Ecology, 87(8), 2006, pp.

1967–1972
Ó 2006 by the the Ecological Society of America

JELLYFISH AGGREGATIONS AND LEATHERBACK TURTLE FORAGING

PATTERNS IN A TEMPERATE COASTAL ENVIRONMENT
JONATHAN D. R. HOUGHTON,1 THOMAS K. DOYLE,2 MARK W. WILSON,2 JOHN DAVENPORT,2 AND GRAEME C. HAYS1,3
1
Department of Biological Sciences, Institute of Environmental Sustainability, University of Wales Swansea, Singleton Park,
Swansea SA2 8PP United Kingdom
2
Department of Zoology, Ecology and Plant Sciences, University College Cork, Lee Maltings, Prospect Row, Cork, Ireland

Abstract. Leatherback turtles (Dermochelys coriacea) are obligate predators of gelatinous

zooplankton. However, the spatial relationship between predator and prey remains poorly
understood beyond sporadic and localized reports. To examine how jellyfish (Phylum
Cnidaria: Orders Semaeostomeae and Rhizostomeae) might drive the broad-scale distribution
of this wide ranging species, we employed aerial surveys to map jellyfish throughout a
temperate coastal shelf area bordering the northeast Atlantic. Previously unknown, consistent
aggregations of Rhizostoma octopus extending over tens of square kilometers were identified in
distinct coastal ‘‘hotspots’’ during consecutive years (2003–2005). Examination of retro-
spective sightings data (.50 yr) suggested that 22.5% of leatherback distribution could be
explained by these hotspots, with the inference that these coastal features may be sufficiently
consistent in space and time to drive long-term foraging associations.
Key words: aerial survey; Dermochelys coriacea; foraging ecology; gelatinous zooplankton; jellyfish;
leatherback turtles; planktivore; predator–prey relationship; Rhizostoma octopus.

INTRODUCTION remains the leatherback turtle Dermochelys coriacea that

Understanding the distribution of species is central to ranges widely throughout temperate waters during

REPORTS
many ecological studies, yet this parameter is sometimes summer and autumn months (e.g., Brongersma 1972).
difficult to measure even for species that may be The reliance of such a large animal (up to 916 kg;
abundant and play important trophic roles. In some Davenport 1998) on an energetically improbable diet
cases it is the environment itself that makes surveys remains intriguing, as the demands of growth and
difficult, such as in remote rainforests or the deep sea, metabolism have to be met in the context of reproduc-
while in other cases it is aspects of the animals own tive periods lasting several months with foraging and
biology that impedes studies. For example, within breeding grounds often separated by thousands of
plankton research there are a range of nets and kilometers (Hays et al. 2004a, b, Arai 2005). In the
autonomous recorders that have been used routinely Atlantic, this constitutes large-scale migration from
for many decades to assess the distributions of species tropical breeding sites to more temperate foraging
(Harris et al. 2000). Yet many gelatinous zooplankton grounds. For example, leatherbacks nesting within the
such as jellyfish (Phylum Cnidaria: Orders Semaeosto- wider Caribbean basin are known to migrate to the cool,
meae and Rhizostomeae) remain poorly studied because temperate waters around the Canadian seaboard, West-
they either (a) break up in nets and/or (b) are so patchily ern Europe, and beyond (Ferraroli 2004, Hays et al.
distributed that they are difficult to study using tradi- 2004a, b). At these distant foraging grounds leather-
tional techniques (Mills 2001). Despite this, the group is backs have been observed to consume great quantities of
highly topical within marine systems with much recent jellyfish (up to 200kg/d; Duron-Dufrenne 1987) with
attention directed towards their influence as predators turtles regularly seen in areas where jellyfish are
(e.g., Arai 1988). Comparatively, their role as prey abundant at the surface (James and Herman 2001).
remains largely understudied with many species depicted Despite such tantalizing insights, it remains unknown
incorrectly as a trophic dead end as their low nutritional as to how or whether temperate jellyfish aggregations
value makes them an unlikely food item for vertebrates drive the broad-scale distribution and foraging behavior
(Arai 2005). However, there is a growing body of of leatherback turtles. This long-standing question has
evidence to suggest the contrary with an emerging list of implications for ecologists and conservationists alike
potential predators ranging from larval fish (Arai 2005) following recent reports that implicate fisheries by-catch
through to large oceanic sea birds (Catry et al. 2004). as a proximate cause for the regional declines in leather-
However, perhaps the best known jellyfish predator back turtles (Spotila et al. 2000, Lewison et al. 2004).
Therefore, with the clear goal of better understanding the
Manuscript received 7 June 2005; revised 23 February 2006;
association of predator and prey, we conducted low-level
accepted 20 March 2006. Corresponding Editor: E. Sala. aerial surveys over three years to identify and map
3
Corresponding author. E-mail: g.hays@swan.ac.uk jellyfish aggregations throughout the Irish Sea, an area
1967
 1968 JONATHAN D. R. HOUGHTON ET AL. Ecology, Vol. 87, No. 8

