaken on islands in the Galapagos
(Carrion, et al., 2011) and Ogasawara Islands (Hashimoto,
2010).
Fernando de Noronha
Fernando de Noronha and Atol das Rocas Reserves in
Brazil was assigned UNESCO World Heritage status in
2001. Fernando de Noronha is an archipelago, comprising
the primary island of the same name and 20 secondary
islands and islets, lying 345 km north east of Brazil in the
tropical Atlantic Ocean. The inhabited centre of the island
is classified as an Environmental Protection Area (APA),
while the uninhabited forested outer areas of the island are
part of the Marine National Park (PARNAMAR). Both
areas are environmentally administered at the federal level
by ICMBio, but socio-politically administered at the state
693
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
level by neighbouring Pernambuco state on the continent.
The resident population of Fernando de Noronha is
estimated at around 3,000 people (IBGE 2016). Tourism is
the major enterprise on Fernando de Noronha (de Oliveira,
2003), and an estimated 500 tourists arrive and depart
each day. This has led the state government to impose a
daily tourist tax for environmental protection <http://
www.ilhadenoronha.com.br/ailha/taxadepreservacao_em_
noronha.php>. However, it is only the regulation of visitor
numbers and not proceeds of the tax which contribute
directly to environmental protection.
Today the major invasive species on Fernando de
Noronha are cats (Felis catus), black (Rattus rattus)
and brown rats (R. norvegicus), and the introduced tegu
(Salvator merianae) lizard (Abrahão, et al., 2019). In
Brazil, invasive species are not widely acknowledged as
a threat to biodiversity (Bellard & Jeschke, 2016), and any
management of invasive species on Fernando de Noronha
typically reflects a public health and continental mind
set. Wildlife is managed only in the context of vectors of
disease (Magalhães, et al., 2017) while cats are managed
as companion animals with strict laws administered from
the governing Pernambuco state which do not permit lethal
control of cats unless their own welfare is suffering (Dias,
et al., 2017). The tegu is a CITES listed native species
from continental South America, which is also likely to
be having severe predatory impacts on the island fauna
(Abrahão, et al., 2019).
Management of invasive species on Fernando de
Noronha lacks an island conservation context which
acknowledges the severe impact such species are having on
the biodiversity of the island, and does not engage in lethal
control (Russell, et al., 2016b). These biodiversity impacts
are not able to be considered alongside other social and
economic issues on Fernando de Noronha, as independent
levels and agencies of government are in charge of each
separately. Strategic environmental assessment would
allow proposals for the management of invasive species
on Fernando de Noronha to be placed within their
broader social context, where invasive species can be
considered both as public health pests and companion
animals. Impact assessments of invasive species on both
the environment and society are absent but could be
contemporaneously created. The island’s environmental
aesthetic (e.g. beaches) is known to be the main driver
of tourism, and generates considerable wealth each year,
but it is unknown what role the island’s biota (e.g. unique
endemic species) play in tourism. Strategic environmental
assessment would allow the costs of invasive species on
the wider economy to be properly calculated, alongside
the potential added value to tourism from invasive species
management if not eradication. It would play a role in
assessing institutional preparedness for embarking on
invasive species management and incorporating invasive
species management in wider environmental issues such
as pollution and island development. In doing so this
would ensure that invasive species management was not
marginalised against other critical development issues on
the island such as poverty and unemployment.
Lord Howe Island
Lord Howe Island, in Australia, was inscribed UNESCO
World Heritage status in 1982. Lord Howe Island is an
archipelago, comprising the primary island of the same
name and 27 secondary islands and islets, 600 km east of
Australia in the Tasman Sea. The island is administered as
part of the state of New South Wales and for legal purposes
is regarded as an unincorporated area administered by the
Lord Howe Island Board which reports to the New South
Wales Minister for Environment and Heritage. The resident
population of the island is around 350 people. Tourism is
694
the primary enterprise on Lord Howe Island but the Kentia
palm (Howea forsteriana) industry also contributes to the
local economy (Gillespie & Bennett, 2017).
The major invasive species on Lord Howe Island are
black rats and mice (Mus musculus) (Wilkinson & Priddell,
2011). Eradication of rodents from Lord Howe Island would
accrue both biodiversity (Hutton, et al., 2007) and economic
benefits (Gillespie & Bennett, 2017). It would specifically
facilitate reintroduction of the critically endangered Lord
Howe Island stick insect (Dryococelus australis) (Hutton,
et al., 2007) from its last remaining wild habitat on nearby
tiny, precipitous Ball’s Pyramid. Eradication of the rats
and mice on Lord Howe Island was first proposed in
2001 followed by a series of technical feasibility studies
(Saunders & Brown, 2001, Parkes, et al., 2003). Planning
commenced in 2006 (Wilkinson & Priddell, 2011) and a
draft eradication plan was published in 2009 (LHI Board,
2009). Whereas a number of other eradications of invasive
species have occurred on islands belonging to Australia
(Priddell, et al., 2011), the eradication of rodents on
Lord Howe Island would be the first to take place on an
inhabited island, particularly in the strict sense of our more
comprehensive definition of inhabitation (i.e. communities
and facilities). However, the original proposal to eradicate
rodents from Lord Howe Island was met with prolonged
resistance by elements of the island community.
Management of invasive species on Lord Howe
Island is undertaken in an island conservation context
which acknowledges the severe impact such species are
having on the biodiversity and economy of the island and
engages in lethal control. Nonetheless, on Lord Howe
Island resistance to rodent eradication was prolonged
from a lack of application of social tools (Russell, et al.,
2018), although at the time Lord Howe Island was one of
the first inhabited islands where rodent eradication was
being actively pursued. Ultimately, a number of tools from
SEA have now been applied independently, including an
environmental impact assessment (LHI Board, 2016),
economic cost-benefit analysis (Gillespie & Bennett,
2017), and human health risk assessment (O’Kane, 2017)
Strategic environmental assessment would have allowed
the planning of rodent eradication on Lord Howe Island
to take place using the most appropriate tools for engaging
with a resident community that had unanticipated levels
of hostility towards the overall proposal. Tools from an
SEA framework would have helped identify the various
underlying threads of the resistance to rodent eradication
in a community that was already accepting of lethal rodent
control for the same values at those proposing rodent
eradication. Whereas it was initially believed providing
more evidential information on the need for eradication
and the expected biodiversity benefits alone would be
sufficient to gain support for rodent eradication (Wilkinson
& Priddell, 2011), this is now known to play only a
small role in invasive species planning (Crowley, et al.,
2017a), and SEA would have provided tools for a greater
participatory process in the rodent eradication planning on
Lord Howe Island.
Summary
Although Fernando de Noronha and Lord Howe
Island are very similar in geography, they share only a
few consistencies in governance and structure, e.g. on
both islands the government remains the land-owner and
residents are all lease-holders. Otherwise, the generally
vast differences in cultures and governance (Reis &
Hayward, 2013) mean that planning for invasive species
management must be considered in very different contexts
on each island. On Fernando de Noronha SEA would have
fostered the consideration of invasive species impacts
within wider environmental and societal issues, whereas
Russell & Taylor: Environmental assessment for inhabited islands
on Lord Howe Island SEA would have provided guidance
on the appropriate tools for community engagement to
move beyond rodent control to eradication. Thus, the
over-arching framework of SEA would have been applied
differently on each island to reflect their different contexts
and experiences.
NEW ZEALAND ISLANDS
New Zealand has led the world in invasive mammal
eradications, with about one third of its islands having been
cleared of all invasive mammals (Towns, et al., 2013).
These successes have spurred the country to propose the
Predator Free New Zealand ambition to eradicate stoats
(Mustela erminea), rats and brushtail possums (Trichosurus
vulpecula) from the entirety of the archipelago by 2050
(Russell, et al., 2015). A necessary stepping stone to this
goal would entail removing invasive mammals from the
large offshore islands of Aotea (Great Barrier Island) and
Rakiura (Stewart Island), which would immediately raise
the amount of offshore island predator-free land area from
10% to 50%. Discussions and limited planning for invasive
mammalian predator eradication from both islands have
taken place but using different methods to understand
the wider context of, and barriers to, invasive mammal
eradication.
Aotea
Aotea comprises a main island of 27,761 ha and
numerous surrounding islands and islets, located 17 km
north-east from the northern North Island of New Zealand.
The island falls within the rohe (tribal boundaries) of Ngati
Rehua and has about 800 residents. Seventy percent of
the land is owned by the New Zealand Government and is
administered by the Department of Conservation. Invasive
mammalian predators include cats, black rats, Pacific rats
(R. exulans) and mice. Mustelids, brushtail possums and
hedgehogs (Erinaceus europaeus) are notably absent.
Large predator control projects at the sub-island level
currently occur at Windy Hill Sanctuary (770 ha) and
Glenfern Sanctuary (230 ha), and invasive mammals
have been removed from numerous surrounding offshore
islands (Clout & Russell 2006). A number of bird species
are currently at risk of island extirpation including redcrowned parakeets (Cyanoramphus novaezelandiae) and
tomtits (Petroica macrocephala), and the last remaining
kokako (Callaeas wilsoni) were removed in 1994 to nearby
Hauturu. Whole-island eradication of feral cats and rodents
was first proposed in 2003, but was met with prolonged
resistance by elements of the island community (Ogden &
Gilbert, 2011).
A number of tools from SEA have been applied
independently on Aotea to better understand the position of
the local community towards invasive mammalian predator
eradication. In 2015 a participatory process was initiated
in the community to understand community perspectives
and aspirations towards the overall ecology of the island
(McEntee & Johnson, 2015; McEntee & Johnson, 2016).
This participatory process identified that the community’s
perspective on invasive mammal eradication could not
be disassociated from their broader economic and social
aspirations, and that any investment in invasive mammal
eradication had to be part of a broader investment in the
community itself. It also identified underlying conflicts in
the community such as the tension between the value of
isolation versus the desire to increase tourism, and between
the desire to control invasive predators versus the value of
a toxin-free environment.
A social profiling exercise was also undertaken in 2015
alongside an assessment of the community’s attitudes to
invasive species management (Aley, 2016; Russell, et al.,
2018). This exercise found that there was a higher level
of uncertainty with respect to supporting eradication than
found on other neighbouring islands, but the social profile
of Aotea was not markedly dissimilar to other neighbouring
islands in the Hauraki Gulf, although all the islands
were markedly different from a corresponding sample in
neighbouring Auckland city. This suggested overall that
the community’s position on invasive mammal eradication
was potentially driven by unique recent experiences and
exposure to ideas, rather than anything in its social profile,
although there did appear to be an overriding island
archetype for all the islands in the study, even though one
had already had invasive rats eradicated from it (Russell,
et al., 2018).
Rakiura
Rakiura comprises a main island of 174,600 ha and
numerous surrounding islands and islets, located 27 km
south from the southern South Island of New Zealand. The
island falls within the rohe of Ngai Tahu and has about
450 residents. Eighty-five percent of the land is owned
by the New Zealand Government and is administered by
the Department of Conservation. Invasive mammalian
predators include cats, black rats, brown rats, Pacific rats,
brushtail possums and hedgehogs. Mustelids and mice are
notably absent. Large predator control projects at the subisland level currently occur at Mamaku Point Conservation
Reserve (172 ha; previously Dancing Star Conservation
Estate), and invasive mammals have been removed from
numerous surrounding offshore islands (Clout & Russell,
2006). Although a number of endangered bird species
rare on the main islands of New Zealand are abundant
on Stewart Island, the last remaining kakapo (Strigops
habroptilus) were removed in 1992 to nearby offshore
Whenua Hou. Whole-island eradication of feral cats and
rodents was first proposed in 2008 (Beaven, 2008), but was
also met with local resistance.
Rakiura is another case where a number of tools from
SEA have been applied independently in an ad hoc manner,
not preceded by any attempt to better understand the position
of the local community towards invasive mammalian
predator eradication. In 2013 a technical feasibility study
for removing all invasive mammal predators from Rakiura
was undertaken (Bell & Bramley, 2013). This technical
feasibility study found that the eradication of invasive
mammalian predators from Rakiura was not possible with
today’s technology, but a sub-island level project around
Halfmoon Bay would be feasible. Subsequently a subisland level project (4,800 ha) consisting of a predatorproof fence protecting the northern peninsula at Halfmoon
Bay was proposed as an interim step to achieving a
predator-free Rakiura, including technical reports on the
predator-proof fence design (Bell, 2014a) and predator
eradication methodology (Bell, 2014b). The report on the
fence design emphasised the necessity of a predator-proof
fence in order to achieve invasive mammalian predator
eradication on the peninsula, while the report on predator
eradication methodology presented a suite of options for
the community to be consulted upon.
In 2014, an economic cost-benefit assessment of
invasive mammalian predator eradication for both
Rakiura and Halfmoon Bay was also undertaken (Morgan
& Simmons, 2014). This report found that eradication
was unlikely to have a net positive economic gain from
tourism alone but became positive with the addition of
ecosystem service valuation. The report also emphasised
that anticipated economic and social benefits from invasive
mammal eradication may not necessarily eventuate unless
the community had a plan and processes in place to
capitalise upon them. Despite the substantial investment in
technical scoping and community lobbying for a predator695
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
free Rakiura and the Halfmoon Bay project, there remains
a level of resistance to both projects on the island along
with multiple local proposals and efforts towards enhanced
biodiversity (Russell, et al. 2017b).
Summary
The human communities on both Aotea and Rakiura exist
in a similar cultural space, and the islands have remarkably
similar ecological histories of bird loss, despite being at
opposite latitudes of New Zealand. However, both islands
illustrate the importance of drawing on the full set of tools
available in SEA to build a comprehensive understanding of
the perceived and real barriers to implementing an invasive
species eradication programme. On Aotea, an SEA approach
would have brought the technical and economic aspects of
predator eradication into the community discussion earlier,
alongside the social elements. When done properly this
could have reduced uncertainty in the technical aspects of
the proposed eradication, and addressed broader livelihood
elements, particularly with respect to the economy, which
are important issues on the island. In contrast, on Rakiura
an SEA approach would have identified much earlier in
the planning process the importance of including social
assessment alongside technical and economic costbenefit assessment, and drawn all three threads together
simultaneously to identify that the most immediate barriers
to predator eradication on Rakiura, or even in Halfmoon
Bay, reflect existing political structures and economic
development issues on the island.
DISCUSSION
In this paper we have outlined the process of SEA and
how it might specifically be applied to wildlife management,
with an emphasis on invasive species management and
eradication on inhabited islands. We have reflected on
lessons learnt from case studies on four inhabited islands
around the world where invasive species are the primary
threat to biodiversity, while also impacting on other
elements of island livelihoods. Strategic environmental
assessment captures a broad suite of tools, including EIA
and SIA. Not all of the tools which are a part of SEA may
need to be implemented on every island, and SEA allows
the application of more context-specific tools such as SIA,
and subsequent community engagement and collaborative
planning. Importantly, SEA is not a single, linear or one-off
process. As stated at the outset, it is an ongoing adaptive
and iterative process which adds value to and builds
capacity in existing systems. For eradications on inhabited
islands the target system is the island community itself,
including both human and non-human organisms. This
should come as no surprise as it is now readily accepted
that environments with humans in them must be managed
as joint socio-ecological systems (González, et al., 2008).
We encourage eradication project managers to identify
at the outset which SEA tools should be applied in any given
project (e.g. Crandall et al. 2018), and to implement those
tools in a consistent manner across projects. Governments
should also develop a standardised planning and reporting
process for invasive species eradication programmes.
However, it is important to note that SEA is not a panacea
to the challenges faced by practitioners wishing to
implement invasive species eradication programmes on
islands. Strategic environmental assessment can still be
prone to biases either towards values or technical evidence
in decision-making processes (Kørnøv & Thissen, 2000).
In some cases, whole-island eradication may simply not be
an optimal nor achievable goal, due to technical, ecological
social, economic or political barriers (Russell, et al.,
2015). This does not mean eradication should not remain
an aspirational goal (e.g. Predator Free New Zealand),
but that in the meantime focus is directed to conservation
696
interventions which maximise return-on-investment in the
broadest sense, e.g. invasive species management at the
sub-island level.
The application of SEA in a conservation context has
the added benefit of bringing wildlife management and
invasive species eradication more strongly into the ambit
of a broader application of SEA to island development.
This would enable the wider benefits of invasive species
eradication to be realised, such as on public health (de
Wit, et al., 2017) and in primary industry (Nimmo-Bell,
2009). It would also allow the benefits to be incorporated
into the international Sustainable Development Goals,
such as reduced inequalities through the more equitable
distribution of resources for invasive species eradication
across developed and developing island nations. For
instance, in small island developing states (Russell, et al.,
2017a), which are predominantly tropical and home to
unique biodiversity not found elsewhere and at risk from
invasive species such as mammalian predators (Russell &
Holmes, 2015). Undertaking an SEA approach to invasive
species eradication on islands will ultimately ensure the
longevity of eradications on islands, and alongside enabling
eradications on islands in the first instance, will have
immediate benefits in the implementation and maintenance
of biosecurity on islands (e.g. Russell, et al., 2017a).
ACKNOWLEDGEMENTS
Funding for components of this work were provided by
the Ministry of Education, Brazil CAPES grant PVE Project
Number 88881.065000/2014-1 and the Department of
Conservation, New Zealand contract 36467 to Uniservices.
Thanks to Carlos Abrahão, Andrew Walsh, Judy Gilbert
and reviewers for their constructive suggestions.
REFERENCES
Abrahão, C., Russell, J.C., Silva, J.C.R., Ferreira, F. and Dias, R. (2019).
‘Population assessment of a novel island invasive: tegu (Salvator
merianae) of Fernando de Noronha’. In: C.R. Veitch, M.N. Clout, A.R.
Martin, J.C. Russell and C.J. West (eds.) Island invasives: scaling up to
meet the challenge, pp. 317–325. Occasional Paper SSC no. 62. Gland,
Switzerland: IUCN.
Aley, J. (2016). ‘Environmental and Pest Management Attitudes of
Hauraki Gulf Island Communities’. MSc thesis. Auckland, New
Zealand: University of Auckland.
Allen, B.L., Cox, T.E., Fleming, P.J., Meek, P.D. and Russell, J.C. (2018).
‘Wildlife conservation management on inhabited islands’. Australasian
Journal of Environmental Management 25(1): 1–4.
Beaven, B. (2008). Scoping the Potential to Eradicate Rats, Cats and
Possums from Stewart Island/Rakiura. Oban: Stewart Island/Rakiura
Community & Environment Trust.
Bell, E.A. (2019). ‘It's Not All Up in the Air: The development and use
of ground-based rat eradication techniques in the UK’. In: C.R. Veitch,
M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) Island
invasives: scaling up to meet the challenge, pp. 79–87. Occasional
Paper SSC no. 62. Gland, Switzerland: IUCN.
Bell, P. and Bramley, A. (2013). Eliminating Predators from Stewart
Island: Scoping Report to Investigate Issues of Technical Feasibility.
Wellington, New Zealand: Department of Conservation.
Bell, P. (2014a). ‘Predator Free Halfmoon Bay Project: Technical Aspects
of the Project – Preliminary Outline’. Wellington, New Zealand:
Department of Conservation.
Bell, P. (2014b). Predator Free Halfmoon Bay Project: The Predator
Fence – Preliminary Outline’. Wellington, New Zealand: Department
of Conservation.
Bellard, C. and Jeschke, J.M. (2016). ‘A spatial mismatch between
invader impacts and research publications’. Conservation Biology
30(1): 230–232.
Carrion, V., Donlan, C.J., Campbell, K.J., Lavoie, C. and Cruz, F. (2011).
‘Archipelago-wide island restoration in the Galápagos Islands: reducing
costs of invasive mammal eradication programs and reinvasion risk’.
PLOS One 6(5): e18835.
Clout, M.N. and Russell, J.C. (2006). ‘The Eradication of Mammals from
New Zealand Islands’. In: F. Koike, M.N. Clout, M. Kawamichi, M. de
Poorter and K. Iwatsuki (eds.) Assessment and Control of Biological
Invasion Risks, , pp. 127–141. Gland, Switzerland: IUCN.
Courchamp, F., Hoffmann, B.D., Russell, J.C., Leclerc, C. and Bellard,
C. (2014). ‘Climate change, sea-level rise, and conservation: keeping
island biodiversity afloat’. Trends in Ecology & Evolution 29(3): 127–
130.
Russell & Taylor: Environmental assessment for inhabited islands
Crandall, S.G., Ohayon, J.L., de Wit, L.A., Hammond, J.E., Melanson,
K.L., Moritsch, M.M., Davenport, R., Ruiz, D., Keitt, B., Holmes, N.D.,
Packard, H.G., Bury, J., Gilbert, G.S. and Parker, I.M. (2018). ‘Best
practices: social research methods to inform biological conservation’.
Australasian Journal of Environmental Management 25(1): 6–23.
Crowley, S.L., Hinchliffe, S. and McDonald, R.A. (2017a). ‘Conflict
in invasive species management’. Frontiers in Ecology and the
Environment 15(3): 133–141.
Crowley, S.L., Hinchliffe, S. and McDonald, R.A. (2017b). ‘Invasive
species management will benefit from social impact assessment’.
Journal of Applied Ecology 54(2): 351–357.
de Oliveira, J.A.P. (2003). ‘Governmental responses to tourism
development: three Brazilian case studies’. Tourism Management 24(1):
97–110.
de Wit, L.A., Croll, D.A., Tershy, B., Newton, K.M., Spatz, D.R., Holmes,
N.D. and Kilpatrick, A.M. (2017). ‘Estimating burdens of neglected
tropical zoonotic diseases on islands with introduced mammals’. The
American Journal of Tropical Medicine and Hygiene 96(3): 749–757.
Dias, R.A., Abrahão, C.R., Micheletti, T., Mangini, P.R., de Oliveira
Gasparotto, V.P., de Jesus Pena, H.F., Ferreira, F., Russell, J.C. and
Silva, J.C.R. (2017). ‘Prospects for domestic and feral cat management
on an inhabited tropical island’. Biological Invasions 19(8): 2339–2353.
Flyvbjerg, B. (2014). ‘What you should know about megaprojects and
why: an overview’. Project Management Journal 45(2): 6–19.
Gillespie, R. and Bennett, J. (2017). ‘Costs and benefits of rodent
eradication on Lord Howe Island, Australia’. Ecological Economics
140: 215–224.
Glen, A.S., Atkinson, R., Campbell, K.J., Hagen, E., Holmes, N.D.,
Keitt, B.S., Parkes, J.P., Saunders, A., Sawyer, J. and Torres, H. (2013).
‘Eradicating multiple invasive species on inhabited islands: the next big
step in island restoration?’ Biological Invasions 15(12) :2589–2603.
González, J., Montes, C., Rodríguez, J. and Tapia, W. (2008). ‘Rethinking
the Galapagos Islands as a complex social-ecological system:
Implications for conservation and management’. Ecology and Society
13(2): 13.
Hanson, C.C., Jolley, W.J., Smith, G., Garcelon, D.K., Keitt, B.S., Little,
A.E. and Campbell, K.J. (2015). ‘Feral cat eradication in the presence
of endemic San Nicolas Island foxes’. Biological Invasions 17(4):
977–986.
Hashimoto, T. (2010). ‘Eradication and ecosystem impacts of rats in the
Ogasawara Islands’. In: K. Kawakami and I. Okochi (eds.) Restoring
the Oceanic Island Ecosystem, pp.153–159. Tokyo, Japan: Springer.
Holmes, N.D., Griffiths, R., Pott, M., Alifano, A., Will, D., Wegmann, A.S.
and Russell, J.C. (2015). ‘Factors associated with rodent eradication
failure’. Biological Conservation 185: 8–16.
Howald, G., Donlan, C., Galván, J.P., Russell, J.C., Parkes, J., Samaniego,
A., Wang, Y., Veitch, D., Genovesi, P., Pascal, M., Saunders, A. and
Tershy, B. (2007). ‘Invasive rodent eradication on islands’. Conservation
Biology 21(5): 1258–1268.
Hutton, I., Parkes, J.P. and Sinclair, A.R.E. (2007). ‘Reassembling island
ecosystems: the case of Lord Howe Island’. Animal Conservation 10(1):
22–29.
IBGE. 2016. Censo Demográfico Fernando de Noronha (Pernambuco).
<http://cod.ibge.gov.br/V6F>.
Jones, H.P., Holmes, N.D., Butchart, S.H., Tershy, B.R., Kappes,
P.J., Corkery, I., Aguirre-Muñoz, A., Armstrong, D.P., Bonnaud, E.,
Burbidge, A.A., Campbell, K., Courchamp, F., Cowan, P.E., Cuthbert,
R.J., Ebbert, S., Genovesi, P., Howald, G.R., Keitt, B.S., Kress, S.W.,
Miskelly, C.M., Oppel, S., Poncet, S., Rauzon, M.J., Rocamora, G.,
Russell, J.C., Samaniego-Herrera, A., Seddon, P.J., Spatz, D.R., Towns,
D.R. and Croll, D.A. (2016). ‘Invasive mammal eradication on islands
results in substantial conservation gains’. Proceedings of the National
Academy of Sciences, USA 113(15): 4033–4038.
Keitt, B., Griffiths, R., Boudjelas, S., Broome, K., Cranwell, S., Millett,
J., Pitt, W. and Samaniego-Herrera, A. (2015). ‘Best practice guidelines
for rat eradication on tropical islands’. Biological Conservation 185:
17–26.
Kørnøv, L. and Thissen, W.A. (2000). ‘Rationality in decision-and policymaking: Implications for strategic environmental assessment’. Impact
Assessment and Project Appraisal 18(3): 191–200.
Lord Howe Island Board (2009). Draft Lord Howe Island Rodent
Eradication Plan. Lord Howe Island: Lord Howe Island Board.
Lord Howe Island Board (2016). Lord Howe Island Rodent Eradication
Project – Public Environment Report. Lord Howe Island: Lord Howe
Island Board.
Magalhães, F.J., Ribeiro-Andrade, M., Souza, F.M., Lima Filho, C.D.,
Biondo, A.W., Vidotto, O., Navarro, I.T. and Mota, R. A. (2017).
‘Seroprevalence and spatial distribution of Toxoplasma gondii infection
in cats, dogs, pigs and equines of the Fernando de Noronha Island,
Brazil’. Parasitology International 66(2): 43-46.
McEntee, M. and Johnson, S. (2015). Enabling an Ecological Vision for
Aotea Great Barrier Island: Understanding Community Perspectives
and Aspirations. Auckland, New Zealand: Aranovus Ltd.
McEntee, M. and Johnson, S. (2016). Aotea Great Barrier Island Ecology
Vision: Weaving the Tapestry. Auckland, New Zealand: Aranovus Ltd.
Morgan, G. and Simmons, G. (2014). Predator-Free Rakiura: An
Economic Appraisal. Wellington, New Zealand: The Morgan
Foundation.
Morgan, R.K. (2012). ‘Environmental impact assessment: The state of the
art’. Impact Assessment and Project Appraisal 30(1): 5–14.
Nimmo-Bell. (2009). Economic Costs of Pests to New Zealand. Technical
Paper No: 2009/31. Wellington, New Zealand: MAF Biosecurity.
O’Kane, M. (2017). Report on the Human Health Risk Assessment for
the Lord Howe Island’s Proposed Rodent Eradication Program. NSW
Chief Scientist & Engineer.
OECD (2006). Applying Strategic Environmental Assessment: Good
Practice Guidance for Development Co-operation. Organisation for
Economic Co-operation and Development. DAC Guidelines and
Reference Series, OECD Publishing.
Ogden, J. and Gilbert, J. (2011). ‘Running the gauntlet: advocating rat and
feral cat eradication on an inhabited island – Great Barrier Island, New
Zealand’. In: C.R. Veitch, M.N. Clout and D.R. Towns (eds.) Island
invasives: eradication and management, pp. 467–471. Occassional
Paper SSC no. 42. Gland, Switzerland: IUCN and Auckland, New
Zealand: CBB.
Oppel, S., Beaven, B.M., Bolton, M., Vickery, J. and Bodey, T.W. (2011).
‘Eradication of invasive mammals on islands inhabited by humans and
domestic animals’. Conservation Biology 25(2): 232–240.
Parkes, J., Ruscoe, W., Fisher, P. and Thomas, B. (2003). Benefits,
Constraints, Risks and Costs of Rodent Control Options on Lord Howe
Island. Unpublished report to the Lord Howe Island Board. Lincoln,
New Zealand: Landcare Research.