spanning several thousand square kilometers that is ‘‘zones’’ (20 3 15 km), accounting for ;72% of all turtle
regularly visited by migratory leatherbacks. By examin- records (n ¼ 103) and a spatial coverage of ;25% of the
ing jellyfish assemblages on such a spatial and temporal entire area.
scale, we were able to make an initial assessment of how To control for area, we converted the number of live
historical sightings of leatherbacks might reflect the sightings to a density value (i.e., individuals/100 km2).
distribution of their gelatinous prey. From this, leatherback distribution did not appear to be
random. Therefore, we redistributed the 103 observa-
METHODS tions between the coastal zones, with the chances of a
Mapping the prey: aerial surveys turtle occurring in each zone proportional to the relative
Estimates of jellyfish abundance were made during sea area of that zone. This procedure was iterated 1000
aerial surveys from an altitude of 152 m (500 ft) at a times. Lastly, to produce an index of Rhizostoma
constant speed of 185 km/h (100 knots). Numbers of abundance, the total number of jellyfish observed in
jellyfish observed within a 250-m observation strip each each section over the three survey years was corrected
side of the aircraft were tallied every five minutes by for effort to give the mean number of jellyfish observed
each observer and combined to give an estimate of within a given 5-min survey period (7710 m2).
abundance (detailed methods are described in Houghton Data analysis: removing potential biases
et al. 2006). Aggregations were mapped (51.2–55.08 N,
3.0–8.58 W) from June to October over consecutive years We examined International Council for the Explora-
(2003–2005) for three scyphozoan species: the barrel tion of the Seas data (available online)5 to derive effort for
jellyfish, Rhizostoma octopus; the lion’s mane, Cyanea a range of fishing activities within U.K. waters: demersal,
capillata; and the compass, Chrysaora hysoscella. pelagic, shrimp and Nephrops, and other shellfish
Random transects were carried out periodically (excluding Nephrops and shrimp). Such diverse activities
throughout the three years to ensure that the aggrega- cannot be described by a simple common metric and, as
tions had remained spatially consistent on an intra- and such, the overall value of different species by area
interseasonal scale (Fig. 1). To validate aerial observa- (financial yield per ICES unit) is taken as a proxy for
tions, jellyfish stranding events were also recorded fishing effort. Data were available as arbitrary yet
directly comparable categories, from the highest yield
REPORTS

through regular beach surveys at low tide. Data were

collected for all U.K./Irish schyphozoan species (denoted by a score of 5 in our analysis) to no yield at all
although only the three species observed from the air (scored 0). Next, we estimated the number of recreational
were considered for the present study. For Rhizostoma moorings in each zone, making the broad assumption
octopus bell diameter was also recorded. Lastly, from the that pleasure craft would remain active within their
air, the location and abundance of leatherback turtles respective zones, with coastal demographics (present
were noted using standard distance sampling techniques population) the final factor included in our analysis.
(Buckland et al. 2001). These combined factors could not explain the distribu-
tion of leatherback sightings (Kendall’s Tau_b; P .
Mapping the predator: historical 0.05), implying they had not rendered the data unusable
databases—leatherback turtles for assessing potential links between predator and prey.
We examined an existing data set for anecdotal RESULTS
sighting and stranding data for leatherback turtles
around the United Kingdom and Ireland. The database Mapping jellyfish aggregations from the air
(entitled ‘‘TURTLE’’; available online)4 was compiled by During 2003–2005 (June–October), we surveyed a
Marine Environmental Monitoring (MEM), a member combined total of 7700 km2 (2003, 3034 km2; 2004, 2941
of the DEFRA ‘‘Collaborative U.K. Cetacean and km2; 2005, 1725 km2; Fig. 1a, c, and e). During 2003 and
Marine Turtle Strandings Project.’’ This database is a 2004 we encountered three areas where Rhizostoma were
public access resource compiling U.K. and Irish records repeatedly observed in high densities (i.e., .800
of sightings and strandings dating back to 1748. We individuals/5 min): Carmarthen Bay in South Wales,
restricted our analysis to the area consistent with our Tremadoc Bay in North Wales, and Rosslare on the
aerial survey program (51.2–55.08 N, 3.0–8.58 W). southeast coast of Ireland (Fig. 1). Survey coverage was
Sightings data for leatherbacks revealed a strong revised in 2005, with the Solway Firth identified as a
coastal bias (distance from shore ¼ 7.4 6 0.6 km (mean fourth area where Rhizostoma occurred in significant
6 SE); minimum 0.0 km, maximum 35.8 km; n ¼ 143). To numbers (
600 individuals/5 min; Fig. 1f).
account for this, a 15-km buffer was created around the Rhizostoma aggregations sometimes extended over
entire coastline (ArcView 3.2; ESRI 1999) with sightings several tens of kilometers. Although we do not know the
beyond this removed from the analysis (see Appendix). density of Rhizostoma beneath the surface, on occasion
The resulting area was divided into 61 separate coastal jellyfish were so abundant that we could only conclude

4 5
hhttp://www.strandings.comi hhttp://www.marlab.ac.uki
August 2006 JELLYFISH AND LEATHERBACK TURTLES 1969

REPORTS
FIG. 1. Areas covered during aerial surveys are shown for (a) 2003, (c) 2004, and (e) 2005. Each square represents the midpoint
of a 5-min survey unit (7710 m2). Distribution of Rhizostoma aggregations are also shown for (b) 2003, (d) 2004, and (f) 2005. Data
are total abundances for the period between July and September (leatherback peak season) in each year. Each circle represents a
measure of abundance during a single 5-min observation period. Relative scale of aggregations is shown in panel (d) ranging from
.1000 to 10–50 jellyfish/5 min. Locations of hotspots are shown in panel (b): A, Carmarthen Bay; B, Rosslare harbor; C,
Tremadoc Bay. A fourth possible hotspot (D, Solway Firth) is also shown although this site was only surveyed once under good
conditions, thus preventing a full assessment of its temporal and spatial constancy.

there were .1000 individuals seen in five minutes of aggregations, implying that they did not occur predom-
flying (i.e., a density of at least 0.13 individuals/m2 at the inantly at the surface, did not form aggregations, or (in
surface). Assessing how the aggregations extended below the case of Aurelia aurita) were not readily visible from
the surface was unfortunately beyond the scope of the the air. These species were excluded from the further
present study. This is an undeniable limitation to our analysis given the more qualitative nature of these data.
estimates of abundance and will hopefully form the basis Regarding shoreline data, 135 beaches were surveyed
of future fine-scale studies at sea level. Nonetheless, even to validate aerial observations. 1226 individual surveys
if we restrict our interpretation of our data to two were conducted, amounting to 1112 km surveyed over
dimensions we are still left with the conclusion that these the three years. Rhizostoma octopus was found to strand
‘‘hotspots’’ extend over tens of square kilometers and on a year round basis, with both small and large
contain many millions of jellyfish (Fig. 1). Apart from individuals present at each time of the year (see
Rhizostoma, other species were never seen in huge Appendix).
 1970 JONATHAN D. R. HOUGHTON ET AL. Ecology, Vol. 87, No. 8

Nevertheless, to assess whether sightings occurred in the

same places over time, or whether different hotspots
emerged at different times (e.g., Solway Firth at one
particular point in time and Carmarthen Bay at another
point) data were grouped by decade from 1950 onwards
(Fig. 2a). This revealed no distinct pattern other than to
confirm that turtles had been sighted widely throughout
the Irish Sea in each decade examined.
Comparison of the 1000 randomized leatherback
distributions with the observed distribution revealed
the number of leatherbacks in a single zone (n ¼ 16) and
the number of zones with no leatherbacks observed (n ¼
24) were larger than would be expected by random
chance (P , 0.001 for both). This nonrandom grouping
was further confirmed using a chi-square test (v2 ¼
253.70, df ¼ 61, P , 0.01).

Linking predator with prey

During 2003 and 2004 three live and one dead
leatherback turtle were observed from the air with two
of the live animals found within 1 km of Rhizostoma
aggregations (Fig. 2). There are 25 previous reports from
REPORTS

FIG. 2. (a) All leatherback sightings (for a given section of

the Irish Sea) from 1950 to 2005 (n ¼ 143). Data are plotted
according to decade: 2000–2005 (open circles); 1990s (solid
circles); 1980s (open squares); 1970s (solid squares); 1960s
(open diamonds); and 1950s (solid diamonds). (b) Sightings
where turtles were associated with jellyfish (solid triangles) and
when foraging activity was confirmed (open triangles). The
three live turtles sighted during 2003 and 2004 aerial surveys are
marked with stars. The three sightings confirming predation on
Rhizostoma octopus were in Carmarthen Bay (51.658 N, 4.538 W
and 51.618 N, 4.738 W) and Tremadoc Bay (52.808 N, 4.368 W).
Predation of Chrysaora hysoscella was observed on a single
occasion in Tremadoc Bay (54.678 N, 3.738 W).