PCE. (2017). Taonga of an Island Nation: Saving New Zealand's Birds.
Wellington, New Zealand: Parliamentary Commissioner for the
Environment.
Priddel, D., Carlile, N., Wilkinson, I. and Wheeler, R. (2011). ‘Eradication
of exotic mammals from offshore islands in New South Wales,
Australia’. In: C.R. Veitch, M.N. Clout and D.R. Towns (eds.) Island
invasives: eradication and management, pp. 337–344. Occassional
Paper SSC no. 42. Gland, Switzerland: IUCN and Auckland, New
Zealand: CBB.
Reis, A. and Hayward, P. (2013). ‘Pronounced particularity: A comparison
of governance structures on Lord Howe Island and Fernando de
Noronha’. Island Studies Journal 8(2): 285–298.
Russell, J.C. and Holmes, N.D. (2015). ‘Tropical island conservation: rat
eradication for species recovery’. Biological Conservation 185: 1–7.
Russell, J.C., Innes, J.G., Brown, P.H. and Byrom, A.E. (2015). ‘Predatorfree New Zealand: Conservation country’. BioScience 65(5): 520–525.
Russell, J.C., Cole, N.C., Zuël, N. and Rocamora, G. (2016a). ‘Introduced
mammals on Western Indian Ocean islands’. Global Ecology and
Conservation 6: 132–144.
Russell, J.C., Jones, H.P., Armstrong, D.P., Courchamp, F., Kappes, P.J.,
Seddon, P.J., Oppel, S., Rauzon, M.J., Cowan, P.E., Rocamora, G.,
Genovesi, P., Bonnaud, E., Keitt, B.S., Holmes, N.D., Tershy, B.R.
(2016b). ‘Importance of lethal control of invasive predators for island
conservation’. Conservation Biology 30(3): 670–672.
Russell, J.C., Meyer, J.Y., Holmes, N.D. and Pagad, S. (2017a). ‘Invasive
alien species on islands: impacts, distribution, interactions and
management’. Environmental Conservation 44(4): 359–370.
Russell, J.C., Taylor, C.N. and Aley J.P. (2018). ‘Social assessment
of inhabited islands for wildlife management and eradication’.
Australasian Journal of Environmental Management 25(1): 24–42.
Russell, K.J., Taylor, C.N., Balanovic, J.X., Aley, J.P., Harbow, M.A. and
Russell, J.C. (2017b). Predator Free Rakiura Social Impact Assessment.
A report for the Department of Conservation. Auckland, New Zealand:
Uniservices Ltd.
Saunders, A. and Brown, D. (2001). An Assessment of the Feasibility of
Eradicating Rodents from the Lord Howe Island Group. Unpublished
report to the Lord Howe Island Board.
Sims, L. (2012). ‘Taking a learning approach to community-based
strategic environmental assessment: results from a Costa Rican case
study’. Impact Assessment and Project Appraisal 30(4): 242–252.
Taylor, C.N., Bryan, C.H., Goodrich, C.G. (2004) Social Assessment:
Theory, Process and Techniques. Middleton, Wisconsin: Social Ecology
Press.
Taylor, C.N. and Mackay, M. (2016). ‘Practice issues for integrating
strategic social assessment into the setting of environmental limits:
insights from Canterbury, New Zealand’. Impact Assessment and
Project Appraisal 34(2): 110–116.
Tetlow, M.F. and Hanusch M. (2012). ‘Strategic environmental
assessment: the state of the art’. Impact Assessment and Project
Appraisal 30(1): 15–24.
Towns, D.R., West, C.J. and Broome, K.G. (2013). ‘Purposes, outcomes
and challenges of eradicating invasive mammals from New Zealand
islands: an historical perspective’. Wildlife Research 40(2): 94–107.
Varnham, K., Glass, T. and Stringer, C. (2011). ‘Involving the community
in rodent eradication on Tristan da Cunha’. In: C.R. Veitch, M.N. Clout
and D.R. Towns (eds.) Island invasives: eradication and management,
pp. 504–507. Occassional Paper SSC no. 42. Gland, Switzerland: IUCN
and Auckland, New Zealand: CBB.
Wilkinson, I.S. and Priddel, D. (2011). ‘Rodent eradication on Lord Howe
Island: challenges posed by people, livestock, and threatened endemics’.
In: C.R. Veitch, M.N. Clout and D.R. Towns (eds.) Island invasives:
eradication and management, pp. 508–514. Occassional Paper SSC no.
42. Gland, Switzerland: IUCN and Auckland, New Zealand: CBB.
697
A. Tye
Tye, A. Towards a guidance document for invasive species planning and management on islands
Towards a guidance document for invasive species planning and
management on islands
A. Tye
International Union for Conservation of Nature, COI Blue Tower 3rd floor, Rue de l’Institut, Ebène, Mauritius. Present
address: 2 School Lane, King’s Ripton PE28 2NL, UK. <alantye@gmail.com>.
Abstract In 2012 a process was initiated to produce a guidance document for invasive species management on islands,
as an objective of a regional invasive species project in the Western Indian Ocean (WIO) islands, implemented by IUCN.
The consultative process for producing the document began with requests and discussions via regional and global island
and invasives email distribution lists. Initial responses revealed a consensus on the need for a guidance document for
programmatic planning. A draft was therefore constructed around existing Pacific regional guidelines and a draft manual
that had initially been written for the WIO, with new supplementary sections suggested by respondents. The new draft
was discussed and revised in workshops at two international conferences. The document is now organised into three main
sections: the first on how to use it, the second a checklist of the essential components of a comprehensive island invasives
programme (to ensure nothing is overlooked when planning), and the third providing detailed guidance on the planning
and decision-making processes. The document is intended to provide a comprehensive framework and procedural guide
for invasive species planning on islands. Further consultations took place by email, and a later draft was tested by a
number of users writing various kinds of invasive species strategy and action plan. Publication will be in English, French
(both published 2018) and Spanish (scheduled for 2019).
Keywords: best practice, consultative planning, NBSAP, networks, NISSAP, prioritisation
INTRODUCTION
FIRST STEPS
The intentional and unintentional movement by people
of organisms around the world, many of which become
invasive in the areas to which they are introduced, is an
international problem of particular concern to islands.
The effective management of invasive species on islands
therefore requires comprehensive and coordinated action
by international agencies, governments, NGOs, the private
sector and local communities. Small islands and developing
states do not have the resources to tackle all invasive
threats by themselves, and in some regions collaborative
regional plans and strategies have been developed to
promote coordinated planning and action between islands
and nations and also to guide international agencies in
providing assistance to them. Many countries have also
developed National Invasive Species Strategies and Action
Plans (NISSAPs), as have a number of individual islands
(for brevity, all such plans are herein termed NISSAPs). The
Convention on Biological Diversity recognises invasive
species as a serious threat, including in its Aichi Target 9,
and encourages countries to include plans for managing
invasive species in their National Biodiversity Strategies
and Action Plans (NBSAPs). However, the NBSAPs and
NISSAPs of different islands and island countries vary
greatly in their comprehensiveness in dealing with invasive
species problems (Doherty & Boudjelas, 2010; Boudjelas,
in press).
The Inva’Ziles project began with a broad interpretation
of its commitment to produce a guidance document for
invasives management on islands, by compiling a first
draft of a manual attempting to cover the whole range of
actions necessary for an invasive species programme, in
the following chapters:
The Inva’Ziles Project, implemented by IUCN from
2012 to 2018, provided assistance to the islands of the
western Indian Ocean (WIO) region in managing biological
invasions. One of the project’s objectives was to develop
guidance for islands and island nations to help them prevent
and manage the spread of invasive species and reduce their
impacts on biodiversity and people’s livelihoods. This
paper describes the process leading to the production of
a guidance document specifically for invasive species
planning on islands worldwide. It explains the purpose of
the document and outlines progress towards its publication,
including input from the 3rd Island Invasives conference in
Dundee, July 2017.
INTRODUCTION
● Importance of biological diversity
● Significance of biological invasion as a disruption of
biodiversity
BIOLOGICAL INVASION AS A PROCESS
ELEMENTS OF AN INVASIVES SPECIES
STRATEGY
●
●
●
●
●
●
Regional coordination and exchange
Risk Assessment
Prevention without quarantine
Pathways of introduction
Early detection and rapid response
Management of established invasions
MONITORING
CAPACITY
●
●
●
●
Institutions
Awareness
Information
Conventions
IMPLEMENTATION OF INVASIVES SPECIES
MANAGEMENT
●
●
●
●
Policies, laws and regulations
Institutions and capacity
Roles and responsibilities of the public
International and regional responsibilities
GLOBAL CHANGE AND INVASION
In: C.R. Veitch, M.N. Clout, A.R. Martin, J.C. Russell and C.J. West (eds.) (2019). Island invasives: scaling
698
up to meet the challenge, pp. 698–702. Occasional Paper SSC no. 62. Gland, Switzerland: IUCN.
Tye: Guidance for invasive species planning
While all these topics are important, there are good
resources already available to help planners and managers
with many of these activities, including project design,
border biosecurity, methods of controlling various
species of invasive animal, plant and other organisms,
raising awareness, etc. Examples include the guidelines
and toolkits of the Global Invasive Species Programme
(www.issg.org/gisp_guidelines_toolkits.htm) on marine
biofouling (Jackson, 2008), marine pest management
(Hilliard, 2005), legal and institutional frameworks (Shine,
et al., 2000; Shine, 2008), best prevention and management
practice (Preston, et al., 2000; Wittenberg & Cock, 2001),
and economic analysis (Emerton & Howard, 2008) along
with their accompanying training courses (www.issg.org/
gisp_training_coursematerials.htm). There are also many
excellent materials developed in individual regions, such as
the rodent and cat eradication resource kits for the Pacific
(Pacific Invasives Initiative, 2011) and UK (GB Nonnative Species Secretariat, 2017), as well as the Pacific
kits for invasive plant (Pacific Invasives Initiative, 2015)
and ant (New Zealand MFAT, 2016) management and their
accompanying training courses. It would be impossible
within a single document to improve on all of these and
others. Further, it was considered doubtful whether general
explanations of biological invasions and their impacts on
biodiversity would be necessary for the intended primary
users of the document: invasive species planners, managers
and researchers on islands. The introductory material
covering these topics and the discursive style adopted in
the first Inva’Ziles draft limited the amount and clarity of
the guidance provided; for example, the draft did not give
clear guidance on the steps to be taken when planning an
invasives programme, nor on how to prioritise when faced
with many problems and limited financial and human
resources. It was felt that a short document with a clear
purpose and direct guidance would be more useful and
used than something longer and more discursive.
It was therefore decided to carry out consultation in
order to find out what kind of guidance invasive species
workers themselves thought they needed most, so as to
be able to focus the planned document more precisely on
priority gaps in available resources.
CONSULTATION AND REDRAFTING
Given the Inva’Ziles Project’s primary responsibility to
provide assistance to the WIO region, an initial consultation
was carried out by e-mailing a simple questionnaire to the
c. 325 members of the Western Indian Ocean Network
on Invasive Species (WIONIS), asking what kind of
guidance they felt was most needed. It was essential to
give respondents an idea of what might be possible for the
project to produce within the limitations of its timespan and
budget, so, to encourage realistic answers three possibilities
were suggested: a manual-style document resembling the
first draft produced by Inva’Ziles, something focused more
precisely on the planning and decision-making processes,
using the example of the Guidelines for Invasive Species
Management in the Pacific (SPREP, 2009: hereafter termed
the ‘Pacific Guidelines’), or something else.
This was followed by a similar worldwide consultation
using the following global and regional e-mail distribution
lists: aliens-l (1,400 subscribers, global), islands-l (360,
global), carib-ias (310, Caribbean), the Pacific Invasives
Initiative list (1,210), and the Pacific Invasives Partnership
(c. 40). In addition, the same request was sent to specifically
compiled lists of known contacts in the Atlantic and
Mediterranean islands (c. 20 people).
These consultations generated responses from invasive
species planners, scientists and managers, including
experts in all major island biomes, marine and terrestrial
(all contributors up to the submission date of the present
article are named in the Acknowledgments). Of the 43
respondents who indicated a clear preference for the
kind of document they would like to see developed,
two wanted an operations manual for field management
and 41 preferred guidance on planning, with no-one
suggesting any other kind of document. These choices,
taken together with written comments from many other
respondents, indicated a consensus that specific guidance
on programmatic planning was scarce and lacking in
detail, and that this represented a particular resource gap.
The Pacific Guidelines have been widely adopted and used
in that region, and many respondents felt that an updated
and internationalised version of this would be highly
appropriate for other island regions.
The decision was therefore made to produce a document
addressing this need for planning guidance, taking the
Pacific Guidelines as a model, updating and hopefully
improving it, and at the same time endeavouring to make
the document as useful and relevant as possible to islands
worldwide. A skeleton was then produced by adapting the
text of the Pacific Guidelines for a global set of users, and
adding ideas for new sections suggested by the drafting
team, questionnaire respondents and others. The new
sections were then partially populated by adapting text
from the Inva’Ziles first draft manual.
To expand the consultation process, we used
opportunities created by international and regional
meetings to obtain further input. Workshops were therefore
organised at the IUCN ‘World Conservation Congress’
(WCC) in Hawai’i, September 2016, and the 3rd ‘Island
Invasives’ conference (3II), Dundee, July 2017. The first of
these meetings attracted (as expected) a broad cross-section
of conservationists, while the second drew a substantially
different group, consisting primarily of invasive species
management practitioners and researchers. Both meetings
generated contributions from people working on a wide
range of aspects of the invasives threat to islands, from all
parts of the world.
At the WCC, two events were organised with the
objective of obtaining input. First, the IUCN held a
major introductory event on ‘Islands at risk: meeting
the global challenge of Invasive Alien Species’, at which
three initial presentations (one on the guidance document)
were followed by work-groups on the three topics. The
guidance work-group attracted some 30 people, of whom
14 offered to make additional contributions later, as the
drafts developed. The second WCC event was a roundtable
discussion organised by the Pacific Invasives Partnership,
which attracted about 20 people, most of whom had not
attended the first working session. At both of these sessions,
input was obtained not only for the global guidance
document, but also for a planned revision of the Pacific
Guidelines, led by the Pacific Invasives Partnership.
Comments and ideas received at the WCC were
incorporated into a second draft, which included
supplementary sections solicited meanwhile from
volunteer experts on particular topics. During this process
it became clear that guidance on two areas in particular was
desired: the planning process itself, including prioritisation
and decision-making, and how to increase support for
invasives management among politicians, their electorates
(the public), and local communities experiencing problems
caused by invasives. As a result, these two areas grew to
constitute the largest supplementary sections.
At the 3II, the IUCN gave an introductory presentation
in plenary to explain the purpose of a working session on
the guidelines that evening. Some 50 people came to the
evening session (approximately 15% of the conference
attendees), which was organised into three work-groups
699
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
covering different sections of the draft, namely: planning
and decision-making; awareness, support and capacity;
research and practical management. Twenty of those who
attended offered to contribute further.
The steps towards producing this document are
illustrated in Fig. 1, and the location or geographical
interest of the identified contributors summarised in Table
1. At the time of writing this paper, we were in the process
of incorporating comments from the 3IIsland A major
outcome from 3II was confirmation from practitioners
that the fundamental need for this document was genuine
and widespread, and also that guidance on how to use the
document should be given clearly within.
THE CURRENT DRAFT
The aim thus became to provide a comprehensive
framework and procedural guide for anyone planning an
invasives programme on islands, including international
and regional agencies, conservation NGOs, relevant
government agencies (agriculture, biosecurity, environment
…), conservation managers, research planners, and anyone
else who has to find, plan and prioritise funds and resources
for invasives management.
The latest and final draft met the target limit of 48
pages plus covers (the Pacific Guidelines comprises 24
pages including covers), has now been organised into three
main sections (plus a “Resources” section). The first of
the main sections explains the purpose of the document,
how to use it, and who the intended users are. The second
section is a checklist of the essential components of a
comprehensive island invasives programme, to ensure
nothing is overlooked when planning (this part still
resembles the Pacific Guidelines, which consists mainly
of such a checklist). The third section describes in detail
how to conduct the processes mentioned by many people
as being particularly problematic, especially how to
plan, how to prioritise, how to make decisions, and how
to increase collaboration, support and involvement by
different target groups ranging from local communities to
senior policy- and decision-makers. Throughout, there are
links to additional resources on each topic.
The document provides decision-making guidance at
both programmatic planning and field project planning
levels, including how to prioritise, how to choose
management goals, and how to win political and community
support for the actions planned. It should help international
agencies to identify their niche for invasives work on
islands and to identify island priority needs that match their
agency’s expertise. It will help national and local agencies
and managers to identify and prioritise actions within their
jurisdiction, design a NISSAP, benefit from the experience
of other countries and organizations, and justify projects
to decision-makers and donors. Content of the three main
sections is organised as follows:
INTRODUCTORY MATERIAL
● Purpose of the document, intended users, how to use
● Background
THE GUIDELINES CHECKLIST
● Foundations (planning, decisions, support, capacity,
legal)
● Information (baseline, monitoring, prioritisation,
research)
● Management (borders, established invaders,
restoration)
Fig. 1 Timeline of the process of
producing the guidance document.
700
Tye: Guidance for invasive species planning
HOW TO PLAN
ACKNOWLEDGEMENTS
● Getting people involved, raising support for the plan,
mutual help networks
● Programme planning, NISSAPs and others
● Prioritising, hard decisions, decision tools, risk
analysis
● Neglected areas
● Planning for global change
● Project planning, other decision tools
The Inva’Ziles Project (official title Preparation
and testing of a comprehensive model for preventing
and managing the spread of invasive species on island
ecosystems) was implemented by the International
Union for Conservation of Nature (IUCN), funded by
the European Union and hosted by the Indian Ocean
Commission. Geoffrey Howard and Olivier Tyack (IUCN)
drew up the initial concept and first draft of the guidance
document. Kosi Latu (SPREP) kindly gave permission to
use the Pacific Guidelines as a model. Kevin Smith, Olivier
Hasinger (IUCN) and David Moverley (SPREP) organised
working sessions at the 2016 ‘World Conservation
Congress’, and Kevin, together with Souad Boudjelas
(Pacific Invasives Initiative) and Jill Key (GB Non-Native
Species Secretariat), also organised the session at the
3rd ‘Island Invasives’ conference; they have all provided
invaluable support and contributions throughout. Dick
Veitch encouraged me to write this article for the current
proceedings, Katharina Lapin (IUCN) constructed Fig. 1,
and Olivier Hasinger and Kevin Smith reviewed a draft. The
following had contributed to drafts by the date of submission
of this article by providing texts, commenting on versions,
suggesting areas of concern, or in other ways (this list does
not include people who contributed at conferences but did
not leave their details): Ademola Ajagbe, Katy Beaver,
Alex Bond, Elsa Bonnard, Olaf Booy, Rafael Borroto,
Souad Boudjelas, Nancy Bunbury, Earl Campbell, Dario
Capizzi, Juli Caujapé-Castells, Alison Copeland, Ana
Costa, Franck Courchamp, Phil Cowan, Steve Cranwell,
Cathleen Cybèle, Curt Daehler, Maria Cristina Duarte,
Julia Dunn, Rui Bento Elias, Marko Filipovic, Julian Fitter,
Frauke Fleischer-Dogley, Jason Goldberg, Ines Gómez,
Viliami Hakaumotu, Sjurdur Hammer, Olivier Hasinger,
Ben Hoffmann, Geoffrey Howard, Stephanie Hudin, Jason
Jack, Patricia Jaramillo, Marie-May Jeremie-Muzungaile,
Gabe Johnson, Chris Kaiser-Bunbury, Springer Kaye,
Inti Keith, John Kelly, Jill Key, Michael Kiehn, Cynthia
Kolar, Christoph Kueffer, Janice Lord, Ian MacDonald,
Gwen Maggs, Christy Martin, Kelly Martinou, John
Mauremootoo, Mathilde Meheut, Tommy Melo, JeanYves Meyer, Joel Miles, Aileen Mill, James Millett, Nitya
Mohanty, Craig Morley, David Moverley, Bradley Myer,
Rachel Neville, Ray Nias, Kimberley O’Connor, Warea
Orapa, Shyama Pagad, Julián Pérez, John Pinel, Bruce
Potter, Parmenanda Ragen, Frida Razafinaivo, Tim Riding,
Gérard Rocamora, James Russell, Susana Saavedra, Adrian
Schiavini, Richard Selman, Nirmal Shah, Andy Sheppard,
Greg Sherley, Junko Shimura, Didier Slachmuylders,
Kevin Smith, Antonio Soares, Yohann Soubeyran, Vikash
Tatayah, Anna Traveset, Olivier Tyack, Magdalena Vicens,
The identification of these priority areas for guidance
has largely been determined by the views of the respondents.
These priorities differ somewhat from the critical areas for
action identified almost 20 years ago in the Pacific, when
the following were considered to need special attention
(SPREP, 2000):
● Shortage and inaccessibility of information on
invasive species and best management practice
● Lack of awareness of the impacts of invasive
species
● Insufficient networking, coordination and
collaboration
● Inadequate legislation, regulations, cross-sectoral
policies, and enforcement
● Shortage of trained personnel, and inadequate
facilities
● Insufficient funding.
The three items in bold are closely related to the current
priority needs identified by our recent consultations. The
differences reflect both the fact that some of the other areas,
particularly best management practices, have since been
addressed by resources specially designed to assist with
them, but also the fact that our new document is aimed
at perceived needs for guidance itself, rather than at other
kinds of need (e.g. adequate facilities, trained personnel,
laws etc.).
FINAL STEPS
The final draft was circulated once more to the core
group of committed contributors (ultimately just over
100 people contributed) as well as to all of the e-mail
distribution lists cited above. Special contributions were
solicited from experts on particular themes. The later drafts
were tested in a number of planning processes, including
for the first NISSAP of the Comoro Islands in mid-2018.
The English and French versions were published in print
and online in mid-2018 and the Spanish version published
online in early 2019.
Table 1 Islands and island regions represented by the 96 identified contributors so far (each contributor assigned
to only one category). OIT signifies overseas island territories of any kind, irrespective of their political status;
n = no. of contributors.
Islands
Australia
Azores
Bermuda
Canaries
Cape Verde Islands
Caribbean
Cuba
France & OITs
Galapagos
n
1
3
1
2
2
1
1
4
2
Islands
Indian Ocean
Isle of Man
Japan
Kosrae
Lord Howe Island
Madagascar
Mauritius
Mediterranean
New Zealand
n
3
1
1
1
1
1
4
3
3
Islands
Pacific
Palau
Papua New Guinea
Seychelles
Tonga
UK & OITs
USA, Hawai’i & OITs
Global, multi-regional or
unknown
n
8
1
1
7
1
4
9
30
701
Island invasives: scaling up to meet the challenge. Ch 3D Strategy: Scaling up
Jeanne Wagner, Josua Wainiqolo, Katherine Walls, Andrew
Walsh, Masahito Yoshida, Glyn Young, Kristi Young.
Thanks to all, and if I’ve missed anyone, please let me
know!
REFERENCES
Boudjelas, S. (in press). A review of national and island plans for the
management of invasive species in the Western Indian Ocean region.
Pp. XX–XX in Tye, A. (ed.) The State of Knowledge and Planning for
Invasive Species Management in the Western Indian Ocean. IUCN,
Gland.
Doherty, N. and Boudjelas, S. (2010). Invasive Species Management in the
Pacific: A Review of National Plans and Current Activities. Unpublished
report for the Pacific Invasives Partnership. Auckland: Pacific Invasives
Initiative. <www.issg.org/cii/Electronic%20references/pii/references/
pii_ism_in_the_pacific_a_review_of_national_plans_and_current_
activities.pdf>.
Emerton, L. and Howard, G. (2008). A Toolkit for the Economic Analysis
of Invasive Species. Global Invasive Species Programme.
GB Non-native Species Secretariat. (2017). UK Rodent Eradication Best
Practice Toolkit. <www.nonnativespecies.org/index.cfm?pageid=613>.
Hilliard, R. (2005). Best Practice for the Management of Introduced
Marine Pests - A Review. Global Invasive Species Programme.
Jackson, L. (2008). Guidelines for the Prevention and Management of
Marine Biofouling and Invasive Species. Global Invasive Species
Programme.
New Zealand MFAT. (2016). Pacific Invasive Ant Toolkit. <www.piat.org.
nz/>.
Pacific Invasives Initiative (2011). Resource Kit for Rodent and Cat
Eradication. <http://rce.pacificinvasivesinitiative.org/>.
Pacific Invasives Initiative. (2015). Resource Kit for Invasive Plant
Management.
<http://pacificinvasivesinitiative.org/ipm/ipm.
pacificinvasivesinitiative.org/index.html>.
Preston, G., Brown, G. and van Wyk, E. (eds.) (2000). Best Management
Practices for Preventing and Controlling Invasive Alien Species. Cape
Town: Working for Water Programme.
Shine, C. (2008). A Toolkit for Developing Legal and Institutional
Frameworks for Invasive Alien Species. Global Invasive Species
Programme.
Shine, C., Williams, N. and Gündling, L. (2000). A Guide to Designing
Legal and Institutional Frameworks on Alien Invasive Species.
Environmental Policy and Law Paper No. 40. Gland, Switzerland:
IUCN.
SPREP. (2000). Draft Invasive Species Strategy for the Pacific Islands
Region. Apia: Secretariat of the Pacific Regional Environment
Programme.
SPREP. (2009). Guidelines for Invasive Species Management in
the Pacific. Apia: Secretariat of the Pacific Regional Environment
Programme.
Wittenberg, R. and Cock, M.J.W. (eds.) (2001). Invasive Alien Species: A
Toolkit for Best Prevention and Management Practices. Global Invasive
Species Programme.
702
Chapter 4: Abstracts
These abstracts are for papers which were presented at the
conference, either as oral presentations or poster papers,
but for which the authors have chosen not to prepare and
publish a full written paper.
These abstracts are given in the alphabetical order of the
prime author of the paper with the address of only that first
author included.
703
Aguirre-Muñoz, A.; F. Méndez-Sánchez, L. Luna-Mendoza, A. Ortiz-Alcaraz, J. Hernández-Montoya, Y. Bedolla-Guzmán, M. Latofski-Robles, E. Rojas-Mayoral, N. Silva-Estudillo,
F. Torres-García, M. Félix-Lizárraga, A. Fabila-Blanco, A. Hernández-Ríos, E. Bravo-Hernández, F. Solís-Carlos, C. Jáuregui-García and D. Munguía-Cajigas. Mexico’s progress and commitment
to comprehensive island restoration
Mexico’s progress and commitment to comprehensive island restoration
A. Aguirre-Muñoz, F. Méndez-Sánchez, L. Luna-Mendoza, A. Ortiz-Alcaraz, J. Hernández-Montoya,
Y. Bedolla-Guzmán, M. Latofski-Robles, E. Rojas-Mayoral, N. Silva-Estudillo, F. Torres-García, M. Félix-Lizárraga,
A. Fabila-Blanco, A. Hernández-Ríos, E. Bravo-Hernández, F. Solís-Carlos, C. Jáuregui-García and
D. Munguía-Cajigas
Grupo de Ecología y Conservación de Islas, A.C., Avenida Moctezuma 836 Zona Centro, Ensenada, Baja California
22800 Mexico. <alfonso.aguirre@islas.org.mx>
For the past 18 years, Mexico has taken bold steps to systematically eradicate invasive mammals. Mexico´s 4,111 islands
host 8.3% of the country’s plants and land vertebrates. They harbour one in three seabirds worldwide, placing Mexico
as the third most diverse country. Invasive mammals have had a big toll on Mexico’s biodiversity, with 17 out of 21
confirmed vertebrate extinctions occurring on islands. The Mexican conservation organisation Grupo de Ecología y
Conservación de Islas (GECI), in collaboration with Mexico’s federal government, and a wide network of national and
international donors, has been leading the National Programme for Island Restoration that has grown in scope. The first
eradications on small islands fostered trust amongst partners, setting the foundations for complex eradications on bigger
islands requiring innovation, capacity development, and research. Island biosecurity is now a priority for long-term
tangible results. This programme evolved to be truly comprehensive, including post-eradication restoration to strengthen
island resilience, and the social construction of a cultural approach integrating interests from conservation and local
fishing communities. Results to date include: (1) eradication of 58 populations of invasive mammals from 37 islands;
(2) publication of both a National Island and Invasive Species Strategy, identifying conservation priorities; (3) ongoing
active restoration of seabird colonies and native plant communities; (4) original applied research and ad hoc infrastructure
and equipment to support restoration; (5) legal protection of all Mexican islands; (6) assessing the effects of climate
change on islands’ biodiversity and human populations; and (7) formation of in-house specialists through postgraduate
studies in collaboration with research institutes and universities from Mexico and elsewhere. As for the future, we foresee
two priorities: (1) remove invasive mammals from all Mexican islands by 2030; and (2) promote the creation of an
“International Islands Institute” that could operate under a wide international collaboration and interdisciplinary approach.
A. Aguirre-Muñoz, F. Méndez-Sánchez, L. Luna-Mendoza, A. Ortiz-Alcaraz, J. Hernández-Montoya, Y. Bedolla-Guzmán, M. Latofski-Robles, E. Rojas-Mayoral, N. Silva-Estudillo, F. Torres-García,
M. Félix-Lizárraga, A. Fabila-Blanco, A. Hernández-Ríos, E. Bravo-Hernández, F. Solís-Carlos, C. Jáuregui-García and D. Munguía-Cajigas
Andreozzi, P.C.; R. Griffiths, D. Moverley, J. Wainiqolo, R. Nias, S. Boudjelas, D. Stewart, S. Cranwell, M. Smith and P. Cowan. The Pacific invasives partnership – a model for regional
collaboration on invasive alien species
The Pacific invasives partnership – a model for regional collaboration
on invasive alien species
P.C. Andreozzi, R. Griffiths, D. Moverley, J. Wainiqolo, R. Nias, S. Boudjelas, D. Stewart, S. Cranwell, M. Smith
and P. Cowan
U.S. Department of Agriculture, Animal and Plant Health Inspection Service, Washington DC,
USA. <Phillip.C.Andreozzi@aphis.usda.gov>
Invasive Alien Species (IAS) are a fundamental challenge facing Pacific Island Countries and Territories (PICTS),
impacting economies, habitats, food security, biodiversity, livelihoods and quality of life. These negative and substantial
impacts are being acknowledged by PICTs leaders as well as on the international stage. As the inter-relatedness of IAS
and other fundamental challenges such as climate resilience, oceans and sustainability are understood and acknowledged,
strategies to integrate IAS and biosecurity concepts into international efforts will require invasive species expertise and
guidance. The Pacific Invasives Partnership (PIP) is a group created by the Pacific Roundtable for the Conservation of
Nature that has evolved into a broad advocate for IAS outreach and an incubator for collaborative IAS efforts in the
Pacific. PIP comprises volunteer IAS experts from regional, national, NGO and international groups that work in two or
more PICTS and want to advance IAS issues. By taking a “rising tide floats all boats” approach, PIP members work to
raise the profile and understanding of IAS as a fundamental, underpinning issue to PICT economies, environments and
future sustainability. PIP successes over the past five years include reports and briefing materials prepared for the Pacific
Islands Forum Leaders meeting, provision of advice and assistance for Pacific invasive species Global Environment
Facility projects, leading and supporting regional and sub-regional projects on regional biosecurity, invasive ant and
rodent eradication and prevention, and the successful raising of the IAS profile at various international fora. PIP is a
successful model of regional collaboration on invasive alien species and could be used as a model for similar efforts in
other island regions of the world.
P.C. Andreozzi, R. Griffiths, D. Moverley, J. Wainiqolo, R. Nias, S. Boudjelas, D. Stewart, S. Cranwell, M. Smith
and P. Cowan
704
Bird, J.; J. Shaw, R. Alderman and R. Fuller. A review of monitoring of biodiversity responses to island invasive species eradications
A review of monitoring of biodiversity responses to island invasive
species eradications
J. Bird, J. Shaw, R. Alderman and R. Fuller
J. Bird, J. Shaw, R. Alderman and R. Fuller
Centre for Biodiversity & Conservation Science, University of Queensland, School of Biological Sciences,
University of Queensland, St Lucia, Queensland 4072 Australia. <jez.bird@uq.edu.au>
A recent review examined the benefits from invasive species eradications on islands worldwide. While the review
concluded that island eradications are overwhelmingly beneficial for native biodiversity, a response to eradication was
only demonstrated for 22 of the 532 islands treated. While many studies advocate monitoring, there appears to be a gap,
either between eradication effort and monitoring effort, or between monitoring and analysing/reporting responses. We
focussed on regions of the Pacific, Australia and the Caribbean to document the level of monitoring on islands where
eradications have taken place. We collated published and unpublished literature and spoke to key practitioners in the region
to investigate targets for monitoring, duration and frequency of monitoring, and the ability of implemented monitoring
work to detect responses. We also investigated drivers of monitoring such as type of funder or implementing organisation
behind the eradication operation. The study’s findings highlight apparent biases in monitoring effort, they provide a
benchmark of current monitoring effort, and open the debate on when and where monitoring should be undertaken and
how best to develop optimal monitoring strategies.
Booker, H.; D. Appleton, D. Bullock, R. MacDonald, E. Bell, D. Price, P. Slader, T. Frayling, A. Taylor and S. Havery. A review of seabird recovery on Lundy Island, England, over a decade
following the eradication of brown and black rats
A review of seabird recovery on Lundy Island, England, over a decade
following the eradication of brown and black rats
H. Booker, D. Appleton, D. Bullock, R. MacDonald, E. Bell, D. Price, P. Slader, T. Frayling, A. Taylor and S. Havery
Species and Habitats, Royal Society for the Protection of Birds, RSPB, Headquarters, the Lodge, Sandy,
Bedfordshire SG19 2DL, UK. <Helen.Booker@rspb.org.uk>
Lundy, a 450 ha island situated 19 km off the Devon coast in the UK’s Bristol Channel, is internationally important
for its marine life and its waters were established as the UK’s first Marine Nature Reserve in 1971. Lundy is home to
eleven seabird species, including Manx shearwater (Puffinus puffinus), for which the UK has a global responsibility
and Atlantic puffin (Fratercula arctica), a globally threatened species. Steep declines in Lundy’s seabird populations,
with puffins nearing extinction and low numbers of Manx shearwaters, led to the establishment of the Seabird Recovery
Project in 2001. The project aimed to improve the conditions for these burrow-nesting seabirds through the eradication
of brown and black rats. From 2002–2004 a ground-based operation was undertaken, and in 2006 Lundy was officially
declared rat-free. The seabird populations of Lundy have been well studied with detailed regular data spanning the last 35
years. Over the last decade, as a result of rat removal, seabird numbers on the island have doubled and storm petrels have
colonised. By 2013, the breeding population of Manx shearwaters increased more than ten-fold to an estimated 3,451
pairs. In 2004, the puffin population on Lundy fell to an all-time low with only five individuals, but in 2013, more than 80
individuals were recorded. Here we discuss the observed seabird responses to the eradication and present the most recent
results of the monitoring surveys from 2017. These impressive results highlight the importance of and need for effective
biosecurity to reduce the risk of re-incursion of rats. Lundy is a popular tourist destination with a working farm; therefore,
the regular transportation of cargo remains a high biosecurity risk. A revised biosecurity and incursion response plan is
now being finalised.
H. Booker, D. Appleton, D. Bullock, R. MacDonald, E. Bell, D. Price, P. Slader, T. Frayling, A. Taylor and S. Havery
Boser, C.L. Eradicating invasive ants in conservation areas
C.L. Boser
Eradicating invasive ants in conservation areas
C.L. Boser
The Nature Conservancy, 532 E Main St, Suite 200 Ventura, California 93001, USA. <cboser@tnc.org>
Established invasive invertebrates, such as Argentine ants (Linepithema humile), can have long-term and cascading adverse
ecological impacts for native communities. In Mediterranean ecosystems, they out-compete most native ant species and
harm plants such that they interfere with pollination, reducing seed set. In 2013–2016 we developed and carried out a
unique treatment protocol on four Argentine ant infestations on Santa Cruz Island, California, totalling 410 ha. We used
polyacrylamide beads, hydrated with 6 ppm thiamethoxam and 25% sucrose water distributed at a rate of 148 litres per
hectare via helicopter and hopper. We treated the four infestation areas 14 times, for total cost of US$1,400 per ha. Two
monitoring strategies used lures and visual searching on 74 ha in 2013–2015, with costs at US$2,200 and US$500 per
ha. The less costly, targeted strategy revealed one spot population totalling 0.3 ha. This population was located at the edge
of a treatment site, possibly indicting that the 50 m buffer added to that delimited infestation was insufficient. Follow up
treatments were conducted on that site and Argentine ants were not detected in subsequent monitoring rounds. Monitoring
will continue 2016–2020 throughout all four treatment areas, aided by a fine-scale model of probability of detection and
probability of persistence by vegetation type, and detection dogs. Packaged with patience and persistence, these treatment
and monitoring protocols show promise as an eradication tool. Preliminary data indicate that the treatment may also be
effective in eradication programmes for other invasive ant species.
705
Brazier, M. Big island, small invader: eradicating invasive fish on a national scale
Big island, small invader: eradicating invasive fish on a national scale
M. Brazier
M. Brazier
National Fisheries, Environment Agency, Bridge End Depot, Causeway Road, Kendal, Cumbria,
North West, UK. <matt.brazier@environment-agency.gov.uk>
Topmouth gudgeon (Pseudorasbora parva) is an invasive non-native cyprinid from Asia. Listed as a Species of Union
Concern under the EU Invasive Alien Species Regulations, it is considered one of the most potentially damaging non-native
fish species to invade Western Europe. Introduced to Great Britain (GB) in 1984, evidence indicated that if topmouth
gudgeon established in GB, the impacts on our native species and habitats could be severe. The threats were clear, and
the case for action robust. However, in 1980s and 1990s GB authorities lacked a coherent invasive species strategy,
regulatory powers were ineffective, there was no focused expertise or capacity and the tools and techniques necessary
to control such a tenacious invasive species had not been developed or adopted. Topmouth gudgeon spread inexorably
across England and Wales, until 2004. By 2004, with seven populations identified, the authorities were no closer to a
solution. However, using an innovative biocide-based approach, a local Environment Agency team successfully eradicated
topmouth gudgeon from a fishery in the Lake District. This led to a number of small scale, ad hoc eradications, but as
confirmed populations climbed to 14, sustainable removal of the species from GB was not considered feasible. In 2011,
supported by the GB Invasive Species Strategy, the Environment Agency utilised their growing expertise and capacity
to develop a specialist team and equipment and implemented a Water Framework Directive National Programme; their
ambitious objective: total eradication of topmouth gudgeon from GB by 2018. Scaling up from small scale, localised
eradication to a national landscape scale programme to eradicate an aquatic invasive species was unprecedented and
presented significant strategic, legal, operational, economic and political challenges. This paper documents that 12-year
journey, highlighting the challenges, discussing how they were overcome, the lessons learnt, and considers the future
potential and direction of this work.
Population growth of seabirds after the eradication of
introduced mammals
R. Buxton and M. Brooke
R. Buxton and M. Brooke
Department of Fish, Wildlife, and Conservation Biology, Colorado State University, 1474 Campus Delivery,
Fort Collins, CO 80523, USA. <rachel.buxton@colostate.edu>
Eradication of introduced mammals to restore island ecosystems has become increasingly common, with more than
1,000 successful projects around the world. Various benefits for native fauna have been documented, including reduced
predation and positive demographic response. However, evidence that these eradications lead to increases in populations
of seabirds, which are important island ecosystem engineers, is sparse. The limited amount of monitoring is partly
because of seabirds’ long life cycle, meaning that several years or even decades may elapse before populations respond
to eradication. Drawing on data from across the world, we assemble population growth rates (lambda, λ) of 181 seabird
populations of 69 species following successful eradication projects. After successful eradication, the median growth rate
was 1.12 and populations with positive growth (λ > 1; n = 151) greatly outnumbered those in decline (λ < 1; n = 23) and
those that exhibited no change (λ = 1; n =7). Population growth was faster at newly-established colonies compared to
those already established, and in the first few years after eradication before the species’ age of first breeding. Because λ
was higher before first-time breeders are recruiting back into the colony, this suggests that immigration is important for
colony growth. Population growth was also faster among gulls and terns compared to other seabird groups and when
several invasive mammals were eradicated together in the course of the restoration project. This reflects the relative
lack of philopatry among gulls and terns and reinforces current best practice – the removal of all invasive mammals
where feasible. These results may help prioritise sites for future eradication projects and determine where active seabird
population management is required after eradication.
Assessment of the possible effects of biological control agents of
Lantana camara and Chromolaena odorata in Davao City,
C. Canlas, C. Gever, P. Rosialda, Ma. N. Quibod,
Mindanao, Philippines
P. Buenavente, N. Barbecho, C. Layusa and M. Day
C. Canlas, C. Gever, P. Rosialda, Ma. N. Quibod, P. Buenavente, N. Barbecho, C. Layusa and M. Day
Biology Department, Adamson University, 9666 Dona Cipriana Street, Gat-Mendoza, Brgy. Vitalez.
<canlascristine@ymail.com>
Invasive plants have an impact on global biodiversity and ecosystem function, and their management is a complex
and formidable task. Two of these invasive plant species, Lantana camara and Chromolaena odorata, are found in the
Philippines. Lantana camara has the ability to suppress the growth of and outcompete neighbouring plants. Chromolaena
odorata causes serious agricultural and economical damage and causes fire hazards during dry season. In addition, both
species have been reported to poison livestock. One of the known global management strategies to control invasive plants
is the introduction of biological control agents. These natural enemies of the invasive plants reduce population density and
impacts of the invasive plants, resulting in the balance of the nature in their invasion. Through secondary data sources,
interviews, and field validation (e.g. microhabitat searches, sweep netting, opportunistic sampling, photo-documentation),
we investigated whether the biocontrol agents previously released by the Philippine Coconut Authority (PCA) in
their Davao Research Center to control these invasive plants are still present and are affecting their respective host
weeds. We confirm the presence of the biocontrol agent of L. camara, Uroplata girardi, which was introduced in 1985,
and Cecidochares connexa, a biocontrol agent of C. odorata released in 2003. Four other biocontrol agents were found
to affect L. camara. Signs of damage (e.g. stem galls in C. odorata, and leaf mines in L. camara) signify that these
biocontrol agents have successfully established outside of their release site in Davao. Further investigating the extent
of the spread of these biocontrol agents in the Philippines and their damage to the two weeds will contribute to the
management of invasive plant species in the country.
706
Buxton, R. and M. Brooke. Population growth of seabirds after the eradication of introduced mammals
Canlas, C.; C. Gever, P. Rosialda, Ma. N. Quibod, P. Buenavente, N. Barbecho, C. Layusa and M. Day. Assessment of the possible effects of biological control agents of Lantana
camara and Chromolaena odorata in Davao City, Mindanao, Philippines
Cecchetti, M.; G. Dell’Omo and B. Massa. Black rat eradication from Linosa Island: work in progress
Black rat eradication from Linosa Island: work in progress
M. Cecchetti, G. Dell’Omo and B. Massa
M. Cecchetti, G. Dell’Omo and B. Massa
University of Exeter, Environment and Sustainability Institute, College of Life and Environmental Science,
Penryn Campus, Cornwall UK; and Berta maris, Linosa. <mc703@exeter.ac.uk>
The black rat (Rattus rattus) is one of the most successful colonising mammals and one of the 100 world’s worst invasive
species. It is a generalist and opportunistic predator, particularly of seabird eggs and chicks on islands where it has
been transported by humans on ships. The Life project “Pelagic Birds: Conservation of the main European population
of Calonectris diomedea and other pelagic birds on Pelagic Islands” on Linosa Island involves the eradication of black
rats, since it is considered the major cause of Scopoli’s shearwater breeding failure. From 15 May to 10 October 2013,
a preliminary phase was carried out to determine the abundance and the distribution of black rats and house mice
(Mus musculus) through captures. In four sessions of captures in eight different representative habitats, a total of 197
rats and 247 mice have been captured. In the same year rats impacted negatively the 34% of the 400 shearwater nests
monitored, having a similar impact on eggs and chicks. On February 2016 we set 2,700 rodenticide stations all around
the island. Then, the rodenticide was replaced in April, June, October and November, with positive results. The rat take
of baits has decreased significantly. In November, an average of 86% of baits were left in the stations, indicating a strong
decrease of the rat population. Continuing the action and the distribution of rodenticide is essential in order to reach the
eradication of this aggressive predator by the end of the year.
Cecchetti, M.; L. Nelli, B. Massa and G. Dell’Omo. Effects of cat, rat, and human predation on Scopoli’s shearwater (Calonectris diomedea) breeding success and nest-site occupancy on
Linosa Island
Effects of cat, rat, and human predation on Scopoli’s shearwater
(Calonectris diomedea) breeding success and nest-site occupancy
on Linosa Island
M. Cecchetti, L. Nelli, B. Massa and G. Dell’Omo
M. Cecchetti, L. Nelli, B. Massa and G. Dell’Omo
University of Exeter, Environment and Sustainability Institute, College of Life and Environmental Science,
Penryn Campus, Cornwall, UK; and Berta maris, Linosa. <mc703@exeter.ac.uk>
Biodiversity on islands is seriously threatened by invasive species, that have been voluntarily or accidentally introduced
by humans. Seabirds, especially small and medium ground-nesting Procellariiformes, are particularly vulnerable to
introduced predators which can negatively affect breeding success and nest occupancy. Linosa is a small Mediterranean
island where thousands of Scopoli’s shearwaters (Calonectris diomedea) breed each year. Their survival is endangered
by the presence of 400 inhabitants, 300 free-roaming cats (Felis catus), and a conspicuous population of rats (Rattus
rattus). Our study aims at evaluating the effects of cat, rat and human predation on the shearwaters’ breeding success and
the effects of breeding failure on nest-site occupancy. From 2013 to 2016 we monitored shearwater nests and collected
data on burrow occupancy, egg deposition, egg hatching, and chick fledging taking notes of cases of failure. Nest
characteristics were also measured. Overall, the shearwater breeding success was 65% and predation by mammals was
the major cause of breeding failure (19%). We analysed the effects of cat and rat predation and poaching on the nest
occupancy in the following year, using generalised linear mixed effect models. We also analysed if nest characteristics
(depth and diameter) and nest position, in terms of distance from houses, roads, trails and coastline, were related to the
probability of predation by cats, rats and poaching. Egg-poaching had a negative effect on the occupancy of the following
year, whereas predation upon eggs by rats and predation upon chicks by cats had a minor effect. We also found that the
nest position didn’t affect the probability of predation by rats and cats and egg poaching. However, increasing in cavity
depth reduces the probability of cat predation.
Clubbe, C. Invasive plants: what can be done about this continuing threat to biodiversity?
C. Clubbe
Invasive plants: what can be done about this continuing
threat to biodiversity?
C. Clubbe
Conservation Science, Royal Botanic Gardens Kew, Richmond, Surrey TW9 3AE, UK. <c.clubbe@kew.org>
Human intervention has led to plants being moved around the world for centuries. This practice has been both
unintentional and intentional. Unintentional when seeds and/or vegetative propagules were transported vicariously
alongside other materials being moved. Intentional when desirable and useful plants were moved around the world, often
linked with colonialisation, arguably the fore-runner of today’s globalisation. Many of these plants became naturalised
only locally or required careful nurturing to survive in their new habitats. However, some of these plant species found
their new environments highly conducive to spread. Removed from controlling factors such as pests and herbivores,
they became established over significant areas posing a serious threat to native biodiversity. Invasive species are now
recognised as a major driver of biodiversity loss globally, with particularly severe impacts on islands. We have reviewed
six global invasive species databases to determine the number of invasive plants globally. Taxonomic reconciliation
has demonstrated that 6,075 vascular plant species are currently documented as invasive. The first part of this talk will
review this in its historical context and consider the implications of the continuing increase in the number and spread of
invasive plant species globally. The second part of the talk will review work by Kew’s UK Overseas Territories team on
invasive plants. The UK Overseas Territories support the most significant UK biodiversity in terms of unique species and
habitats. This biodiversity is under severe threat from invasive species. We have been identifying and mapping invasive
plants, and developing actions plans for their control. The talk will include examples from St Helena, Ascension, Falkland
Islands and British Virgin Islands. Wider implications from this work for dealing with this global threat will be considered.
707
Cranwell, S. Partnerships in the restoration of tropical Pacific islands
Partnerships in the restoration of tropical Pacific islands
S. Cranwell
S. Cranwell
Birdlife International, GPO Box 18332, Suva, Fiji. <steve.cranwell@birdlife.org>
The eradication of invasive alien species from islands is a highly effective conservation action for the recovery of
declining and threatened native species. Among the characteristics necessary for the success of these operations and
the sustainability of the conservation outcomes is a range of technical expertise, cultural and political support, and
financial and organisational capacity. In the tropical Pacific, civil society organisations including the BirdLife Partnership
have taken a lead in implementing invasive vertebrate eradications, and despite capacity limitations have successfully
delivered operations for 40 sites in five countries since 2007. The scale and complexity of these eradications have
increased over time, from focusing on single target species on individual islands to simultaneously addressing multiple
invasives and islands. This growing experience has highlighted the strengths of locally based civil society organisations,
particularly in addressing the cultural and political issues associated with vertebrate eradications, but also the essential
role of partnerships in supporting their technical preparation and financing. The operations to date have benefited multiple
threatened species. However, if invasive species management is to fulfil it’s potential to reduce biodiversity loss on Pacific
islands, political support and local capacity must increase, particularly for biosecurity. Stronger partnerships between
governments and non-governmental organisations are also necessary, both to engage local communities and to meet the
specialised technical preparations and significant financing needs, so that the challenges of island restoration are met with
a response of the requisite pace and scale.
del Mar Leza, M.; A. Marqués, C. Herrera, M. Ángel Miranda, M. Ruiz, A. Pou and C. Guerrero. Vespapp: citizen science to detect the invasive species Vespa velutina.
Vespapp: citizen science to detect the invasive species Vespa velutina
M. del Mar Leza, A. Marqués, C. Herrera, M. Ángel Miranda, M. Ruiz, A. Pou and C. Guerrero
University of Balearic Island, Palma of Majorca, Balearic Island, Spain. <mar.leza@uib.es>
The yellow-legged hornet (Vespa velutina) is an Asian native species recorded across Europe, including mainland Spain,
since 2004. Its first detection in Majorca (Balearic Islands; Spain) took place by researchers at the Laboratory of Zoology
in collaboration with local beekeepers in October 2015. This invasive species has an important impact on biodiversity,
apiculture and human health. Adult wasps are predators of bees, therefore contributing to the loss of honeybee colonies. For
efficient actions to minimise the harms of the invasive species, early detections are crucial. Thus, civic collaboration may
offer an important source of information to determine the presence and distribution of V. velutina. Current technological
advances offer the opportunity for citizens to become active participants of the scientific research (citizen science). Vespapp
is a software, either as a cell phone app or a website, which aims to identify any suspicious observation (hornets and
nests) by sending a picture to a global database. The received information is subsequently confirmed or discarded by an
expert panel. In case of a positive identification, an action protocol is implemented including the placement of traps, nest
removal and monitoring the area. Since the Vespapp launch in June of 2016, the app has been downloaded 1436 times, has
received more than 450 photos and 31 of them have been positive in the Balearic Island and the Iberian Peninsula. These
results have enabled detection and removal of a total of nine nests during 2016, which is of great importance in controlling
the expansion of the V. velutina considering the early stage of invasion in the Balearic Islands.
M. del Mar Leza, A. Marqués, C. Herrera, M. Ángel Miranda, M. Ruiz, A. Pou and C. Guerrero
Djeddour, D.; N. Maczey and C. Pratt. Wild ginger, a beautiful menace to island ecosystems – can a natural solution be found?
Wild ginger, a beautiful menace to island ecosystems – can a natural
solution be found?
D. Djeddour, N. Maczey and C. Pratt
D. Djeddour, N. Maczey and C. Pratt
CABI Bakeham Lane, Egham, Surrey TW20 9TY, UK. <d.djeddour@cabi.org>
Kahili or wild ginger, (Hedychium gardnerianum (Zingiberaceae)) poses a serious threat to many unique island ecosystems
worldwide including: the Federated States of Micronesia, Cook Islands, French Polynesia, Hawaii, New Zealand, La
Réunion, the Macaronesian Archipelago and Jamaica. Introduced from the foothills of the Himalayas for its ornamental/
commercial value, kahili has escaped cultivation and to become an aggressive coloniser in its introduced, sub-tropical
range. Adaptable to a wide range of habitats, from native wetlands and riparian areas through to forest understorey,
road verges and scrubland, wild ginger forms large, herbaceous, shade tolerant monocultures which outcompete native
vegetation. It has the potential to prevent regeneration of native forests and cause wide scale ecosystem collapse and
biodiversity loss. Wild ginger forms deep rhizome beds, reproduces vegetatively as well as through seed and is spreading
unchecked across extensive and rugged terrain, which make chemical and mechanical control largely ineffectual. Classical
biological control is widely believed to be the only long-term solution for this intractable invader. A biocontrol initiative
for kahili ginger was initiated by CABI in 2008 for Hawaiian and New Zealand stakeholders. Surveys in the native
range identified a number of damaging and limiting natural enemies which continue to be evaluated for specificity in the
UK. The progress, prioritised agents and future prospects are further described.
708
Doube, J. Is poisoning rodents a health hazard?
Is poisoning rodents a health hazard?
J. Doube
J. Doube
SAAS-MedSTAR, SA Health (and Australian Antarctic Division), 5 Taylors Lane,
Strathalbyn, SA 5255, Australia. <james.doube@sa.gov.au>
As large-scale island eradication projects expand, it is inevitable that aerial baiting will occur on inhabited islands. However,
when anticoagulant bait is to be spread all around living areas, community concerns about the safety of such projects
are likely and understandable. Health monitoring of bait handling personnel on the largest island aerial baiting projects
(including Macquarie Island and South Georgia), has shown no significant poisoning. Given the exposure of these
individuals is orders of magnitude beyond that of community members, such monitoring can provide reassurance to far
less exposed individuals. Additionally, lessons can be learnt on how to manage the community perceptions of these issues
for critical conservation projects.
Fleischmann, K.; S. Massy, M. Schmutz, B. Seraphine and J. Millett. When our enemy is our friend: new approaches to managing alien vegetation in Seychelles catchment
forest
When our enemy is our friend: new approaches to managing alien
vegetation in Seychelles catchment forest
K. Fleischmann, S. Massy, M. Schmutz,
B. Seraphine and J. Millett
K. Fleischmann, S. Massy, M. Schmutz, B. Seraphine and J. Millett
University of Seychelles, P.O. Box 1348, Anse Royale, Mahé, Seychelles. <kfleisch@bluewin.ch>
Invasive alien plants are one of the major causes of biodiversity loss with impacts on ecosystems such as alterations
of biogeochemical and hydrological cycles. The Seychelles’ forest is dominated by non-native vegetation arising from
plantation agriculture often referred to as novel ecosystems. Under some circumstances native vegetation shows signs of
recovery, particularly in low light conditions that occur under a forest canopy dominated by exotic species. Conversely,
high light conditions arising from forest disturbance benefit invasive exotic species especially vines such as Merremia
peltata which outcompete native vegetation. The Ecosystem Based Adaptation in the Seychelles project aims to
enhance water-catchment management formulating recommendations for vegetation rehabilitation and establishing
post-rehabilitation monitoring. The project will rehabilitate 600 ha of forest, an ambitious target that requires forestry
management capacity development, policy development and community support to ensure long-term protection and
management of catchment forest. Catchment vegetation quality was assessed using plant endemism, species diversity
and forest rejuvenation indices. Sampling was conducted by transects and permanent monitoring plots in 10 intensive
monitoring sites in water catchments. The project also deployed drone monitoring and light level monitoring using
images taken with a fish eye lens. Rehabilitation has been implemented first on sites with high vegetation quality indices
where management is expected to assist natural regeneration. Management has focused on removal of exotic saplings
and under-canopy shrubs leaving a forest canopy dominated by exotic species including Tabebuia pallida and Falcataria
moluccana intact. This counterintuitive approach is expected to maintain the shade conditions and the microclimate that
will benefit native species over non-native species and facilitate the regeneration of palm dominated native forest. Initial
indications are that closed canopy forest rehabilitation and community supported protection of forests from disturbance
are important management measures for these novel ecosystems and hence for water catchments.