Historical analysis of leatherback database

From the TURTLE database 147 live leatherbacks
were sighted in the Irish Sea between 1908 and 2005
(Fig. 2). However, only four turtles were reported prior
to 1950 (1908, 1938, 1948, and 1949), leaving analysis
prior to this date impractical. Beyond this, the number
of sightings increased from one during 1951–1955 to 71
from 2001 to 2005 (Fig. 3a). However, this apparent
increase most likely reflects the centralization of records
on a national scale, improvements to the reporting
mechanism itself, and increased awareness of leather-
back turtles in British and Irish waters.
Regarding seasonality, leatherbacks were most fre-
quently sighted between July and September (n ¼ 125;
87.67%; Fig. 3b). Arguably, this may reflect an increase FIG. 3. (a) Live leatherback turtle sightings since 1950
in boat traffic during summer months although the (from TURTLE database). (b) Seasonality of live leatherback
absence of unit effort prevented a more detailed analysis. turtles sightings (total n ¼ 143; 1950–2005).
August 2006 JELLYFISH AND LEATHERBACK TURTLES 1971

the region (TURTLE database) that state leatherbacks the vertical distribution and surface behavior of the
were ‘‘feeding/swimming amongst/or associated with species within the Irish Sea. However, our objectives
jellyfish.’’ A further 10 records specifically state individ- were not merely to report numbers, but to consider the
uals feeding on jellyfish. Only four records accurately overall distribution of sighted individuals in terms of
identified the prey species, and of these, three were prey aggregations. Consequently, we feel that this
Rhizostoma and the other was Chrysaora (Fig. 2b). limitation in our analysis does not compromise this
A more empirical association was found when key element of the study.
leatherback density was correlated against the Rhizo- More importantly, given that jellyfish are widespread
stoma index in each of the coastal zones. A Kendall’s throughout temperate coastal waters (Russell 1970), we
Tau_b nonparametric test revealed 22.5% of the must also emphasize that Rhizostoma is simply one
variation in leatherback distribution could be explained species that they feed upon. Indeed in recent years,
by the distribution of Rhizostoma (P , 0.05). Addition- leatherbacks off Nova Scotia have also been observed to
ally, zones of high leatherback abundance were charac- feed on Cyanea capillata (James and Herman 2001) and
terized by higher rates of incidental mortality (Kendall’s off North Carolina, USA, on Stomolophus meleagris
nonparametric test; r2 ¼ 0.42, P , 0.05). (Grant and Ferrell 1993). It should also be noted that
significant numbers of Cyanea capillata have been
DISCUSSION previously shown to strand around the north coast of
Foraging decisions are made over a wide range of Wales (.50 individuals/100 m; Houghton et al. 2006),
spatial and temporal scales, based on differences which must in some way contribute to relatively high
between prey patches, factors within individual patches, numbers of leatherbacks sighted in that area. Moreover,
and the motivational state of a forager (Baum and Rhizostoma itself can sometimes occur in extraordinary
Grant 2001). Although the exact response of a particular abundance outside of the hotspots, as in 1978 when the
predator to changes in the spatial distribution of its prey species was reported all around the coast of Ireland
is often poorly understood, temporal variations in the (O’Connor and McGrath 1978). So while the hotspots
distribution of the predator itself can sometimes provide are certainly not the sole factor determining the
an insight into highly dynamic prey that may be difficult distribution of leatherbacks in U.K. and Irish waters,
to track in other ways (Boyd 1996). This has particular they appear sufficiently consistent in space and time to