Fric, J.; T. Dimalexis, V. Goritsas, A. Evangelidis and I. Nikolaou. Eleonora’s falcon (Falco eleonorae) benefiting from rat eradication – the case of Andros, Greece
Eleonora’s falcon (Falco eleonorae) benefiting from cat eradication
– the case of Andros, Greece
J. Fric, T. Dimalexis, V. Goritsas,
A. Evangelidis and I. Nikolaou
J. Fric, T. Dimalexis, V. Goritsas, A. Evangelidis and I. Nikolaou
Nature Conservation Consultants Ltd., Chalandri, GR-15231 Greece. <jakobfric@n2c.gr>
Three colonies of Eleonora’s falcon (Falco eleonorae) on the island of Andros (Cyclades, Greece) have been monitored
since 2006. On two of these colonies, presence of rats has been recorded at least since 2006, while the third colony was
invaded by rats in 2011. The latter provided a unique situation to study the short-term impacts of rats on the breeding
performance of the Eleonora’s falcon. On the newly rat-infested islet within a single year the number of active nests, the
breeding success and the total number of fledglings were reduced by 47%, 23% and 58% respectively. Rat eradication
operations were successfully carried out on all three islets in 2012 and 2014. At all colonies the breeding performance
improved immediately. At the colony where rats were present for only one breeding season (2011), all breeding parameters
recovered to pre-invasion levels within the rat eradication year. In all colonies, vegetation degradation resulting from
rat foraging had consequently led to lower nesting site quality for falcons. Therefore, rat eradications were followed
by construction of artificial nests which further improved the breeding habitat. In the years following the eradications
14–25% of active nests were artificial and the breeding success in artificial nests was in general higher than in natural
nests. The rat eradication operations in combination with the construction of artificial nests on the islets of Andros indicate
the benefits of these management measures on the breeding performance of the Eleonora’s falcon and highlight the
importance of immediate response to rat infestation. The conservation measures were implemented as part of the LIFE
Nature project ANDROSSPA (LIFE10 NAT/GR/000637).
709
Fric, J. and A. Evangelidis. A review of 12 years of rat eradication operations for the conservation of priority island nesting birds in Greece
A review of 12 years of rat eradication operations for the conservation
of priority island nesting birds in Greece
J. Fric and A. Evangelidis
J. Fric and A. Evangelidis
Nature Conservation Consultants Ltd., Chalandri, GR-15231, Greece. <jakobfric@n2c.gr>
Since 2005, rat eradication operations have been carried out on a total of 41 uninhabited islets and islands in the Aegean
Sea in Greece ranging in size from less than 1 ha up to almost 300 ha with the total area exceeding 1,050 ha. The initial
eradication methodology was developed with the support of the Royal Society for the Protection of Birds and further
optimised through implementation of consecutive eradication operations. The operations were carried out on 16 different
groups of islets and islands with the aim of improving the breeding habitat of a significant proportion of island nesting
bird species of conservation concern, including Eleonora’s falcon (Falco eleonorae), Mediterranean shag (Phalacrocorax
aristotelis desmarestii), Audouin’s gull (Larus audouinii), yelkouan shearwater (Puffinus yelkouan), Scopoli’s shearwater
(Calonectris diomedea diomedea) and European storm-petrel (Hydrobates pelagicus). While the most recent eradication
operations are still underway, previous operations have successfully removed all rats, eliminating egg and chick predation,
as well as, degradation of bird nesting habitats. All rat eradication operations were carried out using brodifacoum-based
bait, deployed mainly through placement of bait stations in association with hand broadcast. No significant negative
impacts on non-target species due to baiting have been recorded. All rat eradication operations have been carried out
through six different LIFE projects co-financed by the European Commission.
Fric, J.; A. Evangelidis, T. Dimalexis, N. Tsiopelas, S. Xirouchakis, C. Kassara and S. Giokas. Improving nesting habitats for the Eleonora’s falcon and seabirds
Improving nesting habitats for the Eleonora’s falcon and seabirds
J. Fric, A. Evangelidis, T. Dimalexis, N. Tsiopelas, S. Xirouchakis, C. Kassara and S. Giokas
Nature Conservation Consultants Ltd., Chalandri, GR-15231 Greece. <jakobfric@n2c.gr>
Rat invasion is considered a major environmental issue regarding the Aegean islands (Greece), which are characterised by
a rich biodiversity of faunistic and floristic taxa of high conservation concern. One of the most emblematic bird species
at national level, Eleonora’s falcon (Falco eleonorae), is severely affected by rat invasion. The Aegean islands constitute
the core of its breeding range, holding more than 80% of the species’ breeding population. In the framework of the LIFE
Nature project “LIFE ElClimA” (LIFE13 NAT/GR/000909), rat eradication operations take place at two uninhabited
island complexes, hosting approximately 6% of the species’ national population, as well as important colonies of other
priority seabird species that are also affected by rat predation, namely the yelkouan shearwater (Puffinus yelkouan) and
Scopoli’s shearwater (Calonecrtis diomedea). Removing rats from a total area of 705 ha is the largest rat eradication
operation ever attempted in the country. Rodenticide baits have been primarily deployed in bait stations to minimise
primary poisoning risk to non-target species, e.g. partridges (Alectoris chukar) and rabbits (Oryctolagus cuniculus), as
well as their predators such as Bonelli’s eagles (Aquila fasciata) and long-legged buzzards (Buteo rufinus), which could
be deprived of their food source. After several months of regular baiting, bait consumption is minimal and the eradication
operations are considered to be at their final stage. Close cooperation with regional and local stakeholders throughout
the field operations aims to ensure optimal involvement of local communities and authorities as well as minimal risk of
future rat reinvasion. The rat eradication operations implemented in the framework of the current project are expected to
contribute to the preservation of the high ecological value of the two island complexes in general and, in particular, to the
improvement of the nesting habitat and conservation status of important bird species in the area.
J. Fric, A. Evangelidis, T. Dimalexis, N. Tsiopelas, S. Xirouchakis, C. Kassara and S. Giokas
Genovesi, P. Broadening the context of invasive species eradications
Broadening the context of invasive species eradications
P. Genovesi
P. Genovesi
ISPRA and IUCN SSC ISSG, Via V. Brancati 48, Rome, ROMA 00144, Italy. <piero.genovesi@isprambiente.it>
Our ability to eradicate harmful organisms has greatly progressed in the last decades, and available information shows
that currently this management tool not only is one of the most effective conservation weapons, but also that it permits
protection of human livelihood. In conservation we tend to refer to eradications as the total and permanent removal of an
invasive species’ population by means of a time-limited campaign; this term is more often used for eradications carried
out on islands, where some general rules apply, such as that all individuals need to be vulnerable to the removal methods,
and that there should be no risk of reinvasions. However, there is a growing number of interventions that go beyond this
definition. Eradications can now target multiple species, and campaigns carried out in densely inhabited regions need to
address significant risks of reinvasions through long term surveillance and rapid response efforts. Furthermore, there have
been eradications carried out at much larger scales than small islands, such as those implemented for human or animal
health purposes (e.g. smallpox or the rinderpest virus eradicated from the globe), or of eradications in mainland areas,
requiring complex geographical planning, and that may set context specific objectives such as management to zero density
in key areas through permanent control efforts. To fully exploit the potential of invasive species control for conservation,
it is important to adapt the lessons learnt in islands eradications, rethinking the paradigms of this conservation tool to the
new challenges that need to be met, as also highlighted by the Honolulu Challenge on Invasive Alien Species adopted in
2016. The New Zealand Predator Free 2050 campaign, planning to eradicate several key invasives at an unprecedented
scale, is indeed a milestone in this direction, providing a basis for broadening the global vision of invasive species
management.
710
Geraldes, P.; T. Melo, P. Oliveira and V. Paiva. Recovery of Santa Luzia Nature Reserve and translocation of the globally endangered Raso lark
Recovery of Santa Luzia Nature Reserve and translocation of the
globally endangered Raso lark
P. Geraldes, T. Melo, P. Oliveira and V. Paiva
P. Geraldes, T. Melo, P. Oliveira and V. Paiva
Marine Program, Sociedade Portuguesa Estudo das Aves, Av Columbano Bordalo Pinheiro, 87, 3º,
Lisbon, 1070-062 Portugal. <pedro.geraldes@spea.pt>
The inhabited island of Santa Luzia is a priority KBA located in Cabo Verde. It holds the entire world population of the
Critically Endangered Raso lark (Alauda razae) and the most important colony of Cabo Verde shearwater (Calonectris
edwardsii). Since 2013 SPEA, Biosfera1 and RSPB have developed a feasibility study for habitat recovery of Santa Luzia,
including an operational plan for cat (Felis catus) eradication and several baseline studies on the local species, both native
and alien. The current project aims to translocate part of the population of Raso larks to the nearby island to increase
the resilience of the population to long periods of droughts that have been increasing with global climatic changes. The
feral cat population, estimated at 126 animals (95% CI 87.5 – 189) individuals, has strong negative impacts on several
species of fauna on the island and will have to be removed to increase the chances of success of the re-introduction of
Raso larks. Mice (Mus musculus) are also present, but at very low densities. Abundance index was calculated throughout
the year and peaked at 0.06 and 0.067 captures/trap/night in February and March respectively (mean abundance index
throughout the year 0.026). Recent data on the cat diet shows high levels of reptile predation and was found to change
markedly depending on annual conditions. In 2010 mice were 79.6% of prey species identified in cat diet, while in 2013
and 2014 cats preyed mostly upon reptiles (91.67% of scats and >70% of prey item biomass). The project will rely
strongly on local staff and will involve local communities in order to build local capacity and to increase awareness of the
problems caused by IAS on islands. We aim to achieve sustainable protection of the habitats and threatened biodiversity
of Raso, Branco and Santa Luzia marine protected areas.
Setting-up a predator-free area on a Macaronesian island using a
pest-proof fence
P. Geraldes, T. Pipa, N. Oliveira, C. Silva and S. Hervías
P. Geraldes, T. Pipa, N. Oliveira, C. Silva and S. Hervías
Marine Program, Sociedade Portuguesa Estudo das Aves, Av Columbano Bordalo Pinheiro, 87, 3º,
Lisbon, 1070-062 Portugal. <pedro.geraldes@spea.pt>
The island of Corvo, with an area of 17.1 km2, is the smallest, westernmost and least populated of the nine islands of
the Azorean Archipelago; 41% of the island is classified as a Special Area for Conservation and Special Protection Area
thus included in the Natura 2000 Network and classified as UNESCO’s Biosphere Reserve. Azorean settlers brought a
number of associated threats to the local fauna and flora, such as the introduction of invasive mammals (rats (Rattus spp.),
mice (Mus musculus), cats (Felis catus), goats (Capra hircus) and sheep (Ovis aries)), which jeopardize the breeding
populations of seabirds. The archipelago still remains of critical importance for the conservation of several petrel species,
namely Cory’s shearwater (Calonectris borealis), little shearwater (Puffinus lherminieri) and Madeiran storm-petrel
(Hydrobates castro). From 2009 to 2012, with funds from the EU LIFE program, a 100% pest proof fence 800 m long was
built on Corvo, Azores. This solution was adopted to create a safe nesting area of 3 ha for shearwaters and petrels breeding
in the island and subject to high predator pressure from feral cats, dogs (Canis familiaris), black-rats (Rattus rattus)
and mice. Following the closure of the area, all predators were removed and biosecurity procedures were adopted. The
vegetation cover inside the fenced area was cleared of alien plants and native flora was abundantly replanted to recover the
natural habitats. Acoustic and visual luring methods for prospecting seabirds were employed, and for three consecutive
years small groups of juveniles Cory’s shearwaters were translocated to the area. The fence withstood hurricane type
winds and very frequent harsh weather conditions for long periods of time with minor maintenance necessary. After four
years the fence demonstrated to be a feasible solution for adequate areas and the first breeding pairs of seabirds were
recorded inside the area during the 2016 breeding season.
Green iguana (Iguana iguana) monitoring and control efforts on
Grand Cayman
J. Haakonsson, F. Rivera-Milan and E. Radford
J. Haakonsson, F. Rivera-Milan and E. Radford
Cayman Islands Department of the Environment, Terrestrial Resources Unit, PO Box 10202,
580 North Sound Road, Grand Cayman KY1-1002,Cayman Islands. <jane.haakonsson@gov.ky>
Effective control management of invasive alien species (IAS) is limited by our understanding of population
dynamics. Monitoring and modelling are essential components of control management. The green iguana (Iguana
iguana) is overabundant on Grand Cayman (estimated density ± SE = 41.363 ± 16.813), and this can cause significant
economic losses (e.g., damage to roads and agricultural crops), pose serious health and safety hazards (e.g., diseases and
accidents), and trigger negative ecological interactions with endemics (e.g., hybridisation with the Sister Island rock
iguana (Cyclura nubila caymanensis)). Therefore, control management is a priority for the Cayman Islands Department
of the Environment (DOE). In this poster, we provide information about green iguana population surveys conducted on
Grand Cayman in August 2014, 2015 and 2016. With the abundance estimates derived from these surveys, we conducted
a model-based assessment of population response to sustained removal effort. Although the green iguana is exposed
to human-induced mortality (e.g. hunting at private property, depredation by feral cats (Felis catus) and dogs (Canis
familiaris), and road kills), the population increased at an annual rate of 60% between 2014 and 2015 and 98% between
2015 and 2016 (not including hatchlings). Herein, we present the results from experimental culls organised by the DOE
in June 2016 in which 18,838 green iguanas were removed mainly from western Grand Cayman. Bounty hunter groups
and skilled hunters under contract both averaged about 100 iguanas killed per day. Removal effort, technique used and
crippling loss (i.e., shot but not retrieved) were among the variables quantified, which also included biological data to
establish a baseline understanding of green iguana population dynamics and response to control management. Applying
basic concepts of harvest theory and decision analysis, the DOE and USFWS are developing cost effective strategies
going forward.
Geraldes, P.; T. Pipa, N. Oliveira, C. Silva and S. Hervías. Setting-up a predator-free area on a Macaronesian island using a pest-proof fence
Haakonsson, J.; F. Rivera-Milan and E. Radford. Green iguana (Iguana iguana) monitoring and control efforts on Grand Cayman
711
Haber, E.; M. Eppinga, M. Ferreira dos Santos, M. Rietkerk and M. Wassen. Predicting the potential habitat of the invasive coral vine (Antigonon leptopus) using remote sensing
and species distribution modelling
Predicting the potential habitat of the invasive coral vine (Antigonon
leptopus) using remote sensing and species distribution modelling
E. Haber, M. Eppinga, M. Ferreira dos Santos, M. Rietkerk and M. Wassen
Copernicus Institute for Sustainable Development, Utrecht University, P.O. Box 80115,
Utrecht, 3508 TC, The Netherlands. <e.a.haber@uu.nl>
The spread of invasive plant species often outpaces the capacity to manage the invasions. Remote sensing can be used
to map the distribution of invasive plant species at a snapshot in time, but it is difficult to predict the future distribution
without incorporating the habitat preferences of the invasive species. Habitat suitability modelling is predictive, but
often suffers from an insufficient number of training points. In this study we combine vegetation classification models
based on remotely sensed imagery with habitat suitability models to predict the potential distribution of an invasive
vine, Antigonon leptopus (Polygonaceae), on two neighbouring Caribbean islands, St. Eustatius and Saba. A Support
Vector Machines (SVM) classification was produced for two WorldView-2 images of St. Eustatius (images acquired on 8
February 2011 and 24 August 2014) to produce maps of presence/absence of the vine. Pixels from the SVM classifications
where A. leptopus was present in both years were used as the dependent variable in the species distribution model for
St. Eustatius. The independent variables tested for the species distribution model were slope, elevation, soil hardness, soil
moisture, drainage area, distance to nearest building, and distance to nearest road. The results suggest that the potential
for A. leptopus invasion can be readily assessed for other islands in the Lesser Antilles. We illustrate this potential for
the neighbouring island of Saba, revealing that the expansion of A. leptopus may approach that of St. Eustatius if no
preventive actions are taken.
E. Haber, M. Eppinga, M. Ferreira dos Santos, M. Rietkerk and M. Wassen
Hammer, S. and J. Russell. The diet of ‘Viking mice’ on Nólsoy, Faroe Islands
The diet of ‘Viking mice’ on Nólsoy, Faroe Islands
S. Hammer and J. Russell
S. Hammer and J. Russell
University of Glasgow, 3R 51 Taylors Lane, Dundee, DD2 1AP UK. <sjurdur@hotmail.com>
Burrowing seabirds can be very vulnerable to rodents. Although there is abundant evidence for the negative impact of rats,
there is some recent evidence that mice (Mus musculus) can also have a detrimental effect on seabird populations. Introduced
by Vikings, mice are the only rodent on Nólsoy in the Faroe Islands, which also hosts one the largest European storm-petrel
colonies in the world. Using stomach dissections and stable isotope analysis we examined for evidence of storm petrel
consumption (eggs or chicks) in mice on Nólsoy. The findings may have implications for rodent management on Nólsoy
and other Ramsar sites in the Faroe Islands.
Hernández-Montoya, J.C.; L. Luna-Mendoza, A. Aguirre-Muñoz, F. Méndez-Sánchez, A. Duarte-Canizales, E. Rojas-Mayoral, S. Hall, Z. Peña-Moreno, S. Figueroa-Flores, D. Cosio-Muriel and
M. Latofski-Robles. Seabird restoration and advances towards the eradication of feral cats on Guadalupe Island, Mexico
Seabird restoration and advances towards the eradication of feral cats
on Guadalupe Island, Mexico
J.C. Hernández-Montoya, L. Luna-Mendoza, A. Aguirre-Muñoz, F. Méndez-Sánchez, A. Duarte-Canizales,
E. Rojas-Mayoral, S. Hall, Z. Peña-Moreno, S. Figueroa-Flores, D. Cosio-Muriel and M. Latofski-Robles
Director de Proyecto Isla Guadalupe, Grupo de Ecología y Conservación de Islas, A.C., Moctezuma 836,
Centro, Ensenada, Baja California 22800, Mexico. <julio.montoya@islas.org.mx>
Guadalupe Island (24,172 ha; 1,298 m) is located in the Pacific Ocean, 260 km off the Baja California Peninsula. It is
inhabited (ca. 150 people) and is part of a Biosphere Reserve, managed by Mexico’s National Commission for Natural
Protected Areas (CONANP) in collaboration with Grupo de Ecología y Conservación de Islas, A.C, a professionalised
Mexican NGO. Guadalupe has 223 vascular plant species (12% endemic), and hosts 139 taxa of birds, including seven
endemic races, six of which are considered extinct. Goats (Capra hircus), cats (Felis catus) and house mice (Mus musculus)
were introduced by the end of 19th century. Now a goat-free island, the feral cat is the most serious threat to biodiversity,
especially to surface- and burrow-nesting birds. The island hosts the most important breeding colony of Laysan albatross
(Phoebastria immutabilis) in the Eastern Pacific. Upon its colonisation in 1983, albatross adults and chicks have been
subject to severe predation by feral cats. To protect the albatross population, since 2003 we have done cat control around
the breeding area, now improved by the construction of a 700 m exclusion fence that protects 65 ha. Thanks to these
efforts, the number of breeding pairs has increased exponentially, with more than 400 to date. With a long-term vision
and the support from the National Fish and Wildlife Foundation (NFWF) and the Alliance WWF-Fundación Carlos
Slim, as of March 2017 we have moved from cat control to eradication. Timeframe for the eradication campaign will be
4.5 years. The methods will involve hunting, trapping (leg-hold traps) and detection dogs. Since the island is inhabited,
biosecurity measures are crucial since the eradication’s start. The achievement of the eradication will benefit native and
endemic seabirds and landbirds–especially those endangered–preventing their extinction.
J.C. Hernández-Montoya, L. Luna-Mendoza, A. Aguirre-Muñoz, F. Méndez-Sánchez, A. Duarte-Canizales,
E. Rojas-Mayoral, S. Hall, Z. Peña-Moreno, S. Figueroa-Flores, D. Cosio-Muriel and M. Latofski-Robles
712
Hudin, S. From island studies to mainland management
From island studies to mainland management
S. Hudin
S. Hudin
Natural Areas Conservancies Federation, 6 rue Jeanne d’’Arc, Orléans, 45000 France.
<stephanie.hudin@reseau-cen.org>
Islands have been the first and foremost natural habitats impacted by alien invasive species. With some delay, mainland
ecosystems are going through the same effects. Difficulties to manage and mitigate the effects of AI on indigenous species
are even greater, and the task to define a strategy more complex, on the continent. The numerous studies and reports
have helped taking the challenge up in some territories, and in France, it was decided to plan and organise efforts at the
hydrological scale of the great river, the Loire. Since 2002, exchanges between on-field managers and stakeholders have
permitted the creation of a network that has emerged as an example, as it edited a first interregional strategy of management
of alien species. The flow of information, the common edition of documents and supports for the management were the
first on the to-do list. Now, as Europe has announced its first 37-long list of priority species, the Loire working group is
revising its third version of a prioritised list and editing a first mapping of more than 60 species. Most of these species
came or were helped by the connected water system of the large river and its tributaries. As an interconnected habitat
system within the continent, the Loire basin can be compared to an island and as such has a lot to inform the managers
from the island alien invasives techniques used to eradicate the species. The creation of an atlas helps visualise which
species should be targeted for such efforts, and where to start. So, the achievement of more than 15 years is only the start.
Jesse, W.; J. Ellers, J. Behm and M. Helmus. Differential effects of human impact and habitat type on exotic and native species diversity on oceanic islands
Differential effects of human impact and habitat type on exotic and
native species diversity on oceanic islands
W. Jesse, J. Ellers, J. Behm and M. Helmus
W. Jesse, J. Ellers, J. Behm and M. Helmus
Ecological Science - Animal Ecology, Vrije Universiteit Amsterdam, De Boelelaan 1085,
1081HV Amsterdam, The Netherlands. <w.a.m.jesse@vu.nl>
Human land use is considered one of the main drivers of species compositional change. While some species experience
population decline as a result of human activities, others optimally exploit human-impacted environments. We hypothesised
that such contrasting responses could in part be attributable to species’ native or exotic origin. Our objective was to assess
the effect of human impact, defined as the addition of man-made substrates, on the taxonomic and functional composition
of exotic and native reptile assemblages of two anthropogenically impacted Caribbean islands. We extensively surveyed
insular reptile communities and recorded species abundance and richness data. Functional traits were obtained from
literature and used to construct functional diversity metrics for every sampled community. Of the composite environmental
variation among 114 sample plots, 46% could be reduced onto two PCA axes, resulting in a habitat structure axis (29%)
as well as a human impact axis (17%). PCA axes were subsequently regressed against various taxonomic and functional
abundance and diversity indices. Habitat structure and human impact independently affect abundance and diversity
indices across both islands. The direction of these effects largely depends on exotic or native origin. Exotic species are
never found in forest habitat, whereas native abundances peak in tropical forest. Exotic abundances are primarily affected
by human impact levels while native abundances show no significant association. Exotic species occur in higher numbers
on St Martin, which is likely due to regional shipping intensity rather than within-island factors. Furthermore, on St
Martin, exotic species significantly increase functional trait diversity by occupying unique functional niche space in
impacted environments. However, we found no indication of environmental filtering of functional trait values as a result
of human impact, rather habitat structural change seems to shift community trait values towards beneficial levels for
survival in non-forested environments.
Kanavy, D. and D. Threadgill. Genetic pest management technologies to control invasive rodents
Genetic pest management technologies to control invasive rodents
D. Kanavy and D. Threadgill
D. Kanavy and D. Threadgill
Molecular and Cellular Medicine, Texas A&M University, Texas A&M Health Science Center,
Joe H. Reynolds Medical Bldg. Rm. 440, College State, TX 77840, USA. <kanavy@tamhsc.edu>
Many strategies exist to manage invasive pests on islands, ranging from poison to trapping, with varying degrees
of success. Genetic technologies are increasingly being applied to insect pests, but so far, not to vertebrates. We are
implementing a genetic strategy to eradicate invasive mouse populations as another tool for pest control. Mus musculus,
the common house mouse, is one of the most widespread invasive species. Mice threaten human health, agriculture, and
biodiversity on many islands, particularly seabirds. Seabirds are endangered indirectly through competition for resources
or predators being attracted by the mice or directly with mice attacking chicks and eggs. Rodenticides are the most
common method of eradicating mice, but their use leads to poisoning of non-target species and has limited efficacy against
mice. An approach that could eliminate non-target species impact would be to engineer daughterless mice linked to a
gene drive system for self-sustained propagation. For this project, we have investigated exploiting a naturally occurring
gene drive, the t-complex. Using the tw2 haplotype of the t-complex, we observed the tw2 haplotype being transmitted to
offspring with a transmission distortion ratio of 95.3%. The daughterless phenotype is being accomplished by inserting the
Sry gene (male sex-determining gene) into an autosome containing the tw2 haplotype via CRISPR/Cas9 gene editing. The
presence of Sry will induce testis formation, regardless of the sex chromosomes naturally inherited. When Sry is inserted
into the t-complex, the desired gene will spread through the population, eliminating female offspring. This model system
will support studies to evaluate the effectiveness of crashing an invasive population without adversely affecting other
species. While still in the beginning stages, this is a novel idea and once this method has been perfected, it will open the
way to use this genetic strategy for the eradication of other invasive mammal species.
713
Keith, I.; J. Carlton and G. Ruiz. A new look at Galapagos fouling communities
A new look at Galapagos fouling communities
I. Keith, J. Carlton and G. Ruiz
I. Keith, J. Carlton and G. Ruiz
Marine Biology, Charles Darwin Foundation, Av. Charles Darwin, Puerto Ayora, Santa Cruz,
Galapagos Islands, Puerto Ayora, Ecuador. <inti.keith@fcdarwin.org.ec>
The maritime history of the Galapagos Islands begins in 1535 with the accidental discovery of the archipelago. For the
past 500 years the islands have endured a significant amount of terrestrial plant and animal introductions and, to some
extent, freshwater invasions; however, the number of marine introductions reported has been significantly lower. Research
has been conducted looking at the fouling communities of the Galapagos Marine Reserve (GMR) to provide a clearer
picture of the true scale of marine non-native species present in ports and harbours of the GMR. Settlement plates were
deployed for three and 14 months on floating docks on the Islands of Santa Cruz and San Cristobal. As a result, numerous
new records of introduced species of hydroids, polychaete worms, bryozoans, and ascidians, amongst other taxa, have
been documented for the Galapagos. The continued increase of marine traffic from many sources to the Galapagos Islands
concomitantly increases the risk of arrival of non-native species to this region. While research on terrestrial invasive species
is well established, research on marine invasive species and their impacts in the GMR has been less investigated. The
Charles Darwin Foundation (CDF), the Galapagos National Park Directorate (GNPD) and the Galapagos Biosecurity
Agency (ABG) have been working together to improve the marine biosecurity standards for the GMR, and some clear
advances are now in place. A synthesis of marine biosecurity based on prevention, early detection and management of
marine non-native species is presented and potential management strategies discussed.
Kelly, J.; K. Springer, C. Stringer, A. Schofield and T. Glass. Planning processes for eradication of mice on Gough Island
Planning processes for eradication of mice on Gough Island
J. Kelly, K. Springer, C. Stringer,
A. Schofield and T. Glass
J. Kelly, K. Springer, C. Stringer, A. Schofield and T. Glass
Species and Habitats, Royal Society for the Protection of Birds, RSPB, Headquarters,
the Lodge, Sandy, Bedfordshire SG19 2DL, UK. <John.Kelly@rspb.org.uk>
Gough Island, part of the remote Tristan da Cunha group in the South Atlantic, is considered one of the most important
seabird islands on the planet. A UK Overseas Territory and World Heritage Site, Gough supports millions of breeding
seabirds and the UK’s only Critically Endangered bird species, the Tristan albatross and Gough bunting. Invasive house
mice (Mus musculus) were introduced in the 1800s and prey on hundreds of thousands of chicks each year. It has been
predicted that the Tristan albatross faces extinction within c. 30 years unless the mice are eradicated. Led by the RSPB and
the Government of Tristan da Cunha, the Gough Island Restoration Programme aims to eradicate mice from Gough Island
using aerial baiting containing anticoagulant toxin; a methodology established during previous island eradications. Now
in the operational planning phase, the programme aims for mouse eradication on Gough in 2019. Applications for various
approvals are required and a captive bird management programme designed to protect land birds vulnerable to secondary
poisoning. As well, robust operational planning and detailed logistical planning need to be completed. Situated around
2,800 km from Cape Town, South Africa, Gough Island presents challenges including its remoteness, terrain, weather and
cave systems. Long lead in times for planning are required, reflecting the scale and complexity of logistics and regulatory
requirements.
Knapp, J.; C. Boser, J. Randall, E. O’Byrne and S.A. Morrison. Perils of saving the smallest for the last: lessons learnt about sequencing eradications on Sant Cruz Island, CA
Perils of saving the smallest for the last: lessons learnt about
sequencing eradications on Santa Cruz Island, CA
J. Knapp, C. Boser, J. Randall, E.
O’Byrne and S.A. Morrison
J. Knapp, C. Boser, J. Randall, E. O’Byrne and S.A. Morrison
The Nature Conservancy, 532 E. Main Street, Suite 200, Ventura, California, USA. <jknapp@tnc.org>
The biota of many islands have been damaged by invasive species, but in a growing number of cases island invaders have
been successfully eradicated. Many eradication projects target vertebrate species whose size and harmful effects make
them particularly conspicuous. Unfortunately, smaller and less conspicuous invaders, including invertebrates and plants,
may be overlooked before or following successful eradications, and their continued presence can limit the attainment of
some of the management goals that may have motivated the earlier eradications. For example, vegetation recovery that
often follows removal of herbivores can make eradication of remaining invaders more difficult. Vertebrate eradications
can result in the release of perceived “secondary” invaders, which can compromise the benefits of the initial eradication.
We review the suite of eradications that have occurred or are underway on Santa Cruz Island, USA, which have focused
on plant, invertebrate, avian, and mammalian taxa. We discuss the biological impacts of – including the long-term
management challenges created by – decisions regarding which taxa were eradicated when. We recommend that prior
to undertaking any eradication all invasive species and the resources they threaten be evaluated with regard to how the
sequence of eradications may positively or negatively affect any eradication efforts that may follow.
714
La Morgia, V.; D. Paoloni, P. Aragno and P. Genovesi. Citizens’ attitude towards the removal of grey squirrels in Italy: what support do we need?
Citizens’ attitude towards the removal of grey squirrels in Italy:
what support do we need?
V. La Morgia, D. Paoloni,
P. Aragno and P. Genovesi
V. La Morgia, D. Paoloni, P. Aragno and P. Genovesi
Institute for Environmental Protection and Research – ISPRA, Ozzano Emilia (BO), Via Ca’’ Fornacetta 9,
Ozzano Emilia (BO), 40064 Italy. <valentina.lamorgia@isprambiente.it>
Grey squirrels (Sciurus carolinensis) were introduced in Umbria, Central Italy, in 2000. Since then, they have successfully
occupied a range of about 50 km2. The LIFE U-SAVEREDS Project now aims to eradicate this isolated population, but
the squirrel distribution is centred on the city of Perugia and animals are particularly abundant in public urban parks
and private house gardens. Thus, part of the public opinion opposes the project activities. For this reason, the overall
management strategy involves both direct (capture and euthanasia) and indirect (capture and surgical sterilisation) removal
of the animals. Further, a Decision Support System including the evaluation of social issues was specifically developed. It
identified spatial intervention priorities and it allowed the start-up of grey squirrel management in areas where the
overall social context was favourable. At the same time, we implemented a targeted information campaign to increase
the population’s knowledge on the issue of invasive alien species and, most important, to actively involve citizens in the
Project. As a consequence, several citizens agreed to collaborate on the eradication campaign. Following the intervention
in different management units, characterised by a different acceptance level of the eradication campaign, we now evaluate
how the citizens’ collaboration affected the outcome of Project activities. The percentage of accessible land (ranging
from 84 to 21%) for each management unit was quantified through mapping and modelling in GIS environment, and was
compared to the outcome of direct removal of the animals. In 2016, 470 animals were removed, and preliminary results
suggest that the spatial configuration of accessible lands also plays an important role in the eradication. Considering both
social and technical issues, simulations were finally implemented to assess the success probability of the eradication
campaign at local scales.
Lennon, Z.; H. Wittmer and N. Nelson. Computer modelling of complex interstitial spaces to protect endemic island lizards from invasive mice
Computer modelling of complex interstitial spaces to protect endemic
island lizards from invasive mice
Z. Lennon, H. Wittmer and N. Nelson
Z. Lennon, H. Wittmer and N. Nelson
School of Biological Sciences, Victoria University Wellington, Wellington, New Zealand. <zoe.lennon@vuw.ac.nz>
New Zealand is home to a large diversity of endemic lizards, with 42 gecko (Diplodactylidae) and 55 skink (Scincidae)
taxa, ~ 84% of which are classified as Threatened or At Risk. Habitat destruction and invasive mammalian predators are
responsible for much of this decline. Endemic lizard species are afforded legal protection in New Zealand, meaning that
when populations are threatened by human activity such as road construction, individual animals must be salvaged and
moved to a safe location (mitigation translocation). Mitigation translocations of lizards in New Zealand often involve
habitat enhancement, for instance building new rock pile habitat. However, there is little research to show if habitat
enhancement actually has the intended effect of providing better habitat for lizards, or if there might be undesirable side
effects such as creating habitat for invasive predators like mice (Mus musculus). I describe a novel technique using a
computer game physics engine (Unity, PhysX) to investigate the best rock pile design to protect translocated skinks while
hindering the movement of mice. I achieve this by measuring the interstitial spaces in virtual rock piles to determine which
compositions (sizes, shapes of constituent rocks) will maximise spaces skinks are able to fit through while minimising
spaces mice are able to fit through, enabling skinks to avoid predation by mice. My virtual approach to this problem
allows me to model complex spaces which were unable to be measured using previous, physical techniques. Predictions
from modelling are confirmed using data from computed tomography (CT) scans of real rock piles. The design that results
from this research will be tested in a real mitigation translocation to determine whether skinks have higher survival in
my rock pile designs. This research will inform understanding of invasive predator/prey interactions and conservation of
species threatened by invasive mammals.
Li, J.; C. Zhao and X. Zhao. An integrated physical control method on Spartina alterniflora
An integrated physical control method on Spartina alterniflora
J. Li, C. Zhao and X. Zhao
J. Li, C. Zhao and X. Zhao
Chinese Research Academy of Environmental Sciences, Beijing, China. <lijsh@craes.org.cn>
Spartina alterniflora is a noxious invasive plant due to its ecological impact. An integrated method of mowing plus
shading (MS) was conducted for control of Spartina alterniflora in Guangxi, China. Plant height, node number, node
length, basal stem diameter, aboveground biomass and population density of this weed were used to compare the
effectiveness of mowing and MS. Results showed that all characters of S. alterniflora were significantly decreased by
mowing plus shading (P<0. 05), and only node number, plant height and aboveground biomass were suppressed by
mowing alone. It was indicated that clonal growth and sexual reproduction of S. alterniflora were absolutely inhibited
by mowing plus shading in the whole growth season. We also found the restraining effect of mowing plus shading was
positively correlated with shading degree. The light transmittances of single layer shading net, double layers shading net
and triple layers shading net were 15.27%, 2.29% and 0.31%, respectively, and rhizome survival rate were 3.68%, 2.09%
and 1.70% in November respectively. Above-ground parts were all dead in November before mowing plus single layer
shading treatment, while they were all dead at July in mowing plus double layers shading treatment and mowing plus
triple layers shading treatment. In the future, mowing plus shading may be used as an effective method of controlling
S. alterniflora.
715
Libeau, M.; R. Pouteau, R. Taputuarai and J.-Y. Meyer. Predicting the risk of plant invasion on islands: the case of Miconia calvescens in the Marquesas, French Polynesia (South Pacific)
Predicting the risk of plant invasion on islands: the case of Miconia
calvescens in the Marquesas, French Polynesia (South Pacific)
M. Libeau, R. Pouteau, R. Taputuarai
and J.-Y. Meyer
M. Libeau, R. Pouteau, R. Taputuarai and J.-Y. Meyer
Délégation à la Recherche, Government of French Polynesia, B.P. 20981, 98713 Papeete,
Tahiti, French Polynesia. <mel.libeau@gmail.com>
Miconia calvescens (Melastomataceae), a small tree native to Central and South American rainforests, is a dominant plant
invader in the Society Islands (French Polynesia), Hawaii, New Caledonia, and tropical Australia, thus listed as one of the
world’s 100 worst invasive species. This fast growing, early reproducing and prolific seed producer (small fleshy fruits
dispersed by birds over long distances) with a long-lasting soil seed bank (several decades) was first detected 20 years ago
in the Marquesas (French Polynesia), a remote archipelago with a unique and endangered native flora (48% of endemism
and 145 threatened species). Despite some eradication efforts, several new outbreaks have been located in the last few years
on the largest island of Nuku Hiva. In this alarming context, it is urgently needed to determine the potential distribution of
the species in order to assess the risk of invasion and refine the areas for further surveys and control. Species distribution
models (SDMs) are numerical tools that project species distribution from the combination of species occurrences with
environmental variables. Fitting an SDM on the basis of Marquesas populations to predict the future of Miconia over the
archipelago would violate the equilibrium assumption behind SDMs. Moreover, projecting the environmental envelope
occupied by the species in its native range would ignore inherent characteristics of island ecosystems (e.g. low species
richness, low functional redundancy, competitive release, vacant niches, restricted and specialised habitats) that leave
them much more vulnerable than continents to biological invasions. As a result, the environmental distribution of Miconia
across the similarly-sized high-elevation islands of the Society and the Hawaiian archipelagos was projected over the
Marquesas. The different SDMs agree that Miconia will spread over a large area of native lowland rainforest and montane
cloud forest in Nuku Hiva unless appropriate control strategies are rapidly adopted.
Lindholm, A. and B. König. The secret life in Switzerland of an island pest, the house mouse
The secret life in Switzerland of an island pest, the house mouse
A. Lindholm and B. König
A. Lindholm and B. König
Department of Evolutionary Biology and Environmental Studies, University of Zurich,
Winterthurerstrasse 190 CH-8057 Zurich, Switzerland. <anna.lindholm@ieu.uzh.ch>
House mice (Mus musculus) can have harmful effects on island biota, and are frequently the targets of eradication efforts. The
success of eradication strategies will be influenced by how well the biology of the house mouse is understood. We have
carried out a long-term study of a free-living population of wild house mice in Switzerland, following mice in the population
from cradle to grave (or disappearance). Adult mice are chipped and a system of antennas installed at the entrances to nests
have allowed us to monitor the movements of house mice and observe their social lives in unprecedented detail. House
mice live in large but fairly closed social groups of males and females, sharing several nests. Competition between males
and between females has led to dramatic reproductive skews in both sexes and high rates of infanticide, despite ad libitum
food availability. Multiple paternity within litters is common. Cooperation between breeding females within a social
group also occurs, in communal nursing of all pups present in the same nest. Population density has increased over time,
giving rise to larger group sizes. How this increase in social tolerance is achieved is unclear. Furthermore, population
size recovered rapidly from an epidemic that killed ca 30% of adults. We are currently focused on understanding factors
influencing reproductive suppression, dispersal likelihood, social tolerance and cooperation between females, including
genetic influences, such as the t haplotype. Our studies may be useful in predicting the outcome of interventions to house
mouse populations.
Little, A.; A. Aguirre-Muñoz, G. Seutin, L. Wein, P. Nantel, H. Berlanga, F. Méndez-Sánchez, J. Putnam, E. Iñigo-Elías and G. Howald. Catalysing conservation of islands through
collaboration: a North American perspective
Catalysing conservation of islands through collaboration:
a North American perspective
A. Little, A. Aguirre-Muñoz, G. Seutin, L. Wein, P. Nantel, H. Berlanga, F. Méndez-Sánchez, J. Putnam,
E. Iñigo-Elías and G. Howald
U.S. Fish and Wildlife Service, 1901 Spinnaker Drive, Ventura, CA 93001, USA. <annie_little@fws.gov>
The countries of North America are inextricably linked through shared species, habitats, and ecosystems. Over the last
several decades, significant efforts have been made to protect and restore unique island ecosystems within the three
nations. Many of the significant advances have been through bi and trilateral collaboration. In recognition of the value
of cross border collaboration, in 2014, the governments of Canada, United States and Mexico signed an agreement
to protect fragile island ecosystems and their imperilled species. This agreement, endorsed under the scope of the
Trilateral Committee for Wildlife and Ecosystem Conservation and Management, strengthens the on-going collaboration
between the three nations on the conservation and restoration of island ecosystems and their adjacent coastal and
marine environments. Through coordinated efforts, government and NGO partners are accelerating investment in island
conservation programmes across North America with a focus on invasive species, biosecurity, restoration, and regulatory
processes. Activities include prioritisation of invasive species on a continental scale, sharing of expertise and technology,
strengthening institutional capacities, and leveraging (shared) funding and support. These partnerships have accelerated
conservation outcomes across North America, including the eradication of invasive species in Canada, protection of rare
species and ecosystems in the United States, and a systematic and comprehensive programme to conserve and restore
islands in Mexico.
A. Little, A. Aguirre-Muñoz, G. Seutin, L. Wein, P. Nantel, H. Berlanga, F. Méndez-Sánchez, J. Putnam, E. Iñigo-Elías and G. Howald
716
McHenry, E.; X. Lambin, T. Cornulier and D. Elston. The value of monitoring and the price of uncertainty in the management of an invasive population
The value of monitoring and the price of uncertainty in the management
of an invasive population
E. McHenry, X. Lambin, T. Cornulier and D. Elston
E. McHenry, X. Lambin, T. Cornulier and D. Elston
School of Biological Sciences, Zoology Building, University of Aberdeen, Tillydrone Ave,
Aberdeen, UK. <ewan.mchenry@abdn.ac.uk>
Improving decision-making regarding resource allocation for the control of invasive populations often requires monitoring
to obtain information on the state of the population. The cost incurred by monitoring detracts from the resources available
for direct control, and so, for monitoring to be feasible, the information gained must have greater value to management than
the costs of obtaining it. We aim to provide generalisable recommendations on the use of monitoring data to inform the
management of invasive species. Here we present a simulation study inspired by the control of invasive American mink in
Scotland. Mink populations exhibit seasonal dynamics with highly dispersive juvenile and intrasexually territorial adult
life stages. Control effort was simulated to be dependent on season and perceived variation in the abundance of settled
adults. Imperfect monitoring can result in false positive or negative detections of adults, allowing the value of reducing
uncertainty by increasing monitoring effort to be explicitly considered in terms of its impact on the invasive population
and unplanned overspending of effort budgets. The modelling framework allows the relative value of monitoring effort
to be assessed for different control strategies. Future work will utilise large-scale mink control data and surveys of a
threatened endemic prey species, the water vole, to estimate the level of mink control required for a high probability of
persistence of water vole metapopulations. This will inform future simulation work identifying the balances between
monitoring and intervention that maximise the probability of favourable conservation outcomes for fixed cost.
Miranda, M.A.; C. Barceló, D. Borràs, A. González, M. Leza and C. Paredes-Esquivel. Invasive arthropods of ecological, agricultural and health importance recently introduced in the Balearic Islands
(Spain)
Invasive arthropods of ecological, agricultural and health importance
recently introduced in the Balearic Islands (Spain)
M. A. Miranda,
C. Barceló, D. Borràs,
A. González, M. Leza and
C. Paredes-Esquivel
M. A. Miranda, C. Barceló, D. Borràs, A. González, M. Leza and C. Paredes-Esquivel
Biology, University of the Balearic Islands, Cra. Valldemossa km 7,5, Palma de Mallorca,
Balearic Islands 07122 Spain. <ma.miranda@uib.es>
The Balearic Islands archipelago (Mallorca, Menorca, Ibiza and Formentera) is located in the western part of the
Mediterranean Sea. Like other places in the area, the Balearic Islands are exposed to the introduction of several highly
invasive species, some of them even world-wide distributed. In fact, the Balearic Islands have a long record of introduced
species including different taxa of animals. Here we focus on those invasive arthropod species that were introduced during
the last decade and have high impact on ecosystems, agriculture and human health. We present a description of the current
situation of the incursion, spread and impact of the tomato leafminer (Tuta absoluta, Gelechiidae); the red palm weevil
(Rhynchophorus ferrugineus, Curculionidae); the Asian tiger mosquito (Aedes albopictus, Culicidae) and the Asian hornet
(Vespa velutina, Vespidae). We conducted an analysis of the path of entry of the different species to the Balearic Islands,
considering means of transport including commodities and human transportation. We also analysed the current impact of
the presence of the above-mentioned species on agriculture (i.e. increase use of insecticides), landscape (i.e. palm trees
destruction), human health (i.e. vector-borne diseases) and ecosystems (i.e. impact on bee population). Results indicate
that some invasive species, such as T. absoluta could be effectively managed by farmers after a period of adaptation of
control procedures to the new pest. The impact on landscape by species such as the red palm weevil has notably increased
since its introduction and its expansion is currently uncontrolled. Species such as the Asian tiger mosquito have changed
the perception of citizens on the risk of vector-borne diseases, due to the current expansion and its possible implication
on arbovirus transmission. Finally, the recent detection of the Asian hornet, has deeply increased concern about the role
of bees as an essential component of ecosystems.
Mohanty, N.P.; G.J. Measey, A. Sachin, G. Selvaraj and K. Vasudevan. Using key-informant surveys to reliably and rapidly estimate the distributions of multiple insular invasive species
Using key-informant surveys to reliably and rapidly estimate the
distributions of multiple insular invasive species
N.P. Mohanty, G.J. Measey, A. Sachin,
G. Selvaraj and K. Vasudevan
N.P. Mohanty, G.J. Measey, A. Sachin, G. Selvaraj and K. Vasudevan
Centre of Excellence for Invasion Biology, Department of Botany and Zoology,
Stellenbosch University, Stellenbosch, South Africa. <nitya.mohanty@gmail.com>
Knowledge of invasive species’ distributions is critical to manage established populations. Distribution at large spatial
scales can be rapidly estimated through public surveys, though reliability of such information must be tested. We gathered
detection/non-detection data for the Indian bullfrog (Hoplobatrachus tigerinus), the common myna (Acridotheres tristis),
the house sparrow (Passer domesticus), and the giant African snail (Achatina fulica) through interviews in 91 sites on
inhabited islands of the Andaman Archipelago. We interviewed 855 key informants comprising farmers, plantation
workers, and aquaculturists, from January to March and September to December 2015. Additionally, we obtained
detection/non-detection data for the Indian bullfrog (75 sites), the common myna (65 sites), the house sparrow (39
sites), and the giant African snail (29 sites) through systematic visual encounter surveys and opportunistic records. We
corrected the informant data for false positive detections in an occupancy framework and estimated the distribution of the
four species. The Indian bullfrog occurred on all islands, except Baratang, Long, and Little Andaman Islands. The giant
African snail was ubiquitous, occurring on all islands. The distribution of the common myna was most likely influenced
by roads, while ports might be significant for the house sparrow invasion. The findings substantiate the efficacy of public
surveys in generating rapid distribution information on multiple invasive species simultaneously.
717
Negrín Pérez, Z.; D. Da Re, M. Bernardos and B. Garrido. Time germination response to temperature and light conditions in Ulex
Time germination response to temperature and light conditions in Ulex
Z. Negrín Pérez, D. Da Re,
M. Bernardos and B. Garrido
Z. Negrín Pérez, D. Da Re, M. Bernardos and B. Garrido
La Laguna University, San Cristóbal de La Laguna, Tenerife, Canary Islands, Spain. <znegrinp@ull.es>
The Canary Islands are widely recognised as an outstanding biodiversity hotspot worldwide. Biological invasion, together
with wildfire, are two of the main factors of biodiversity loss in the islands, due to low habitat diversity, their simplified
trophic webs and the high rates of endemism. Ulex europaeus is an invasive species, which is in the early period of its
naturalisation, but it is already affecting two of the richest ecosystems of the island: laurel and pine forests. Previous
studies were focused on shade and post-fire conditions as key factors in the growth of young plants, while less attention
was oriented to factors linked to seeds germination. The goals of this study are to understand the role of light exposure and
temperature shocks in U. europaeus germination, and to highlight the optimal conditions. In this study, seeds experienced
three different light exposures (total darkness, 70% shade and full light) with eight different temperature ranges (from
30º to 130º C). Then seeds were exposed to temperature shocks for 1, 5 and 10 minutes. The results of DCA and standard
statistical analysis show that light exposure has a low relationship with seed germination. Significant differences were
found between temperature and time germination: a short exposure to temperatures between 40º to 70 ºC has a positive
effect on the germination of U. europaeus, although higher temperatures inhibit germination. These results enable a
greater understanding of the relationships of U. europaeus and environmental conditions of fire zones, but further studies
that take into consideration the role of litter and ashes are needed also.
Oliveira, N.; P. Geraldes, I. Fagundes, P. Oliveira and J. Andrade. Rat eradication from Berlengas Island, Portugal
N. Oliveira, P. Geraldes,
I. Fagundes, P. Oliveira and
J. Andrade
Rat eradication from Berlengas Island, Portugal
N. Oliveira, P. Geraldes, I. Fagundes, P. Oliveira and J. Andrade
Marine Conservation Department, Sociedade Portuguesa Estudo das Aves, Av Columbano Bordalo
Pinheiro, 87, 3º, Lisbon, 1070-062 Portugal. <nuno.oliveira@spea.pt>
The Berlengas Archipelago, six miles off the coast of Portugal, is composed of a main island of 95 ha and five small
islets. It holds the only colony of Cory’s shearwaters (Calonectris borealis) on continental Portugal, and the largest
Portuguese colonies of shag (Phalacrocorax aristotelis) and yellow-legged gull (Larus michaellis). A breeding population
of Madeiran storm-petrel (Hydrobates castro) of unknown size also breeds on the nearby islets. The native vegetation
includes three endemic species of conservation concern. The presence of IAS in Berlengas (black rat, Rattus rattus)
is considered to have a significant impact on several seabird species and on the island vegetation. It is also thought to
prevent colonisation of the main island by prospecting Madeiran storm-petrels that are often recorded there. Within the
scope of an EU funded LIFE programme, a full rat eradication started in 2014, and is still underway, to restore the local
ecosystem. A grid of 1,000 closed baiting stations (25 m x 25 m) was used with cereal pellets containing the anticoagulant
brodifacoum. Special care was taken to prevent secondary poisoning of non-target species, and a full assessment of the
invasive alien species populations was made before any control action started. Species abundance, local distribution,
inter-annual abundance variation, and genetic characterisation was determined prior to the baiting operations that started
on September 2016. The last confirmed rat sign was registered at the end of October during the weekly monitoring
surveys. After December 2016, the remaining toxic baits were removed from the baiting stations and non-toxic scented
baits were used to detect any remaining signs of rat activity. The operational phase is expected to last at least two years
after the first baiting station was set and we expect that after the eradication the subsequent recovery by seabirds and
native plants will make a substantial conservation contribution at European level.
Palmas, P.; R. Gouyet, T. Ghestemme, A. Matohi, E. Terorohauepa, I. Tauapaohu, C. Blanvillain, J. Zito, D. Beaune and E. Vidal. Response of an open feral cat population to an intensive control
programme for improving the Critically Endangered Fatu Hiva monarch conservation strategy
Response of an open feral cat population to an intensive control
programme for improving the critically endangered Fatu Hiva
monarch conservation strategy
P. Palmas, R. Gouyet, T. Ghestemme, A. Matohi, E. Terorohauepa, I. Tauapaohu, C. Blanvillain, J. Zito,
D. Beaune and E. Vidal
Institut Méditerranéen de Biodiversité et d’Ecologie marine et continentale (IMBE), Aix Marseille Université,
CNRS, IRD, Avignon Université, Centre IRD de Nouméa, BPA5, 98848 Nouméa cedex,
Nouvelle-Ca, Nouméa, New Caledonia. <palmas.pauline@gmail.com>
The Fatu Hiva monarch (Pomarea whitneyi) is an endemic terrestrial bird of Fatu Hiva Island (Marquesas, French
Polynesia) red-listed by IUCN as Critically Endangered since 2000. Recent decline of the remaining populations is
particularly alarming with 30 individuals currently remaining while 275 were still present 10 years ago. Introduced
predators have been identified as the main cause of extirpation, especially ship rat (Rattus rattus), introduced in the 1980s
and feral cats (Felis catus) that reatly impact the remaining population at all bird demographic stages (chicks at nest,
fledging chicks, and adults). An intensive feral cat culling programme has therefore been progressively implemented over
the past five years by SOP-Manu (Birdlife representative in FP) on a 290 ha controlled area to secure part of the Fatu
Hiva monarch population. By using data from 43,845 trap-nights and > 189,000 camera-trap images we evaluated the
effects of this intensive cat control on feral cat abundance in the treated area (three different indices: abundance index,
minimum number of individuals and individual capture histories using the spatially explicit capture–recapture (SECR)
model to calculate densities). In parallel, we fitted cats with GPS collars to (i) understand the recolonisation process from
the untreated adjacent areas and (ii) assess the risk due to domestic and stray cats from the nearby village. These results
will help to refine and optimise feral cat control strategy in this large, mountainous and inhabited island where eradication
could be considered, although difficult. The protected and treated area includes 25 of the 30 remaining individuals whose
only three breeding pairs of this species are on the verge of extinction.
P. Palmas, R. Gouyet, T. Ghestemme, A. Matohi, E. Terorohauepa, I. Tauapaohu, C. Blanvillain, J. Zito, D. Beaune and E. Vidal
718
Palmas, P.; H. Jourdan, E. Bonnaud, F. Rigault, L. Debar, H. De Méringo, E. Bourguet, R. Adjouhgniope and E. Vidal. Feral cats threaten the
outstanding endemic fauna of the New Caledonia biodiversity hotspot: implications for feral cat management strategy
Feral cats threaten the outstanding endemic fauna of the New Caledonia
biodiversity hotspot: implications for feral cat management strategy
P. Palmas, H. Jourdan, E. Bonnaud, F. Rigault, L. Debar, H. De Méringo, E. Bourguet, R. Adjouhgniope and E. Vidal.
Institut Méditerranéen de Biodiversité et d’Ecologie marine et continentale (IMBE), Aix Marseille Université,
CNRS, IRD, Avignon Université, Centre IRD de Nouméa, BPA5, 98848 Nouméa cedex,
Nouvelle-Ca, Nouméa, New Caledonia. <palmas.pauline@gmail.com>
Among invasive species, feral cats (Felis catus) are one of the most successful and harmful predator species leading to
dramatic loss of biodiversity on the world’s islands. Effective feral cat management (eradications, controls) on numerous
islands generally resulted in positive effects for native biodiversity conservation. The lack of feral cat diet study in the
New Caledonia archipelago was an obstacle (i) to assess the importance of feral cat issues and (ii) to provide relevant
guidelines for feral cat population management to mitigate their impacts. Our study aims to evaluate feral cat threats
to the outstanding biodiversity at this major biodiversity hotspot in order to provide recommendations to prioritise
management and preservation of native biodiversity. We investigated feral cat predation by analysing 5,300 cat scats
sampled at 14 selected representative sites giving an accurate picture of the four main natural habitats. Feral cats prey
upon at least 43 vertebrate species, 20 of which are IUCN Red List threatened species. New Caledonia is the home of
30.8% of IUCN threatened species preyed on by feral cats, while representing only 0.12% of the total area of islands
(including Australia). Thus, this study increases at least by 44.4% the number of IUCN threatened species vulnerable
and preyed upon by feral cats across islands worldwide. Threatened vertebrate species preyed on by feral cats are skinks,
flying foxes and petrels, and their predation mainly occur in humid forest and maquis mosaic sites. The results of this
study prompted feral cats to be listed among the top-five priority species for future management in New Caledonia. We
therefore recommend that future actions be prioritised based upon the most critical species situations (most impacted and
endangered native species, i.e. skinks, flying foxes, seabirds), and targeting first some geographic areas of manageable
size already offering some management facilities and support.
P. Palmas, H. Jourdan, E. Bonnaud, F. Rigault, L. Debar, H. De Méringo, E. Bourguet, R. Adjouhgniope and E. Vidal.
Pandoo, S.; P. Ragen, B. Vishnuduth, Z. Jhumka and J. Mauremootoo. Scaling up invasive plant management for ecosystem restoration in Mauritius: successes and challenges
Scaling up invasive plant management for ecosystem restoration in
Mauritius: successes and challenges
S. Pandoo, P. Ragen, B. Vishnuduth,
Z. Jhumka and J. Mauremootoo
S. Pandoo, P. Ragen, B. Vishnuduth, Z. Jhumka and J. Mauremootoo
InSpiral Pathways, 23 Southside, Congresbury, Bristol, Avon BS495BS, UK. <seewajee.pandoo@undp.org>
Objectives: To document progress made in the last 30 years in restoration of Mauritian terrestrial ecosystems with a
primary focus on the invasive plant management component. Methodology: Invasive plant management activities and
results have not been systematically monitored so much of the evidence for management effectiveness is anecdotal. As
part of the UNDP-GEF PAN Project (Expanding coverage and strengthening management effectiveness of the protected
area network on the island of Mauritius) practitioners’ knowledge of plant restoration practices undertaken to date has
been synthesised in a ‘Good Practice Guide for Native Forest Restoration in Mauritius’. This synthesis has allowed us to
take stock of management effectiveness. Results: The area under restoration in mainland Mauritius has increased from
< 10 ha in the 1980s to almost 100 ha from the 1990s to the 2000s to nearly 500 ha today. Per hectare weeding costs
in real terms have been reduced by more than half during this period, principally by moving away from pure manual
weeding to an approach that involves a mixture of manual and chemical approaches, and more effective implementation
arrangements. There are certain common practices in invasive plant management but there are also site and species-specific
weeding approaches, and initiatives that could be scaled up such as utilising weed biomass as a cost-recovery option, and
using mulching as a weed suppression technique. Conclusions: Much progress has been made at both the site level and
nationally for the country’s entire PA estate. The Good Practice Guide will help disseminate this knowledge among new
and existing practitioners as a contribution to management effectiveness.
Parish, J. Implementing an early detection programme on Catalina Island: prioritising landscaped grasses
J. Parish
Implementing an early detection programme on Catalina Island:
prioritising landscaped grasses
J. Parish
Catalina Island Conservancy, PO BOX 2739, Avalon, California 90704, USA. <jparish@catalinaconservancy.org>
Invasive species pose a significant threat to native plant species by increasing the risk of wildland fires, displacing
native species, and altering native habitat. Recent trends in Southern California landscaping have increased the demand
for drought resistant grasses, and often these are non-native species. Catalina Island Conservancy’s Catalina Habitat
Improvement and Restoration Program’s invasive plant project developed an early detection and rapid response project,
the Avalon Grasses Initiative, in 2016 to address recent introductions of three highly invasive grass species installed
in landscaping. The Avalon Grasses Initiative implements “target-based” early detection methodology created by
previous research and early detection efforts conducted on mainland California. Roadside surveys detect populations and
staff walks through the community going door to door to request permission to remove target species and offer native
plants as replacement. Initial surveys detected 30 populations of Cortaderia selloana, Pennisetum setaceum, and Stipa
tenuissima. Control and survey efforts are on-going, but more than 1,000 plants have already been removed and replaced
with native Catalina Island plant species grown in the Conservancy’s native plant nursery.
719
Parrott, D.; G. Massei, R. Ridley, J. Sandon, M. Lambert, D. Cowan and M. Sutton-Croft. Challenges and opportunities for lethal and non-lethal management of
non-native ungulates on islands: feral pigs, goats and cows.................................
Challenges and opportunities for lethal and non-lethal management
of non-native ungulates on islands: feral pigs, goats and cows
D. Parrott, G. Massei,
R. Ridley, J. Sandon,
M. Lambert, D. Cowan
and M. Sutton-Croft
D. Parrott, G. Massei, R. Ridley, J. Sandon, M. Lambert, D. Cowan and M. Sutton-Croft
National Wildlife Management Centre, Animal & Plant Health Agency, Sand Hutton,
York, YO41 1LZ, UK. <dave.parrott@apha.gsi.gov.uk>
The National Wildlife Management Centre (NWMC), which is part of the UK Government’s Animal and Plant Health
Agency (APHA), has supported and delivered the management of non-native species including commensal rodents and
ungulates on a variety of islands across the world. The NWMC utilises a range of both lethal and non-lethal approaches
in these projects. We will present two ungulate case studies highlighting the merits and limitations of each of these
approaches. This includes NWMC’s recent work to reduce the population of feral goats on Great and Little Tobago in the
British Virgin Islands. NWMC worked with the RSPB and the National Parks Trust of the Virgin Islands to directly reduce
this population through humane culling and trained locally-based staff to increase their capacity to deliver similar projects
in the future. Although this project proceeded as intended, lethal control is not suitable in all situations. We have found
that although it can deliver rapid reductions in populations in the short term, and is often the best option where complete
eradication is the aim of the management intervention, it may be unfeasible or be unacceptable due to its impact on the
environment and on animal welfare. Fertility control is increasingly being considered as an alternative long-term solution
to reduce population sizes of problematic species. This non-lethal method can offer a humane, publicly acceptable method
to reduce population sizes. Recent advances in research and development have led to the registration of novel fertility
control agents for wildlife. Species-specific systems to deliver baits containing oral contraceptives to target species are
now available. In addition, the development of new software and mathematical models has allowed researchers to make
predictions of the effects of fertility control on population size. In our second case study, we present experimental data on
the efficacy of fertility control agents on model wildlife species and illustrate examples of species-specific bait delivery
systems.
Pink, C.; D. Algar and P. Green. Diet of introduced black rats Rattus rattus on Christmas Island: setting the scene with stomach and stable isotope analysis
Diet of introduced black rats (Rattus rattus) on Christmas Island:
setting the scene with stomach and stable isotope analysis
C. Pink, D. Algar and P. Green
C. Pink, D. Algar and P. Green
Evolution and Ecology, Latrobe University, P.O. Box 867, Christmas Island,
Western Australia 6798, Australia. <caitlyn.pink@environment.gov.au>
The black rat (Rattus rattus) is an introduced and invasive rodent, negatively affecting endemic species on many islands
worldwide. Black rats have existed on Christmas Island for more than 100 years, and feral cats (Felis catus), also on the
island, are poised for imminent eradication. The risk of meso-predator release needs to be considered, and a combination
of stomach and stable isotope analyses of rats was used to determine potential impacts on native fauna should such a
release occur. Samples of rat stomach, muscle and fur, along with baseline and consumer reference groups were collected
in plateau forest and coastal terrace for stable isotope analysis during the wet and dry season of 2015/16. Stomach
analysis revealed an omnivorous diet, with reproductive parts (flowers, fruits and small seeds) of plants significantly
dominating the invertebrate component. One reptile was found in a single gut, the introduced blind snake (Indothyphlops
bramini) but no birds were detected in stomach contents. Stable isotope analysis showed an omnivorous to predatory
role compared with stomach analysis, but no association with nesting seabird sources. The effect of habitat and season
did not result in major diet shifts, with rats consuming items that primarily followed the C3 pathway. Omnivory was
predominant in plateau forest and carnivory dominated the coastal terraces, while trophic niche width broadened on the
coastal terraces. Homogeneity of diet across habitat and season suggests persistent plant and invertebrate resources may
satisfy nutritional requirements through opportunity or necessity year-round. Little evidence of significant dietary overlap
was shown with feral cats based on stomach data from previous diet studies. Further investigation into the diets and
relative abundance of rats over time is required to reliably gauge their impacts on vulnerable species and communities on
Christmas Island, to justify future rat control actions in the wake of feral cat eradication.
The prospects for biological control of Rubus niveus in the
Galapagos Islands
K. Pollard, D. Kurose, A. Buddie and C. Ellison
K. Pollard, D. Kurose, A. Buddie and C. Ellison
Invasive Species Management, CABI-UK, Bakeham Lane, Egham, Surrey TW20 9TY, UK. <k.pollard@cabi.org>
Following its introduction for its sweet edible fruit in the 1970s, Rubus niveus, native to Indochina, has become one of
the worst invasive weeds on the Galapagos archipelago. It invades open vegetation, scrub and forests where it can grow
to 4 m in height and form dense, impenetrable thickets. As a result, R. niveus can out-compete native flora and decrease
biodiversity; the endemic Scalesia pedunculata forest on Santa Cruz Island is currently threatened by R. niveus. It is
also a serious problem for agricultural land where it increases the cost of weed control and may render land unsuitable
for cultivation. Current control methods are based on mechanical removal followed by chemical control. However, due
to the long-lived seed bank and rapid growth of R. niveus, this has to be repeated, which is both labour intensive and
costly. Classical biological control using coevolved, host-specific natural enemies from the native range of an invasive
species can be an economic and self-sustaining method of weed control. It is important to select natural enemies for
further evaluation that are best-adapted to populations of R. niveus on the Galapagos Islands. The results of on-going
molecular research undertaken to determine which area in the native range the archipelago biotype originated from, will
be presented. In addition, the results of a desk-based analysis and preliminary natural enemy surveys in India and China,
which have revealed a suite of insects and fungal pathogens that target R. niveus, will be discussed.
Pollard, K.; D. Kurose, A. Buddie and C. Ellison. The prospects for biological control of Rubus niveus in the Galapagos Islands
720
Pott, M.; E. Hagen, P. Martínez and M. Díaz. A tool for biodiversity conservation within Chile: renewed interest in island
eradications sparked by successful European rabbit (Oryctolagus cuniculus) eradication
A tool for biodiversity conservation within Chile: renewed interest in
island eradications sparked by successful European rabbit
(Oryctolagus cuniculus) eradication
M. Pott, E. Hagen, P. Martínez and M. Díaz
M. Pott, E. Hagen, P. Martínez and M. Díaz
2100 Delaware Avenue, Suite 1, Santa Cruz, California 95060, USA. <mpott@islandconservation.org>
Choros Island (301 ha) and Chañaral Island (517 ha) are the two largest islands which make up the Humboldt Penguin
National Reserve (RNPH), in northern Chile, within Chile’s National Protected Areas System (SNASPE) designed
to protect the rich resources of the Humboldt Current. The European rabbit (Oryctolagus cuniculus) was introduced
to both islands in the mid-20th Century, triggering erosion and negative impacts on native vegetation and two seabird
species endemic to the Humboldt Current: the Humboldt penguin (Spheniscus humboldti) and the Peruvian diving petrel
(Pelecanoides garnotii). Island Conservation and CONAF (Corporación Nacional Forestal; Chile’s National Forestry
Corporation and RNPH manager) initiated the eradication of European rabbits from Choros Island in 2013 – the first
eradication of invasive species from a Chilean island in a decade. The project was successfully confirmed in 2014,
prompting the partnership to pursue ecological restoration of Chanaral Island in 2015, beginning with the removal of
invasive rabbits. Utilising lessons learnt from work on Choros Island, the eradication on Chañaral Island was initiated
in 2016 and is currently in a monitoring phase. The opportunity to remove all invasive vertebrates from the entirety of a
protected area – RNPH – has built confidence in planning, implementation and monitoring among government officials
and local stakeholders (ecotourism operators) and has facilitated increased momentum in Chile for island biodiversity
conservation through the eradication of invasive species. As a result, CONAF seeks to achieve greater biodiversity
conservation within other islands in the SNASPE, such as the Juan Fernández Archipelago National Park (PNAJF),
representing unique ecosystems severely affected by multiple invasive vertebrates.
Finders keepers? Discovering and securing the rare species
P. Ragen, S. Pandoo, V. Bachraz,
rediscovered in weeded restoration plots
Z. Jhumka and J. Mauremooto
P. Ragen, S. Pandoo, V. Bachraz, Z. Jhumka and J. Mauremooto
InSpiral Pathways, 23 Southside, Congresbury, Bristol, Avon BS495BS, UK. <parmananda.ragen@gmail.com>
Objectives: To document the role of the expansion of weeded areas in increasing the discovery of threatened plants and
how this process can be integrated into systematic plant species recovery programmes. Methodology: Mauritius hosts
some of the most threatened plant species in the world. More than 80% of its remaining 273 endemic plant species are
considered to be threatened. The expansion in weeded areas in recent years has resulted in a number of species rediscoveries
and increases in the known wild populations for other species. Written and verbal records of species rediscovery from
different agencies are consolidated. Results: Results are summarised by numbers of species and number of individuals
rediscovered, location of these discoveries and the fate of the discovered individuals. Some rediscovered individuals have
been successfully utilised for their germplasm for propagation and subsequent reintroduction. However, in most cases,
this process has not been systematic. Conclusions: Finding previously unrecorded species and populations is clearly a
positive thing. However, there are challenges. Weeding, although essential for the long-term health of Mauritian native
forest, can cause short-term negative effects for rare plants and other threatened taxa. Therefore, it is important to develop
weeding approaches that take the requirements of rare plants into account, for example leaving certain exotic species
which act as substrates to epiphytic plants, and gradually removing species in the vicinity of rare plants so that they are
not exposed to a sudden change in micro-climate. These actions have been implemented in certain instances but have been
neglected in others, chiefly because of the lack of knowledge of labourers who are not trained to recognise rare native
plants. Rediscovery does not mean that the species concerned are ‘out of the woods’ so it must be considered to be a part
of an overall rare species recovery plan.
Impacts and control of invasive species: trading off actions
M. Roberts, W. Cresswell and N. Hanley
M. Roberts, W. Cresswell and N. Hanley
James Hutton Institute, Craigiebuckler, Aberdeen, Scotland. <Michaela.Roberts@hutton.ac.uk>
Environmental conservation is chronically underfunded, as a result of both an absolute shortfall in funding, and poor
funding prioritisation. Control of impacts of invasive species on native ecosystems is recognised as of high global
conservation priority, but also requires significant economic investment. Improving prioritisation of invasive species
control options, and identifying alternative funding sources, would therefore greatly improve efficiency in mitigating
degradation caused by invasive species. We incorporate ecological, economic, and social considerations to prioritise
options for control of invasive grazing species on Bonaire, Caribbean Netherlands. We estimate impacts of control of
terrestrial invasive species on the dry-forest, and on the coral reef, linked by changes in terrestrial sedimentation rates. To
address absolute shortfalls in funding, we estimate willingness of SCUBA divers to pay for terrestrial invasive species
control. We find significant negative relationships of donkey density with vegetation ground cover; and a significant
positive relationship of ground cover on the watershed with coral cover at depths below 10m. Using these models we
estimate the impacts on coral cover of strategies to control grazing, including fencing and eradication. Cost curves for
each strategy indicated that fencing of watersheds to exclude grazers presented the most cost effective solution within
a 50-year time frame. We conducted choice experiments with SCUBA divers to estimate willingness to pay for control
of terrestrial invasive species, where this would improve reef health. Willingness to pay exceeded the total costs of both
fencing and eradication. We illustrate that control of terrestrial invasive species can lead to benefits in both terrestrial and
marine ecosystems, and that funding for such projects may be possible via marine stakeholders. The combination of both
terrestrial and marine considerations into invasive species control can greatly improve efficiency, while ensuring funding
is allocated to address all threats to ecosystems under direct use.
Ragen, P.; S. Pandoo, V. Bachraz, Z. Jhumka and J. Mauremooto. Finders keepers? Discovering and securing the rare species rediscovered in weeded restoration plots
Roberts, M.; W. Cresswell and N. Hanley. Impacts and control of invasive species: trading off actions
721
Rogers, A.; J. Shaw and S. Kark. Incorporating interaction networks into conservation: Tasmania as a case study
Incorporating interaction networks into conservation: Tasmania as a
case study
A. Rogers, J. Shaw and S. Kark
A. Rogers, J. Shaw and S. Kark
Biological Sciences, University of Queensland, St Lucia, Queensland 4072, Australia. <a.munro.rogers@gmail.com>
Quantifying the direct impacts of invasive species requires time and resources which are not always available. In systems
with limited information, qualitative interaction networks provide a method in which to explore the potential interactions
between species at the community level. In Tasmania, hollow-breeding bird communities have been invaded by five
hollow-nesting birds and one hollow-using, predatory marsupial, contributing to the decline in populations of several
threatened species. While some interactions between native and invasive species on the island have been well documented,
little information exists on the impact of most invasive species across the island. The aim of this research is to develop a
model which quantifies the likely competitive interactions between hollow-breeding species across Tasmania in order to
determine the potential impacts of unstudied invasive species. Hollow-breeding communities are an ideal community in
which to study competitive interactions because there is direct competition between species over shared resources and it
is possible to include all species in the community. Here we use species traits to model individual species breeding niche
space, and use a metric of niche overlap between species to build qualitative networks representing potential competitive
interactions for entire hollow-breeding communities. This method highlights known impacts of established invasive
species and can be used to model the potential interactions of alien species present but not yet established.
Rojas-Sandoval, J.; P. Acevedo-Rodríguez, M. Datiles, S. Dube, H. Diaz-Soltero, L. Charles, G. Richards, M. Angel Duenas, D. Simpson, E. Ventosa-Febles, J. Ackerman, F. Areces-Berazain,
M. Caraballo-Ortiz, A. Carvajal-Vélez, J. Chabert-Llompart, S. Kaufman, J. Thompson and J. Vélez-Gavilán. Invasive plants of the Caribbean: application of herbarium collections to protect a
regional biodiversity hotspot
Invasive plants of the Caribbean: application of herbarium collections
to protect a regional biodiversity hotspot
J. Rojas-Sandoval, P. Acevedo-Rodríguez, M. Datiles, S. Dube, H. Diaz-Soltero, L. Charles, G. Richards,
M. Angel Duenas, D. Simpson, E. Ventosa-Febles, J. Ackerman, F. Areces-Berazain, M. Caraballo-Ortiz,
A. Carvajal-Vélez, J. Chabert-Llompart, S. Kaufman, J. Thompson and J. Vélez-Gavilán
Compendium Office, CABI, Nosworthy Way, Wallingford, Oxfordshire OX10 6RP, UK. <julirs07@gmail.com>
The Caribbean Islands represents a biodiversity hotspot with over 650 Critically Endangered or Endangered
species. Collation and dissemination of knowledge is a requisite to address the problem of invasive species, a major
driving force of species extinction with many other serious socio-economic impacts. This poster describes how, building
on the keystone work, ‘Catalogue of seed plants of the West Indies’ and analysis of over 14,300 georeferenced herbarium
accessions at the Smithsonian Institution, a project has collated data on over 570 invasive plant species prioritised from
1,879 plants identified as exotic to the region. Expert authors were selected to compile datasheets on each species from
records in the herbarium and in scientific journal articles and authoritative databases. The datasheets were peer reviewed
and submitted to CABI for final style edits and publication in the Invasive Species Compendium (ISC), a scientific
knowledgebase with global coverage and reach. As of February 2017, 253 of 417 completed datasheets have been
published. Inclusion in the ISC provides an Open Access platform for comparison with other taxa and geographic regions
within a sustainable programme where information will be updated. Data are collated and presented with particular
focus on risk assessment, management of pathways, public awareness, policy development, identification, detection, and
options for control. Future work will include in-country gap analyses through consultation and comparison with locally
compiled invasive species lists.
J. Rojas-Sandoval, P. Acevedo-Rodríguez, M. Datiles, S. Dube, H. Diaz-Soltero, L. Charles, G. Richards, M. Angel Duenas, D. Simpson, E. Ventosa-Febles, J. Ackerman, F. Areces-Berazain,
M. Caraballo-Ortiz, A. Carvajal-Vélez, J. Chabert-Llompart, S. Kaufman, J. Thompson and J. Vélez-Gavilán
Ruffino, L. and T. Cornulier. Can large database mining inform invasive non-native species management on islands?...........................
Can large database mining inform invasive non-native species
management on islands?
L. Ruffino and T. Cornulier
L. Ruffino and T. Cornulier
University of Aberdeen, Aberdeen, AB24 3UU, UK. <Lise.Ruffino@jncc.gov.uk>
Global databases, including the IUCN Red List database, the Global Invasive Species Database, and the Threatened Island
Biodiversity Database represent invaluable assets for investigating global patterns of extinction risk in insular vertebrates
and target priority islands and species for conservation or eradication management. In view of the growing number of
studies mining these databases to inform global conservation priorities, we ask two key questions: 1) what questions can
these data most effectively address or not address; and 2) are the recommendations issued useful to practitioners and
policy makers? Here, we critically assess the quality of the evidence used for quantifying global impacts of invasive
non-native species on island vertebrates, and the methodology used in analyses of large publicly-available datasets. We
provide recommendations on how to overcome limitations identified in the data, their processing and reporting, and
suggest perspectives to address critical knowledge gaps.
722
Russell, P. and S. Weaver. Management of numerous introduced plants on Matiu (Somes Island), Wellington, New Zealand
Management of numerous introduced plants on Matiu (Somes Island),
Wellington, New Zealand
P. Russell and S. Weaver
P. Russell and S. Weaver
Scarhead House, Glenfarg, Perth, PH2 9QG UK. <peter@aotearoabiodiversity.co.nz>
Matiu (Somes Island) is a 24.9 ha island in Wellington Harbour, New Zealand. The island has been free of introduced wild
mammals since the late 1980s and provides a valuable opportunity to restore coastal forest ecosystems, including biota now
rare or extinct on the mainland. Despite being only c. 2.5 km from the mainland, experience to date suggests invasion by
invasive plants from the mainland is generally unlikely, although this situation may change in future. Restoration planting
began in 1981. Major efforts to manage numerous plants known to threaten the restoration and protection of the island’s
native biodiversity (“weeds”) began in 1998 and were initially somewhat ad hoc. Due partly to the retention of skilled
personnel the island’s weed management strategy has been refined greatly since 2007, including: enhancing biosecurity
procedures pertaining to weeds; developing a thorough, systematic and regular approach to surveying; considering all
introduced plants and implementing a precautionary approach (erring on the side of controlling plants that may be a threat,
especially if rare and easy to kill); upskilling personnel; more strategically dividing volunteer, staff and contract labour
and prioritising control work (including placing greater emphasis on early detection and nascent foci); and increasing the
diversity of the island’s native vegetation to enhance its resistance to weed invasion. Of 129 plants of concern to date, 73
(57%) are rated as posing a very high, high or moderate threat and 53 (73%) of those are now considered rare, possibly
eradicated or probably eradicated. Major progress has also been made in most other areas of the weed management
strategy, although some tenacious weeds remain a challenge. Lessons learnt on Matiu during the last 20 years may be
applicable to other sites, including larger ones; sites with multiple land uses, owners and management regimes; and sites
with greater chances of weed invasion.
Schmack, J.M.; M.C. Barron, D.F. Ward and J.R. Beggs. Managing Vespula wasp invasion in New Zealand
Managing Vespula wasp invasion in New Zealand
J.M. Schmack, M.C. Barron,
D.F. Ward and J.R. Beggs
J.M. Schmack, M.C. Barron, D.F. Ward and J.R. Beggs
School of Biological Sciences, University of Auckland, Auckland, New Zealand. <j.schmack@auckland.ac.nz>
Introduced Vespula wasps cause severe problems to New Zealand ecosystems. Though vespulid wasps have successfully
invaded most of New Zealand’s offshore islands, little is known about their abundance and population development
on those islands. Anecdotal observations suggest three offshore islands in the Hauraki Gulf and on the coast of the
Coromandel (Little Barrier Island, Korapuki and Tiritiri Matangi) have become vespula wasp-free following successful
mammal eradication. This study aims to investigate the drivers of successful wasp suppression and the prevention
of reinvasion. Wasp monitoring will be conducted on different offshore islands along the northern east coast of New
Zealand’s North Island to measure the relative abundance of wasps and to collect a database on the island’s environmental
parameters. The combination of wasp trapping and a molecular analysis of paternity levels will allow us to estimate nest
densities on offshore islands. The proposed study is novel because it will use a combination of methods (field based and
molecular) to assess the density of vespula wasps in low-density areas (not beech forest). This database will also serve as a
baseline for future investigations on pest dispersal and colonisation processes. It is crucial that we improve understanding
of how different factors influence the development of wasp colonies to elaborate efficient pest control plans. The efficiency
of five novel control methods will be forecasted using population modelling on colony and landscape scales.
Sjodin, B.; R. Irvine, G. Howald and M. Russello. Invasive rat colonisation history and movement dynamics in Haida Gwaii
Invasive rat colonisation history and movement dynamics in Haida Gwaii
B. Sjodin, R. Irvine, G. Howald and M. Russello
B. Sjodin, R. Irvine, G. Howald and M. Russello
University of British Columbia, 1531 Appleridge Rd, Kelowna,
British Columbia V1W3A5, Canada. <Bryson.Sjodin@ubc.ca>
The brown rat (Rattus norvegicus) and black rat (R. rattus) are among the most invasive species worldwide with
distributions encompassing every inhabited continent. Through predation and competition, invasions from both species
have caused range contractions, local extirpation, and extinctions, resulting in reduced biodiversity. On Haida Gwaii,
invasive rats have been implicated in population declines of six seabird species. Eradications were conducted on several
islands where important nesting sites for sea-birds exist. On the Bischof Islands, reappearance of rats post-eradication
has been observed. The objectives of this research are to investigate population history and movement dynamics of
invasive rats in Haida Gwaii. Presently, 551 brown and black rats have been sampled from eighteen islands, collected
from 2008 to 2016. Pre- and post-eradication samples were collected from the Bischofs allowing for an explicit evaluation
of re-emergence versus re-colonisation in these locations. Genomic DNA was extracted from ear samples and used to
construct double digest restriction site associated DNA sequencing libraries and sequenced using the Illumina HiSeq2500
PE125 platform. Single nucleotide polymorphisms (SNPs) were identified, genotyped, and used to assign individuals
to species using a Bayesian clustering approach. The two species were then separated, and SNPs were re-identified
and genotyped for further analysis. Resulting SNP data will be analysed using a series of population genetic and
spatially-explicit analyses to determine the source of re-established populations and quantify the extent and direction of
gene flow throughout the system. Genotypic data are being collected such that they offer full connectibility to a global
SNP database of brown rats to infer potential sources of the populations in Haida Gwaii. Results of these analyses will
help facilitate future eradications and provide useful insights to prevent the spread of rats elsewhere within the system.
723
Skei, B. Garden cans and river rafts – equipped to approach invasive freshwater fish
Garden cans and river rafts – equipped to approach invasive
freshwater fish
B. Skei
B. Skei
Environmental and Biosecurity Measures, Norwegian Veterinary Institute, Nordengbakkan 47,
Levanger, +47 7602 Norway. <bjorn.skei@gmail.com>
How can environmental and fishery managers benefit from a natural toxin when aiming to maintain healthy native aquatic
ecosystems? Rotenone is the only substance on the EU Biocides Regulation 528/2012 product-type 17 (piscicides) list
and considered one of the most environmentally benign toxicants available for eradication of invasive fish. The substance
is distributed in the formulated product CFT Legumine (CFT-L). In the wake of a CFT-L treatment, rotenone persistence
in natural waters differs from a few days to several weeks depending on the season. Unless all parts of a large water body
or catchment can be treated more or less simultaneously, the breakdown of rotenone may allow fish migration back into
previously treated areas, i.e. undermining a successful treatment operation. When aiming for treatment of invasive alien
species against a complex hydrogeological backdrop, standard tools are often pushed towards customised equipment. This
poster presents equipment and techniques used in CFT-L treatment of diverse habitats such as groundwater entries,
ponds and lakes, streams and rivers, tarns and marshlands, opening a toolbox containing garden cans, peristaltic pumps,
backpack pumps and river rafts.
Springer, K. What happens after the helicopters have gone – assessing post-eradication changes on Macquarie Island
What happens after the helicopters have gone – assessing
post-eradication changes on Macquarie Island
K. Springer
K. Springer
16 Rinaldi Ave, The Pines Beach, North Canterbury 7630, New Zealand. <keith.springer@gmail.com>
In 2014 an eradication operation targeting house mice (Mus musculus), black rats (Rattus rattus) and European rabbits
(Oryctolagus cuniculus) on sub-Antarctic Macquarie Island was successfully concluded. Monitoring of outcomes since
that time has been sporadic and some is partly anecdotal, however the changes apparent on Macquarie Island in the
absence of pest species are nonetheless considerable and are significant indicators of the progressive recovery of an island
ecosystem. Vegetation changes following cessation of rabbit grazing are the most visually dramatic and widespread and
are demonstrated partly by use of photo-points. Censuses of some seabird and invertebrate species have documented
changing trends in island populations. Further changes can be expected for decades to come although some changes will
be influenced by changing climatic conditions.
Thibault, M.; F. Brescia and M. Barbet-Massin. Predicting the distribution of island invader bird species under climate change
Predicting the distribution of island invader bird species under
climate change
M. Thibault, F. Brescia and M. Barbet-Massin
M. Thibault, F. Brescia and M. Barbet-Massin
Institut Agronomique Néo-Calédonien (IAC), Equipe ARBOREAL (AgricultuRe BiOdiversité Et vAlorisation),
BP 73, 98890 Païta, New Caledonia. <thibault@iac.nc>
The red-vented bulbul (Pycnonotus cafer), the common myna (Acridotheres tristis) and the red-whiskered bulbul
(P. jocosus) are three passerine bird species native to the Indian subcontinent that were transported to islands from the
early 1900s. Nowadays, the common myna is considered established in 20 island territories, the red-vented bulbul in 11
(32 islands) and the red-whiskered bulbul established in four island territories. Considering that perturbations associated
with human activities will continue to increase during the 21st Century, leading to unprecedented species transportation
rates, understanding potential climatic ranges of these species could be crucial. Moreover, predicting future range shifts
under various climate change scenarios could be very useful in order to better inform management strategies. This is
particularly true for birds as climate is often assumed to be one of the main drivers of the distribution of this taxon at
large spatial scale. Here, we used eight species distribution models, five global circulation models and four representative
concentration pathways using presence data from both the native and alien ranges of the three species. The objectives of
this study were to i) assess the potential invasion risk of the red-vented bulbul, common myna and red-whiskered bulbul;
ii) highlight priority locations for the management of these species and prevention of their introduction; and iii) explore
the likely influence of climate change on the future climatic range of each. Our world climate suitability maps for each
species predict a latitudinal expansion of climatic range. Then, our projections highlight three major potential climatic
pathways for the establishment of the three species around the coasts of Northern Brazil, Guinea Gulf and North-West of
the United States.
724
Vanderwoude, C.; S. Boudjelas, P. Andreozzi, P. Cowan and J. Wainiqolo. Biosecurity Plan for invasive ants in the Pacific
C. Vanderwoude, S. Boudjelas,
P. Andreozzi, P. Cowan and
J. Wainiqolo
Biosecurity Plan for invasive ants in the Pacific
C. Vanderwoude, S. Boudjelas, P. Andreozzi, P. Cowan and J. Wainiqolo
Invasive Species Programme, Secretariat of the Pacific Regional Environment Programme,
PO Box 240, Apia, Samoa. <cas@littlefireants.com>
Invasive ants are very adept hitchhikers and invaders of novel ecosystems. They have the ability to move through a wide
range of international trade pathways. Once established, invasive ants are very difficult to eradicate or control. Their
impacts are felt across many sectors including, agriculture, horticulture, trade, tourism, human health and the
environment. Pacific islands are un-adapted to the presence of invasive ants and largely devoid of native ant species. Their
impacts are often more far-reaching than at other locations, threatening not only delicate and complex island ecosystems,
but the livelihoods and wellbeing of island communities. In the face of climate change, invasive ants will further reduce
the climate resilience and food security of subsistence economies. A best practice integrated biosecurity system is needed
to prevent the entry and establishment of these species as well as mitigate impacts caused by priority invasive ants
currently present in the region. We recommend a regional approach to invasive ant biosecurity be established, which
includes the essential elements of prevention, early detection, rapid response, ongoing management, capacity building,
outreach and research. This system should operate at island, national and regional scales.
Varnham, K.; E. Radford, S. Busuttil, C. Forbes, E. Gibbs-Williams and G. Gerber. Rat control to protect the Turks and Caicos rock iguana:
monitoring and responding to rat activity on a Caribbean island Nature Reserve
Rat control to protect the Turks and Caicos rock iguana: monitoring and
responding to rat activity on a Caribbean island Nature Reserve
K. Varnham, E. Radford, S. Busuttil, C. Forbes, E. Gibbs-Williams and G. Gerber
K. Varnham, E. Radford, S. Busuttil, C. Forbes, E. Gibbs-Williams and G. Gerber
Nature Recovery Unit, RSPB, The Lodge, Sandy, Beds, SG19 2DL. UK. <karen.varnham@rspb.org.uk>
A significant proportion of the global population of the Critically Endangered Turks & Caicos rock iguana (Cyclura
carinata) is found on the small 43 ha island of Little Water Cay, which is managed as a nature reserve by the Turks
& Caicos National Trust (TCNT). Black rats (Rattus rattus) and feral cats (Felis catus) are also found on the island,
which is connected by sand bars to two larger islands and within black rat swimming distance of the large inhabited
island of Providenciales. While rat eradication is not currently thought sustainable, a control programme began in 2015
aiming to control rats to zero/low density, reducing predation pressure on young iguanas. The baiting programme uses
the first-generation anticoagulant rodenticide diphacinone set in bait boxes on a 50 m grid across the island. Following
three weeks of baiting in November 2015 anecdotal changes were observed; increased sightings of young iguanas,
nesting least terns (Sternula antillarum) on the sand bar and presence of a juvenile Antillean nighthawk (Chordeiles
gundlachii). However, a second scheduled baiting round in November 2016 showed that rats were once more found
across the entire island. We therefore devised a monitoring system to observe the speed and distribution of the influx of
rats, predicted to walk across from the adjoining cays and/or swim from Providenciales. After the second baiting season
non-toxic chocolate wax monitoring blocks were set in 20 bait stations across the island and checked weekly for signs
of rat activity. These data will inform the timing and duration of future rat control undertaken on the island, allowing us
to maximise the conservation benefits to iguanas while minimising the amount of rodenticide used, and thus non-target
impacts. TCNT staff have been trained in rat control and monitoring techniques and now lead this project to reduce the
impact of invasive rats on this important species.
Varnham, K.; S. Thomas, L. Bambini, S. Havery and L. Lock. An innovative programme to protect the UK’s seabird islands
An innovative programme to protect the UK’s seabird islands
K. Varnham, S. Thomas, L. Bambini,
S. Havery and L. Lock
K. Varnham, S. Thomas, L. Bambini, S. Havery and L. Lock
Nature Recovery Unit, RSPB, The Lodge, Sandy, Beds, SG19 2DL, UK. <karen.varnham@rspb.org.uk>
The UK supports globally significant populations of seabirds, including 80% of the world’s Manx shearwaters (Puffinus
puffinus) and almost 60% of northern gannets (Morus bassanus), with breeding populations mostly restricted to offshore
islands. However, many of these islands have one or more invasive non-native mammals present which negatively impact
seabirds as well as many other native species. In 2013 the RSPB’s innovative Seabird Island Restoration Project was
established to protect these important islands using three key approaches. Firstly, we have developed a best practice
toolkit for UK ground-based rat eradication projects, to be launched in 2017. This toolkit is based on international
standards but tailored to the UK environmental, legal and social situation, consisting of technical advice documents
on planning and carrying out eradication, biosecurity and incursion response work, as well as templates and series of
worked examples. We have also collaborated on a prioritisation exercise to identify the UK islands where the greatest
conservation gains can be made through eradication of invasive non-native mammals (eradication priorities) and where
the greatest losses would be expected to occur were brown rats (Rattus norvegicus) to arrive on currently rat-free
islands (biosecurity priorities). Finally, we are building UK capacity in island restoration through supporting UK-based
conservation organisations, offering training in biosecurity, safe and effective rodenticide use, and incursion response
planning, as well as writing, and supporting others to write, biosecurity plans, feasibility studies and operational plans. We
have supported and trained two incursion response teams, one in south-west England and one in north Scotland, and plan
to extend this network UK-wide. We believe this combination of working at the sites where the greatest conservation
gains can be made, with well-trained people following tailored best practice guidelines offers the best chance to protect
the UK’s iconic seabird island heritage.
725
Veatch, S.D. Prioritising islands for the eradication of invasive vertebrates in the Arctic
Prioritising islands for the eradication of invasive vertebrates
in the Arctic
S.D. Veatch
S.D. Veatch
American University, 4400 Massachusetts Avenue, N.W., Washington, D.C. 20016, USA. <sarah.d.veatch@gmail.com>
As human activity increases and climate warms, invasive alien species pose a serious, growing threat in the Arctic to
native biodiversity, ecosystems, and inhabitants, particularly those of Arctic island ecosystems. Consequently, Arctic
states have recognised the need to eradicate invasive alien species from Arctic island ecosystems. The Arctic Council – an
intergovernmental forum comprised of eight countries and six Permanent Participants that represent Arctic indigenous
peoples – has defined a collective priority for upcoming action in the Arctic Invasive Alien Species (ARIAS) Strategy and
Action Plan: “actively facilitate the eradication of invasive alien species from island ecosystems throughout the Arctic, as
well as the recovery of native island species and habitats that have been impacted by invasive alien species”. Prioritising
islands for eradication activity is both necessary and strategically important in order to achieve this goal with limited
resources across multiple jurisdictional authorities. This paper will explore the application of a study published in
Conservation Biology (“Prioritising islands for the eradication of invasive vertebrates in the United Kingdom overseas
territories”) to the development of a prioritisation schema for the eradication of terrestrial invasive vertebrates on Arctic
islands. The paper will provide a summary of key findings, including the identification of relevant data gaps; a proposed
Arctic island prioritisation schema for the eradication of terrestrial invasive vertebrates; and a summary of further needs
for input from scientific and policy perspectives. These findings will be applied to the ARIAS Strategy and Action Plan
Steering Committee’s efforts to develop a plan for the eradication of invasive alien species from Arctic island ecosystems.
Veitch, C.R. Changes in forest passerine numbers on Hauturu following rat eradication
C.R. Veitch
Changes in forest passerine numbers on Hauturu following
rat eradication
C.R. Veitch
48 Manse Road, Papakura, New Zealand. <dveitch@kiwilink.net.nz>
Passerines were monitored on Hauturu (Little Barrier Island) over 15 years (1975–89) spanning the period (1976–80)
when feral cats were eradicated from the island and again for the period 2013–2017. All birds seen and heard were
recorded while walking three transects representing an altitudinal range from near sea level to approximately 550 m above
sea level. Analysis of variance statistics were used to test for differences in bird numbers between transects and between
years. Bird species were examined by transect to test for changes in numbers over time. Following cat eradication three
species had increased on some transects, and two species had decreased on some transects, but it was difficult to attribute
changes in bird numbers to the one cause which we were able to study: reduced cat numbers. Following rat (Rattus
exulans) eradication in 2004 there have been significant increases and decreases of forest dwelling passerines. Field work
for this study was completed in March 2017 and the data have yet to be analysed in detail.
Webber, B.; K. Uyehara, T. Luxner and D. Dewey. Habitat features that influence predation of endangered Hawaiian common
gallinule nests by invasive vertebrates in Hanalei and Huleia National Wildlife Refuges
Habitat features that influence predation of endangered Hawaiian
common gallinule nests by invasive vertebrates in Hanalei
and Huleia National Wildlife Refuges
B. Webber, K. Uyehara, T. Luxner and D. Dewey
B. Webber, K. Uyehara, T. Luxner and D. Dewey
Kauai National Wildlife Refuge Complex, U.S. Fish and Wildlife Service, PO Box 1128,
Kilauea, Hawaii 96754, USA. <brynwebber@gmail.com>
Hanalei and Huleia National Wildlife Refuges (NWRs) on the island of Kauai in Hawaii are designated as core wetland
areas essential to the recovery of five endangered Hawaiian waterbird species. These two sites support approximately
50% of the endangered Hawaiian common gallinule (Alae ula, Gallinula galeata sandvicensis) population state-wide. On
Hanalei NWR, taro (Colocasia esculenta) farming provides dense emergent aquatic vegetation needed for breeding
gallinules, but previous research suggests that dikes, water drawdowns, and harvested fields may increase access by
introduced mammalian predators. Although these studies documented egg predation, researchers were unable to determine
nest fates for 25% of the nests using observer-based methods. In this pilot study, we evaluated the use of remote motion
detection cameras as a method to determine gallinule nest fates and elucidated factors related to predation events through
the early brooding phase. We predicted that taro farming practices influence predation by invasive vertebrates (e.g. feral
cats (Felis catus), rats (Rattus spp.)) and negatively affect gallinule nest success in taro fields, when compared to managed
wetland units that have fewer dikes, suitable vegetative cover, and stable water levels. Higher gallinule nest success in
wetland units, coupled with reliable data regarding drawdowns and predation of nests in taro fields, allows managers to
implement more specific management and monitoring methods to control and reduce access of invasive vertebrates that
prey on endangered gallinule nests in these critically-important wetland, riverine, and agricultural landscapes.
726
Will, D.; T. Hall, M. Khalsa and J. Bruch. Using sUAS to direct trap placement in support of feral cat eradication on islands
Using sUAS to direct trap placement in support of feral cat eradication
on islands
D. Will, T. Hall, M. Khalsa and J. Bruch
D. Will, T. Hall, M. Khalsa and J. Bruch
Island Conservation, 2100 Delaware Ave, Suite 1, Santa Cruz, CA 95060, USA. <dwill@islandconservation.org>
Trap location is one the most important elements in a successful trapping programme and requires specialists that can
analyse microhabitats across a landscape and identify areas of likely cat (Felis catus) presence and key travel routes. This is
particularly true when determining the location of walkthrough trap sets. Existing remote sensing data can help specialists
identify macrohabitats where cat activity is suspected but is not collected at a fine enough resolution to resolve microhabitats
or topographical features where cat activity is likely. Using a case study on Kaho’olawe, Hawaii we evaluate how placing
very high resolution sUAS-derived data in the hands of trapping specialists can be used to direct trap placement reducing
the need for time intensive exploration of the landscape. On Kaho’olawe (11,550 ha), there is considerable need to direct
trap placement because the presence of unexploded ordnance (only 10% of the island is cleared to a depth of four feet
and 69% of the island surface-cleared) poses a significant risk to staff safety and greatly increases project risk and cost. In
this case study, we use traditional remote sensing techniques to select three representative study areas that have limited
UXO concerns and estimated high cat habitat suitability. Each study area is mapped at a resolution of less than 5 cm and
resulting products are reviewed in 2D and 3D by trapping specialist to select suitable trap locations. Trapping specialists
evaluate each study area on foot using their normal protocols to determine trap locations. Finally, we evaluate the efficacy
of sUAS direct trap placement by comparing the sUAS derived trap locations with the ground-truthed locations. The
workflows for collecting, processing and analysing sUAS data that we describe should enable managers to determine if
integrating sUAS into trapping programmes is a cost-effective and efficient way to improve project success.
Wynn, M. and D. Driscoll. Removal of invasive, black rats increases activity levels and population density of Christmas Island’s last remaining endemic reptile
Removal of invasive, black rats increases activity levels and population
density of Christmas Island’s last remaining endemic reptile
M. Wynn and D. Driscoll
M. Wynn and D. Driscoll
The Australian National University, Fenner School of Environment and Society, Canberra,
Australian Capital Territory, Australia. <melissa.wynn@anu.edu.au>
Invasive black rats (Rattus rattus) have been implicated in the extinctions of native species across the globe, particularly
where native fauna are predator-naïve and are within insular island systems. Through the process of introduced disease
and predation, Christmas Island in the Indian Ocean has suffered catastrophic extinctions of four endemic mammals and
four reptiles since the early 1900s. Up until now, the endangered Christmas Island giant gecko (Cyrtodactylus sadleiri) has
resisted extinction, but the interactions of this rainforest-dwelling endemic species with invasive and abundant black rats
are unclear. With the recent onset of feral cat eradication by the Australian Government’s Christmas Island National Park,
a greater understanding of the potential for rats to impact on threatened reptile species is critical. Here we will present
novel findings from a large-scale manipulation experiment to determine the impacts of the removal (using poison bait) of
black rats from primary rainforest areas on Christmas Island, and the consequential behavioural and population responses
of giant geckos. Giant gecko activity levels were found to increase as rat activity dropped, and gecko population density
doubled, from 27 to 62 geckos per hectare, when rats were no longer present in high densities in the rainforest, with the
greatest effect occurring in the dry season, eight weeks after initial baiting. Interestingly, insect and forest bird activity
was also observed to increase with the reduction of rat activity, suggesting the role of the black rat as a predator of other
native forest species. This research will assist in predicting the consequences of increased rat predation on Christmas
Island’s last remaining endemic reptile, helping to guide future management of invasive black rats, and suggests the
urgent need for further research on complex interactions between invasive species and native prey on Christmas Island,
and a multi-species approach to any further predator eradication.
Zhao, C.; J. Li and X. Liu. Effect of Spartina alterniflora invasion on benthic macro-invertebrate communities in Guangxi Zhuang Autonomous Region
Effect of Spartina alterniflora invasion on benthic macro-invertebrate
communities in Guangxi Zhuang Autonomous Region
C. Zhao, J. Li and X. Liu
C. Zhao, J. Li and X. Liu
Chinese Research Academy of Environmental Sciences, Beijing, China. <zhaocy@craes.org.cn>
In order to assess the ecological impacts of Spartina alterniflora invasion in Guangxi Zhuang Autonomous Region,
we analysed communities of macro-invertebrates in different habitats and with different invasion times of Spartina
alterniflora. Results showed that Shannon-Wiener index and Simpson diversity indices differed between the S. alterniflora
wetlands and a mangrove wetland, and macro-invertebrate communities in S. alterniflora habitat mainly differed from
those of mangrove habitats based on the non-metric multidimensional scaling used in this study. Perhaps due to the
invasion of S. alterniflora, the bivalve Glauconome chinensis became the predominant species, leading to a greater
macro-invertebrate biomass in S. alterniflora wetlands than in mangrove wetlands. Species composition, biomass and
diversity of macro-invertebrates were assessed between the different invasive years of Spartina alterniflora including
20 years, five years and one year. Results showed that the community structures of macro-invertebrates were distinctly
different between the 20-year Spartina alterniflora communities and the other two communities. The biomass of
macro-invertebrates decreased with the length of time Spartina alterniflora communities were established. No significant
differences of richness of macro-invertebrates were found among different invasive years (p<0.05). The results also
showed all of these changes of macro-invertebrates at different communities or different invasion time were related to the
density of Spartina alterniflora based on multiple linear regressions.
727
AUTHOR INDEX
Author names and paper or abstract titles
listed in the order of the first-named author.
Author Index
Abrahão, C.R.; J.C. Russell, J.C.R. Silva, F. Ferreira and R.A. Dias. Population assessment of a novel island
invasive: tegu (Salvator merianae) of Fernando de Noronha ................................................................................... 317
Aguirre-Muñoz, A.; F. Méndez-Sánchez, L. Luna-Mendoza, A. Ortiz-Alcaraz, J. Hernández-Montoya,
Y. Bedolla-Guzmán, M. Latofski-Robles, E. Rojas-Mayoral, N. Silva-Estudillo, F. Torres-García,
M. Félix-Lizárraga, A. Fabila-Blanco, A. Hernández-Ríos, E. Bravo-Hernández, F. Solís-Carlos,
C. Jáuregui-García and D. Munguía-Cajigas. Mexico’s progress and commitment to comprehensive island
restoration .................................................................................................................................................................. 704
Algar, D.; M. Johnston and C. Pink. Big island feral cat eradication campaigns: an overview and status
update of two significant examples ........................................................................................................................... 238
Andreozzi, P.C.; R. Griffiths, D. Moverley, J. Wainiqolo, R. Nias, S. Boudjelas, D. Stewart, S. Cranwell,
M. Smith and P. Cowan. The Pacific invasives partnership – a model for regional collaboration on
invasive alien species ................................................................................................................................................ 704
Auld, M.; B. Ayling, L. Bambini, G. Harper, G. Neville, S. Sankey, D.B.A. Thompson and P. Walton.
Safeguarding Orkney’s native wildlife from non-native invasive stoats .................................................................. 244
Balchin, J.R.; D.G. Duncan, G.E. Key and N. Stevens. Biosecurity on St Helena Island – a socially inclusive
model for protecting small island nations from invasive species ............................................................................. 468
Ballantyne, L.; D. Baum, C.W. Bean, J. Long and S. Whitaker. Successful eradication of signal crayfish
Pacifastacus leniusculus using a non-specific biocide in a small isolated water body in Scotland .......................... 443
Bardal, H. Small- and large-scale eradication of invasive fish and fish parasites in freshwater systems in Norway ..... 447
Bedolla-Guzmán, Y.; F. Méndez-Sánchez, A. Aguirre-Muñoz, M. Félix-Lizárraga, A. Fabila-Blanco,
E. Bravo-Hernández, A. Hernández-Ríos, M. Corrales-Sauceda, A. Aguilar-Vargas, A. Aztorga-Ornelas,
F. Solís-Carlos, F. Torres-García, L. Luna-Mendoza, A. Ortiz-Alcaraz, J. Hernández-Montoya,
M. Latofski-Robles, E. Rojas-Mayoral and A. Cárdenas-Tapia. Recovery and current status of seabirds on
the Baja California Pacific Islands, Mexico, following restoration actions .............................................................. 531
Bell, E.A. It’s not all up in the air: the development and use of ground-based rat eradication techniques in the UK ..... 79
Bell, E.A.; M.D. Bell, G. Morgan and L. Morgan. The recovery of seabird populations on Ramsey Island,
Pembrokeshire, Wales, following the 1999/2000 rat eradication .............................................................................. 539
Bell, E.; J. Daltry, F. Mukhida, R. Connor and K. Varnham. The eradication of black rats (Rattus rattus) from
Dog Island, Anguilla, using ground-based techniques .............................................................................................. 162
Bell, E.; K. Floyd, D. Boyle, J. Pearson, P. St Pierre, L. Lock, P. Buckley, S. Mason, R. McCarthy, W. Garratt,
K. Sugar and J. Pearce. The Isles of Scilly seabird restoration project: the eradication of brown rats
(Rattus norvegicus) from the inhabited islands of St Agnes and Gugh, Isles of Scilly .............................................. 88
Bell, P.; H. Nathan and N. Mulgan. ‘Island’ eradication within large landscapes: the remove and protect model ........ 604
Bellis, K.X.T.; R.T. Peet, R.L. Irvine, G. Howald and G.J. Alsop. Beyond biodiversity: the cultural context of
invasive species initiatives in Gwaii Haanas............................................................................................................. 494
Bird, J.; J. Shaw, R. Alderman and R. Fuller. A review of monitoring of biodiversity responses to island
invasive species eradications ..................................................................................................................................... 705
Bird, J.P.; K. Varnham, J.D. Shaw and N.D. Holmes. Practical considerations for monitoring invasive
mammal eradication outcomes .................................................................................................................................. 545
Boag, B. and R. Neilson. The potential detrimental impact of the New Zealand flatworm to Scottish islands ............. 356
Bond, A.L.; R.J. Cuthbert, G.T.W. McClelland, T. Churchyard, N. Duffield, S. Havery, J. Kelly, J.L. Lavers,
T. Proud, N. Torr, J.A. Vickery and S. Oppel. Recovery of introduced Pacific rats following a failed
eradication attempt on subtropical Henderson Island, South Pacific Ocean ............................................................. 167
Bond, A.L.; S. O’Keefe, P. Warren and G.T.W. McClelland. Bait colour and moisture do not affect bait
acceptance by introduced Pacific rats (Rattus exulans) at Henderson Island, Pitcairn Islands ................................. 175
Booker, H.; D. Appleton, D. Bullock, R. MacDonald, E. Bell, D. Price, P. Slader, T. Frayling, A. Taylor
and S. Havery. A review of seabird recovery on Lundy Island, England, over a decade following the
eradication of brown and black rats .......................................................................................................................... 705
Boser, C.L. Eradicating invasive ants in conservation areas .......................................................................................... 705
Boser, C.L.; P. Power, A. Little, J. Matos, G.R. Howald, J.M. Randall and S.A. Morrison. Proactive planning
and compliance for a high-priority invasive species rapid response programme ..................................................... 473
Boulton, R.A.; M. Bulgarella, I.E. Ramirez, C.E. Causton and G.E. Heimpel. Management of an invasive
avian parasitic fly in the Galápagos Islands: is biological control a viable option?.................................................. 360
Brazier, M. Big island, small invader: eradicating invasive fish on a national scale ...................................................... 706
Brooke, M.de L. Rat eradication in the Pitcairn Islands, South Pacific: a 25-year perspective ....................................... 95
Broome, K.; D. Brown, K. Brown, E. Murphy, C. Birmingham, C. Golding, P. Corson, A. Cox
and R. Griffiths. House mice on islands: management and lessons from New Zealand ........................................... 100
728
Brown, K.; C.B. Phillips, K. Broome, C. Green, R. Toft and G. Walker. Feasibility of eradicating the large
white butterfly (Pieris brassicae) from New Zealand: Data gathering to inform decisions about the
feasibility of eradication ............................................................................................................................................ 364
Bryce, J. and M. Tonkin. Containment of invasive grey squirrels in Scotland: meeting the challenge ......................... 180
Bunbury, N.; P. Haverson, N. Page, J. Agricole, G. Angell, P. Banville, A. Constance, J. Friedlander, L. Leite,
T. Mahoune, E. Melton-Durup, J. Moumou, K. Raines, J. van de Crommenacker and F. Fleischer-Dogley.
Five eradications, three species, three islands: overview, insights and recommendations from invasive bird
eradications in the Seychelles ................................................................................................................................... 282
Buxton, R. and M. Brooke. Population growth of seabirds after the eradication of introduced mammals .................... 706
Campbell, K.J.; J.R. Saah, P.R. Brown, J. Godwin, F. Gould, G.R. Howald, A. Piaggio, P. Thomas,
D.M. Tompkins, D. Threadgill, J. Delborne, D.M. Kanavy, T. Kuikin, H. Packard, M. Serr and A. Shiels.
A potential new tool for the toolbox: assessing gene drives for eradicating invasive rodent populations.................... 6
Canlas, C.; C. Gever, P. Rosialda, Ma. N. Quibod, P. Buenavente, N. Barbecho, C. Layusa and
M. Day. Assessment of the possible effects of biological control agents of Lantana camara and
Chromolaena odorata in Davao City, Mindanao, Philippines .................................................................................. 706
Capizzi, D.; P. Sposimo, G. Sozio, F. Petrassi, C. Gotti, E. Raganella Pelliccioni and N. Baccetti. Black rat
eradication on Italian islands: planning forward by looking backward ...................................................................... 15
Cárdenas-Calle, M.; J. Pérez-Correa, P. Martinez,, I. Keith, F. Rivera, M. Cornejo, G. Torres, F. Villamar, R.
Zambrano, A. Cárdenas, M. Triviño, L. Troccoli, G. Bigatti, J. Coronel and E. Mora. First report of marine
alien species in mainland Ecuador: threats of invasion in rocky shores ................................................................... 452
Carey, P.W. Simultaneous rat, mouse and rabbit eradication on Bense and Little Bense Islands, Falkland Islands ...... 108
Carter, A.; R. van Dam, S. Barr and D. Peters. Testing auto-dispensing lure pumps for incursion control of
rats with reduced effort on a small, re-invadable island in New Zealand ................................................................. 187
Cecchetti, M.; G. Dell’Omo and B. Massa. Black rat eradication from Linosa Island: work in progress ..................... 707
Cecchetti, M.; L. Nelli, B. Massa and G. Dell’Omo. Effects of cat, rat, and human predation on Scopoli’s
shearwater (Calonectris diomedea) breeding success and nest-site occupancy on Linosa Island ............................ 707
Clubbe, C. Invasive plants: what can be done about this continuing threat to biodiversity? ......................................... 707
Cranwell, S. Partnerships in the restoration of tropical Pacific islands .......................................................................... 708
Da Re, D.; E. Tordoni, Z. Negrín-Pérez , J. M. Fernàndez-Palacios, J. R. Arévalo, R. Otto, D. Rocchini and
G. Bacaro. Modelling invasive plant alien species richness in Tenerife (Canary Islands) using Bayesian
Generalised Linear Spatial Models .......................................................................................................................... 410
del Mar Leza, M.; A. Marqués, C. Herrera, M. Ángel Miranda, M. Ruiz, A. Pou and C. Guerrero. Vespapp:
citizen science to detect the invasive species Vespa velutina. ................................................................................... 708
Djeddour, D.; N. Maczey and C. Pratt. Wild ginger, a beautiful menace to island ecosystems – can a natural
solution be found? ..................................................................................................................................................... 708
Doube, J. Is poisoning rodents a health hazard? ............................................................................................................. 709
Duffy, D.C. and C. Martin. Cooperative natural resource and invasive species management in Hawaiʽi ..................... 497
Duhr, M.; E.N. Flint, S.A. Hunter, R.V. Taylor, B. Flanders, G. Howald and D. Norwood. Control of house
mice preying on adult albatrosses at Midway Atoll National Wildlife Refuge ........................................................... 21
Dyer, M.J.B.; G. Keppel, D. Watling, M. Tuiwawa, S. Vido and H.J. Boehmer. Using expert knowledge and
field surveys to guide management of an invasive alien palm in a Pacific Island lowland rainforest ...................... 417
Fisher, R.N.; J. Niukula, P. Harlow, S. Rasalato, R. Chand, B. Thaman, E. Seniloli, J. Vadada, S. Cranwell,
J. Brown, K. Lovich and N. Thomas-Moko. Community-based conservation and recovery of native
species on Monuriki Island, Fiji ................................................................................................................................ 552
Fisher, S.R.; R.N. Fisher, S. Alcaraz, R. Gallo-Barneto, C. Patino-Martinez, L.F. López Jurado and
C.J. Rochester. Life-history comparisons between the native range and an invasive island population of a
colubrid snake ........................................................................................................................................................... 326
Fleischmann, K.; S. Massy, M. Schmutz, B. Seraphine and J. Millett. When our enemy is our friend: new
approaches to managing alien vegetation in Seychelles catchment forest ............................................................... 709
Floyd, K.; K. Passfield, S. Poncet, B. Myer and J. Lee. Persistence, accuracy and timeliness: finding, mapping
and managing non-native plant species on the island of South Georgia (South Atlantic) ....................................... 424
Fric, J.; T. Dimalexis, V. Goritsas, A. Evangelidis and I. Nikolaou. Eleonora’s falcon (Falco eleonorae)
benefiting from rat eradication – the case of Andros, Greece ................................................................................... 709
Fric, J. and A. Evangelidis. A review of 12 years of rat eradication operations for the conservation of priority
island nesting birds in Greece ................................................................................................................................... 710
Fric, J.; A. Evangelidis, T. Dimalexis, N. Tsiopelas, S. Xirouchakis, C. Kassara and S. Giokas. Improving
nesting habitats for the Eleonora’s falcon and seabirds ............................................................................................ 710
Garden, P.; P. McClelland and K. Broome. The history of the aerial application of rodenticide in New Zealand ......... 114
Genovesi, P. Broadening the context of invasive species eradications ........................................................................... 710
Geraldes, P.; T. Melo, P. Oliveira and V. Paiva. Recovery of Santa Luzia Nature Reserve and translocation of
the globally endangered Raso lark ............................................................................................................................ 711
Geraldes, P.; T. Pipa, N. Oliveira, C. Silva and S. Hervías. Setting-up a predator-free area on a Macaronesian
island using a pest-proof fence .................................................................................................................................. 711
Green, C. Effort required to confirm eradication of an Argentine ant invasion: Tiritiri Matangi Island, New Zealand . 370
729
Griffiths, R.; E. Bell, J. Campbell, P. Cassey, J.G. Ewen, C. Green, L. Joyce, M. Rayner, R. Toy, D. Towns, L.
Wade, R. Walle and C.R. Veitch. Costs and benefits for biodiversity following rat and cat eradication on
Te Hauturu-o-Toi/Little Barrier Island ...................................................................................................................... 558
Griffiths, R.; D. Brown, B. Tershy, W.C. Pitt, R.J. Cuthbert, A. Wegmann, B. Keitt, S. Cranwell and G.
Howald. Successes and failures of rat eradications on tropical islands: a comparative review of eight
recent projects ........................................................................................................................................................... 120
Griffiths, R.; S. Cranwell, D. Derand, T. Ghestemme, D. Will, J. Zito, T. Hall, M. Pott and G. Coulston. Multi
island, multi invasive species eradication in French Polynesia demonstrates economies of scale ........................... 611
Haakonsson, J.; F. Rivera-Milan and E. Radford. Green iguana (Iguana iguana) monitoring and control
efforts on Grand Cayman .......................................................................................................................................... 711
Haber, E.; M. Eppinga, M. Ferreira dos Santos, M. Rietkerk and M. Wassen. Predicting the potential habitat
of the invasive coral vine (Antigonon leptopus) using remote sensing and species distribution modelling ............. 712
Hagen, E.; J. Bonham and K. Campbell. House sparrow eradication attempt on Robinson Crusoe Island, Juan
Fernández Archipelago, Chile ................................................................................................................................... 289
Hammer, S. and J. Russell. The diet of ‘Viking mice’ on Nólsoy, Faroe Islands ........................................................... 712
Hanson, C.C.; T.J. Hall, A.J. DeNicola, S. Silander, B.S. Keitt and K.J. Campbell. Rhesus macaque
eradication to restore the ecological integrity of Desecheo National Wildlife Refuge, Puerto Rico ........................ 249
Harper, G.A.; P. Carr and H. Pitman. Eradicating black rats from the Chagos – working towards the whole
archipelago .................................................................................................................................................................. 26
Harvey-Samuel, T.; K.J. Campbell, M. Edgington and L. Alphey. Trialling gene drives to control invasive
species: what, where and how? ................................................................................................................................. 618
Hernández-Montoya, J.C.; L. Luna-Mendoza, A. Aguirre-Muñoz, F. Méndez-Sánchez, A. Duarte-Canizales,
E. Rojas-Mayoral, S. Hall, Z. Peña-Moreno, S. Figueroa-Flores, D. Cosio-Muriel and
M. Latofski-Robles. Seabird restoration and advances towards the eradication of feral cats on Guadalupe
Island, Mexico ........................................................................................................................................................... 712
Herrera, C.; A. Marqués, V. Colomar and M.M. Leza. Analysis of the secondary nest of the yellow-legged
hornet found in the Balearic Islands reveals its high adaptability to Mediterranean isolated ecosystems ............... 375
Herrera-Giraldo, J.L.; C.E. Figuerola-Hernández, N.D. Holmes, K. Swinnerton, E.N. Bermúdez-Carambot,
J.F. González-Maya and D.A. Gómez-Hoyos. Survival analysis of two endemic lizard species before,
during and after a rat eradication attempt on Desecheo Island, Puerto Rico ............................................................ 191
Holmes, N.D.; B.S. Keitt, D.R. Spatz, D.J. Will, S. Hein, J.C. Russell, P. Genovesi, P.E. Cowan and
B.R. Tershy. Tracking invasive species eradications on islands at a global scale ..................................................... 628
Horn, S.; T. Greene and G. Elliott. Eradication of mice from Antipodes Island, New Zealand ..................................... 131
Horrill, J.C. M.K. Oliver and J. Stubbs Partridge. Lessons on effectiveness and long-term prevention from
broad-scale control of invasive alien species in Scotland’s rivers and lochs ............................................................ 458
Houghton, M.; A. Terauds and J. Shaw. Methods for monitoring invertebrate response to vertebrate eradication ....... 381
Hudin, S. From island studies to mainland management................................................................................................ 713
Hughes, B.J.; R.C. Dickey and S.J. Reynolds. Predation pressures on sooty terns by cats, rats and common
mynas on Ascension Island in the South Atlantic ..................................................................................................... 295
Jesse, W.; J. Ellers, J. Behm and M. Helmus. Differential effects of human impact and habitat type on exotic
and native species diversity on oceanic islands ........................................................................................................ 713
Kaiser-Bunbury, C.N. Restoring plant-pollinator communities: using a network approach to monitor
pollination function ................................................................................................................................................... 568
Kanavy, D. and D. Threadgill. Genetic pest management technologies to control invasive rodents ............................. 713
Keith, I.; J. Carlton and G. Ruiz. A new look at Galapagos fouling communities ......................................................... 714
Keitt, B.; N. Holmes, E. Hagen, G. Howald and K. Poiani. Going to scale: reviewing where we’ve been and
where we need to go in invasive vertebrate eradications .......................................................................................... 633
Kelly, J.; K. Springer, C. Stringer, A. Schofield and T. Glass. Planning processes for eradication of mice on
Gough Island ............................................................................................................................................................. 714
Kennedy, E.S. and K.G. Broome. How do we prevent the obstacles to good island biosecurity from limiting
our eradication ambitions? ........................................................................................................................................ 478
Key, G.E. and N.P. Moore. Tackling invasive non-native species in the UK Overseas Territories ................................ 637
Knapp, D.A.; J.J. Knapp, K.A. Stahlheber and T. Dudley. A little goes a long way when controlling invasive
plants for biodiversity conservation .......................................................................................................................... 643
Knapp, J.; C. Boser, J. Randall, E. O’Byrne and S.A. Morrison. Perils of saving the smallest for the last:
lessons learnt about sequencing eradications on Sant Cruz Island, CA .................................................................... 714
Lago, P.; J.S. Santiago Cabello and K. Varnham. Long term rodent control in Rdum tal-Madonna yelkouan
shearwater colony ...................................................................................................................................................... 196
Lambin, X.; J.C. Horrill and R. Raynor. Achieving large scale, long-term invasive American mink control in
northern Scotland despite short term funding ........................................................................................................... 651
La Morgia, V.; D. Paoloni, P. Aragno and P. Genovesi. Citizens’ attitude towards the removal of grey squirrels
in Italy: what support do we need? ........................................................................................................................... 715
730
Latofski-Robles, M.; F. Méndez-Sánchez, A. Aguirre-Muñoz, C. Jáuregui-García, P. Koleff-Osorio,
A.I. González-Martínez, G. Born-Schmidt, J. Bernal-Stoopen and E. Rendón-Hernández. Mexico’s island
biosecurity programme: collaborative formulation and implementation .................................................................. 484
Lennon, Z.; H. Wittmer and N. Nelson. Computer modelling of complex interstitial spaces to protect endemic
island lizards from invasive mice .............................................................................................................................. 715
Li, J.; C. Zhao and X. Zhao. An integrated physical control method on Spartina alterniflora ..................................... 715
Libeau, M.; R. Pouteau, R. Taputuarai and J.-Y. Meyer. Predicting the risk of plant invasion on islands: the
case of Miconia calvescens in the Marquesas, French Polynesia (South Pacific) .................................................... 716
Lindholm, A. and B. König. The secret life in Switzerland of an island pest, the house mouse .................................... 716
Little, A.; A. Aguirre-Muñoz, G. Seutin, L. Wein, P. Nantel, H. Berlanga, F. Méndez-Sánchez, J. Putnam,
E. Iñigo-Elías and G. Howald. Catalysing conservation of islands through collaboration: a North
American perspective ................................................................................................................................................ 716
L. Luna-Mendoza, A. Aguirre-Muñoz, J.C. Hernández-Montoya, M. Torres-Aguilar, J.S. García-Carreón,
O. Puebla-Hernández, S. Luvianos-Colín, A. Cárdenas-Tapia and F. Méndez-Sánchez. Ten years after
feral goat eradication: the active restoration of plant communities on Guadalupe Island, Mexico .......................... 571
Luxmoore, R.; R. Swann and E. Bell. Canna seabird recovery project: 10 years on ..................................................... 576
Macdonald, N.; G. Nugent, K-A. Edge and J.P. Parkes. Eradication of red deer from Secretary Island, New
Zealand: changing tactics to achieve success ............................................................................................................ 256
Macleod, I.A.; D. Maclennan, R. Raynor, D.B.A. Thompson and S. Whitaker. Large scale eradication of
non-native invasive American mink (Neovison vison) from the Outer Hebrides of Scotland .................................. 261
Maczey, N.; D. Moore, P. González-Moreno and N. Rendell. Introduction of biological control agents against
the European earwig (Forficula auricularia) on the Falkland Islands ...................................................................... 389
Maggs, G.; M.A.C. Nicoll, N. Zuël, D.J. Murrell, J.G. Ewen, C. Ferrière, V. Tatayah, C.G. Jones and
K. Norris. Bridging the research-management gap: using knowledge exchange and stakeholder
engagement to aid decision-making in invasive rat management ............................................................................... 31
Main, C.E.; E. Bell, K. Floyd, J. Tayton, J. Ibbotson, W. Whittington, P.R. Taylor, R. Reid, K. Varnham,
T. Churchyard, L. Bambini, A. Douse, T. Nicolson and G. Campbell. Scaling down (cliffs) to meet the
challenge: the Shiants black rat eradication .............................................................................................................. 138
Martin, A.R.; M.N. Clout, J.C. Russell, C.R. Veitch and C.J. West. Addressing the challenge ......................................xiii
Mauremootoo, J.R.; S. Pandoo, V. Bachraz, I. Buldawoo and N.C. Cole. Invasive species management in
Mauritius: from the reactive to the proactive – the National Invasive Species Management Strategy and its
implementation .......................................................................................................................................................... 503
McHenry, E.; X. Lambin, T. Cornulier and D. Elston. The value of monitoring and the price of uncertainty in
the management of an invasive population ............................................................................................................... 717
Meyer, J.-Y. and M. Fourdrigniez. Islander perceptions of invasive alien species: the role of socio-economy
and culture in small isolated islands of French Polynesia (South Pacific) ................................................................ 510
Millett, J.E.; W. Accouche, J. van de Crommenacker, M.A.J.A. van Dinther, A. de Groene, C.P. Havemann,
T.A. Retief, J. Appoo and R.M. Bristol. Conservation gains and missed opportunities 15 years after rodent
eradications in the Seychelles ................................................................................................................................... 580
Miranda, M.A.; C. Barceló, D. Borràs, A. González, M. Leza and C. Paredes-Esquivel. Invasive arthropods
of ecological, agricultural and health importance recently introduced in the Balearic Islands (Spain) .................... 717
Mohanty, N.P.; G.J. Measey, A. Sachin, G. Selvaraj and K. Vasudevan. Using key-informant surveys to
reliably and rapidly estimate the distributions of multiple insular invasive species ................................................. 717
Moverley, D. Battling invasive species in the Pacific ..................................................................................................... 658
Negrín Pérez, Z.; D. Da Re, M. Bernardos and B. Garrido. Time germination response to temperature and
light conditions in Ulex ............................................................................................................................................ 718
Neville, R.; J.Y. Fujikawa and M. Halabisky. Eradication programmes complicated by long-lived seed banks:
lessons learnt from 15 years of miconia control on O’ahu Island, Hawai’i .............................................................. 430
Oliveira, N.; P. Geraldes, I. Fagundes, P. Oliveira and J. Andrade. Rat eradication from Berlengas Island, Portugal... 718
Oppel, S.; G.T.W. McClelland, J.L. Lavers, T. Churchyard, A. Donaldson, N. Duffield, S. Havery, J. Kelly,
T. Proud, J.C. Russell and A.L. Bond. Seasonal variation in movements and survival of invasive Pacific
rats on sub-tropical Henderson Island: implications for eradication......................................................................... 200
Oppel, S.; S.J. Havery, L. John, L. Bambini, K. Varnham, J. Dawson and E. Radford. Maximising
conservation impact by prioritising islands for biosecurity ...................................................................................... 663
Ortiz-Alcaraz, A.; A. Aguirre-Muñoz, F. Méndez-Sánchez, E. Rojas-Mayoral, F. Solís-Carlos,
B. Rojas-Mayoral, E. Benavides-Ríos, S. Hall, H. Nevins and A. Ortega-Rubio. Ecological restoration of
Socorro Island, Revillagigedo Archipelago, Mexico: the eradication of feral sheep and cats .................................. 267
Palmas, P.; R. Gouyet, T. Ghestemme, A. Matohi, E. Terorohauepa, I. Tauapaohu, C. Blanvillain, J. Zito,
D. Beaune and E. Vidal. Response of an open feral cat population to an intensive control programme for
improving the Critically Endangered Fatu Hiva monarch conservation strategy ..................................................... 718
Palmas, P.; H. Jourdan, E. Bonnaud, F. Rigault, L. Debar, H. De Méringo, E. Bourguet, R. Adjouhgniope and
E. Vidal. Feral cats threaten the outstanding endemic fauna of the New Caledonia biodiversity hotspot:
implications for feral cat management strategy ........................................................................................................ 719
731
Pandoo, S.; P. Ragen, B. Vishnuduth, Z. Jhumka and J. Mauremootoo. Scaling up invasive plant management
for ecosystem restoration in Mauritius: successes and challenges ............................................................................ 719
Parent, C.E.; P. Fisher, W. Jolley, A. Alifano and K.J. Campbell. Assessment of snail exposure to the
anticoagulant rodenticide brodifacoum in the Galapagos Islands ............................................................................. 394
Parish, J. Implementing an early detection programme on Catalina Island: prioritising landscaped grasses ................ 719
Parkes, J.P. Timing aerial baiting for rodent eradications on cool temperate islands: mice on Marion Island................. 36
Parrott, D.; G. Massei, R. Ridley, J. Sandon, M. Lambert, D. Cowan and M. Sutton-Croft. Challenges and
opportunities for lethal and non-lethal management of non-native ungulates on islands: feral pigs, goats
and cows................................. ................................................................................................................................... 720
Pearson, J.; P. St Pierre, L. Lock, P. Buckley, E. Bell, S. Mason, R. McCarthy, W. Garratt, K. Sugar
and J. Pearce. Working with the local community to eradicate rats on an inhabited island: securing the
seabird heritage of the Isles of Scilly ........................................................................................................................ 670
Phillips, C.B.; K. Brown, K. Broome, C. Green and G. Walker. Criteria to help evaluate and guide attempts to
eradicate terrestrial arthropod pests........................................................................................................................... 400
Picó, G.; M.J. Fernández, J.E. Moreno and V. Colomar. Control of the ladder snake (Rhinechis scalaris) in
Formentera using experimental live-traps ................................................................................................................. 332
Pili, A.N.; C.E. Supsup, E.Y. Sy, M.L.L. Diesmos and A.C. Diesmos. Spatial dynamics of invasion and
distribution of alien frogs in a biodiversity hotspot archipelago ............................................................................... 337
Pink, C.; D. Algar and P. Green. Diet of introduced black rats Rattus rattus on Christmas Island: setting the
scene with stomach and stable isotope analysis ........................................................................................................ 720
Pollard, K.; D. Kurose, A. Buddie and C. Ellison. The prospects for biological control of Rubus niveus in the
Galapagos Islands...................................................................................................................................................... 720
Poncet, S.; K. Passfield, A. Kuepfer and M.A. Tabak. The effect of Norway rats on coastal waterbirds of the
Falkland Islands: a preliminary analysis ................................................................................................................... 147
Pott, M.; E. Hagen, P. Martínez and M. Díaz. A tool for biodiversity conservation within Chile: renewed
interest in island eradications sparked by successful European rabbit (Oryctolagus cuniculus) eradication ........... 721
Preston, G.R.; B.J. Dilley, J. Cooper, J. Beaumont, L.F. Chauke, S. L. Chown, N. Devanunthan, M. Dopolo,
L. Fikizolo, J. Heine, S. Henderson, C.A. Jacobs, F. Johnson, J. Kelly, A.B. Makhado, C. Marais,
J. Maroga, M. Mayekiso, G. McClelland, J. Mphepya, D. Muir, N. Ngcaba, N. Ngcobo, J.P. Parkes,
F. Paulsen, S. Schoombie, K. Springer, C. Stringer, H. Valentine, R.M. Wanless and P.G. Ryan. South
Africa works towards eradicating introduced house mice from sub-Antarctic Marion Island: the largest
island yet attempted for mice ...................................................................................................................................... 40
Ragen, P.; S. Pandoo, V. Bachraz, Z. Jhumka and J. Mauremooto. Finders keepers? Discovering and securing
the rare species rediscovered in weeded restoration plots......................................................................................... 721
Reaser, J.K.; G.R. Howald and S.D. Veatch. A plan for the eradication of invasive alien species from Arctic islands . 679
Richardson, M.G. and J.P. Croxall. Achieving post-eradication biosecurity on South Georgia..................................... 489
Roberts, M.; W. Cresswell and N. Hanley. Impacts and control of invasive species: trading off actions ...................... 721
Robertson, P.A.; S. Roy, A.C. Mill, M. Shirley, T. Adriaens, A.I. Ward, V. Tatayah and O. Booy. Invasive
species removals and scale – contrasting island and mainland experience............................................................... 687
Rocamora, G. Eradication of invasive animals and other island restoration practices in Seychelles:
achievements, challenges and scaling up perspectives ............................................................................................. 588
Rogers, A.; J. Shaw and S. Kark. Incorporating interaction networks into conservation: Tasmania as a case study ..... 722
Rojas-Mayoral, E.; F.A. Méndez-Sánchez, B. Rojas-Mayoral and A. Aguirre-Muñoz. Improving the
efficiency of aerial rodent eradications by means of the numerical estimation of rodenticide density....................... 47
Rojas-Sandoval, J.; P. Acevedo-Rodríguez, M. Datiles, S. Dube, H. Diaz-Soltero, L. Charles, G. Richards,
M. Angel Duenas, D. Simpson, E. Ventosa-Febles, J. Ackerman, F. Areces-Berazain, M. Caraballo-Ortiz,
A. Carvajal-Vélez, J. Chabert-Llompart, S. Kaufman, J. Thompson and J. Vélez-Gavilán. Invasive plants
of the Caribbean: application of herbarium collections to protect a regional biodiversity hotspot .......................... 722
Rueda, D.; V. Carrion, P.A. Castaño, F. Cunninghame, P. Fisher, E. Hagen, J.B. Ponder, C.A. Riekena,
C. Sevilla, H. Shield, D. Will and K.J. Campbell. Preventing extinctions: planning and undertaking
invasive rodent eradication from Pinzon Island, Galapagos ....................................................................................... 51
Ruffino, L. and T. Cornulier. Can large database mining inform invasive non-native species management on
islands?........................... ........................................................................................................................................... 722
Russell, J.C. and C.N. Taylor. Strategic environmental assessment for invasive species management on
inhabited islands ........................................................................................................................................................ 692
Russell, P. and S. Weaver. Management of numerous introduced plants on Matiu (Somes Island), Wellington,
New Zealand ............................................................................................................................................................. 723
Saavedra Cruz, S. and S.J. Reynolds. Eradication and control programmes for invasive mynas (Acridotheres
spp.) and bulbuls (Pycnonotus spp.): defining best practice in managing invasive bird populations on
oceanic islands........................................................................................................................................................... 302
Samaniego-Herrera, A.; S. Boudjelas, G.A. Harper and J.C. Russell. Assessing the critical role that land crabs
play in tropical island rodent eradications and ecological restoration ...................................................................... 209
Sandodden, R. Eradication of invasive alien crayfish: past experiences and further possibilities ................................. 405
732
Schiavini, A.; J. Escobar, E. Curto and P. Jusim. First results from a pilot programme for the eradication of
beavers for environmental restoration in Tierra Del Fuego ........................................................................................ 57
Schmack, J.M.; M.C. Barron, D.F. Ward and J.R. Beggs. Managing Vespula wasp invasion in New Zealand ............. 723
Serr, M.; N. Heard and J. Godwin. Towards a genetic approach to invasive rodent eradications: assessing
reproductive competitiveness between wild and laboratory mice .............................................................................. 64
Shiels, A.B.; D. Will, C. Figuerola-Hernández, K.J. Swinnerton, S. Silander, C. Samra and G.W. Witmer.
Trail cameras are a key monitoring tool for determining target and non-target bait-take during rodent
removal operations: evidence from Desecheo Island rat eradication ........................................................................ 223
Siers, S.R.; W.C. Pitt, J.D. Eisemann, L. Clark, A.B. Shiels, C.S. Clark, R.J. Gosnell and M.C. Messaros.
In situ evaluation of an automated aerial bait delivery system for landscape-scale control of invasive
brown treesnakes on Guam ....................................................................................................................................... 348
Sjodin, B.; R. Irvine, G. Howald and M. Russello. Invasive rat colonisation history and movement dynamics
in Haida Gwaii .......................................................................................................................................................... 723
Skei, B. Garden cans and river rafts – equipped to approach invasive freshwater fish .................................................. 724
Sposimo, P.; D. Capizzi, T. Cencetti, F. De Pietro, F. Giannini, C. Gotti, F. Puppo, G. Quilghini, E. Raganella
Pelliccioni, G. Sammuri, V. Trocchi, S. Vagniluca, F. Zanichelli and N. Baccetti. Rat and lagomorph
eradication on two large islands of central Mediterranean: differences in island morphology and
consequences on methods, problems and targets ...................................................................................................... 231
Springer, K. What happens after the helicopters have gone – assessing post-eradication changes on
Macquarie Island ....................................................................................................................................................... 724
Stringer, C.; S. Boudjelas, K. Broome, S. Cranwell, E. Hagen, G. Howald, J. Kelly, J. Millett, K. Springer
and K. Varnham. Married bliss and shotgun weddings: effective partnerships for island restoration ...................... 517
Thibault, M.; F. Brescia and M. Barbet-Massin. Predicting the distribution of island invader bird species
under climate change ................................................................................................................................................. 724
Thibault, M.; E. Vidal, M.A. Potter, F. Masse, A. Pujapujane, B. Fogliani, G. Lannuzel, H. Jourdan, N.
Robert, L. Demaret, N. Barré and F. Brescia. Invasion by the red-vented bulbul: an overview of recent
studies in New Caledonia .......................................................................................................................................... 309
Thompson, R.C. and J.M. Ferguson. Removing introduced hedgehogs from the Uists................................................. 274
Tye, A. Towards a guidance document for invasive species planning and management on islands .............................. 698
Vanderwoude, C.; S. Boudjelas, P. Andreozzi, P. Cowan and J. Wainiqolo. Biosecurity Plan for invasive ants
in the Pacific .............................................................................................................................................................. 725
Varnham, K.; E. Radford, S. Busuttil, C. Forbes, E. Gibbs-Williams and G. Gerber. Rat control to protect the
Turks and Caicos rock iguana: monitoring and responding to rat activity on a Caribbean island Nature Reserve .. 725
Varnham, K.; S. Thomas, L. Bambini, S. Havery and L. Lock. An innovative programme to protect the UK’s
seabird islands ........................................................................................................................................................... 725
Veatch, S.D. Prioritising islands for the eradication of invasive vertebrates in the Arctic ............................................. 726
Veitch, C.R. Changes in forest passerine numbers on Hauturu following rat eradication .............................................. 726
Walsh, A.; A. Wilson, H. Bower, P. McClelland and J. Pearson. Winning the hearts and minds – proceeding to
implementation of the Lord Howe Island rodent eradication project: a case study .................................................. 522
Webber, B.; K. Uyehara, T. Luxner and D. Dewey. Habitat features that influence predation of endangered
Hawaiian common gallinule nests by invasive vertebrates in Hanalei and Huleia National Wildlife Refuges ........ 726
Wegmann, A.; G. Howald, S. Kropidlowski, N. Holmes and A.B. Shiels. No detection of brodifacoum
residues in the marine and terrestrial food web three years after rat eradication at Palmyra Atoll, Central Pacific . 600
West, C.J. and D. Havell. Weed eradication on Raoul Island, Kermadec Islands, New Zealand: progress and
prognosis ................................................................................................................................................................... 435
Will, D.; T. Hall, M. Khalsa and J. Bruch. Using sUAS to direct trap placement in support of feral cat
eradication on islands ................................................................................................................................................ 727
Will, D.; G. Howald, N. Holmes, R. Griffiths and C. Gill. Considerations and consequences when conducting
aerial broadcast applications during rodent eradications ............................................................................................ 71
Will, D.J.; K. Swinnerton, S. Silander, B. Keitt, R. Griffiths, G.R. Howald, C.E. Figuerola-Hernandez and
J.L. Herrera-Giraldo. Applying lessons learnt from tropical rodent eradications: a second attempt to
remove invasive rats from Desecheo National Wildlife Refuge, Puerto Rico .......................................................... 154
Wynn, M. and D. Driscoll. Removal of invasive, black rats increases activity levels and population density of
Christmas Island’s last remaining endemic reptile .................................................................................................... 727
Zhao, C.; J. Li and X. Liu. Effect of Spartina alterniflora invasion on benthic macro-invertebrate
communities in Guangxi Zhuang Autonomous Region ............................................................................................ 727
733
734
INTERNATIONAL UNION
FOR CONSERVATION OF NATURE
WORLD HEADQUARTERS
Rue Mauverney 28
1196 Gland, Switzerland
mail@iucn.org
Tel +41 22 999 0000
Fax +41 22 999 0002
www.iucn.org