REPORTS
relevance within the marine environment where it is drive an increased occurrence of turtles in these specific
often difficult to measure simple variables and life areas over the past 50 years.
history traits by direct observation (Boyd 1996). The This relationship between Rhizostoma spp. and
scale of this problem is proportional to the spatial scale leatherback turtles has been briefly alluded to before
within which the predator and prey operate (e.g., Sims within European waters. Along the Atlantic coast of
and Quayle 1998), which over recent years has driven a France, leatherbacks have been observed to feed
number of technological advances that allow empirical extensively on Rhizostoma spp. while also consuming
data to be gathered from previously intractable environ- Chrysaora hysoscella, Aurelia aurita, and more rarely,
ments. However, as our real time capabilities increase, Cyanea lamarckii (Duguy 1982, Duron-Dufrenne 1987).
we should be careful not to devalue historical data, as Given the geographical proximity of France to the Irish
often they can provide insights beyond the scope of Sea, such findings tentatively suggest that the associa-
contemporary findings alone (Jackson et al. 2001, tion of leatherbacks and Rhizostoma may extend further
Pandolfi et al. 2003). The long-standing conundrum of into the wider region of Europe’s Atlantic fringe.
jellyfish–leatherback turtle foraging associations Furthermore, given the persistent stranding of Rhizo-
presents one such scenario given a wealth of recent stoma throughout the year, the temporal inference is
studies into their migratory habits (e.g., Ferraroli et al. that the species may provide a food source for leather-
2004, Hays et al. 2004b, James et al. 2005) that overlies backs beyond the scope of other scyphozoans that are
an almost nonexistent knowledge of how such behaviors largely absent from the water column during the autumn
are determined by the distribution and seasonality of and winter months (Russell 1970).
their prey. This can largely be explained by the sheer This leads us to a more general discussion of how
scale of leatherback movements rendering direct obser- leatherbacks might interact with prey assemblages over
vations impractical, but also by the fact that postnesting consecutive seasons. For example, long-term telemetry
migrations are dispersed (Ferraroli 2004, Hays et al. of leatherback turtles tagged in their foraging grounds
2004b) and do not appear to center on spatially off Nova Scotia has shown that after migrating south at
restricted foraging grounds. In this context, the co- the end of the summer and overwintering at low
occurrence of leatherbacks and Rhizostoma becomes latitudes, individuals turtles then return to high latitudes
important as it suggests that migratory movements may the following summer and seem to maintain fidelity at
reflect the distribution of particular prey items. None- least to the approximate area they frequented the
theless, in terms of absolute abundance we cannot claim previous year (James et al. 2005). However, these
that the numbers of turtles sighted are representative of tracking results show no evidence for fidelity to specific
the actual numbers present, given no prior knowledge of foraging bays or such localized areas. Consequently, it is
 1972 JONATHAN D. R. HOUGHTON ET AL. Ecology, Vol. 87, No. 8

unlikely that leatherbacks maintain knowledge of the Duguy, R. 1982. Note sur les méduses des Pertuis Charentais.
Rhizostoma hotspots in the Irish Sea and travel directly Annales de la Société des Sciences Naturelles de la Charente-
Maritime 6:1029–1034.
to these bays each year. It is more likely that some Duron-Dufrenne, M. 1987. Premier suivi par satellite en
individual turtles maintain fidelity to the northeast Atlantique d’une tortue luth Dermochelys coriacea. Comptes
Atlantic and that they occasionally stumble into a Rendus Academie des Sciences Paris 304:339–402.
number of hotspots, whereupon they can feed until ESRI. 1999. ArcView. Version 3.2. Environmental Systems
Research Institute (ESRI). Redlands, California, USA.
cooling waters drive their southerly migration.
Ferraroli, S., J.-Y. Georges, P. Gaspar, and Y. Le Maho. 2004.
In summary, the distribution of Rhizostoma octopus Where leatherback turtles meet fisheries. Nature 429:521–
explained almost a quarter of the variance in sighted 522.
leatherbacks over a period of .50 years, on a scale Grant, G. S., and D. Ferrell. 1993. Leatherback turtle,
spanning several thousand square kilometers. We do not Dermochelys coriacea (Reptilia, Dermochelidae)—notes on
near-shore feeding behavior and association with Cobia.
imply that leatherback distribution is entirely driven by Brimleyana 19:77–81.
this single species, and we fully acknowledge that other Harris, R., P. Wiebe, J. Lenz, H.-R. Skjoldal, and M. E.
jellyfish species play an important role. However, we Huntley, editors. 2000. ICES zooplankton methodology
believe our results represent a solid foundation for manual. Academic Press, London, UK.
Hays, G. C., J. D. R Houghton, C. Isaacs, R. S. King, C. Lloyd,
further studies into leatherback foraging behavior that and P. Lovell. 2004a. First records of oceanic dive profiles for
also highlights the broader issue of how jellyfish as prey leatherback turtles, D. coriacea, indicate behavioural plasti-
may form integral links within temperate coastal marine city associated with long-distance migration. Animal Behav-
systems. iour 67:733–743.
Hays, G. C., J. D. R. Houghton, and A. E. Myers. 2004b. Pan-
ACKNOWLEDGMENTS Atlantic leatherback turtle movements. Nature 429:522.
Houghton, J. D. R., T. K. Doyle, J. Davenport, and G. C.
Funding was provided by INTERREG IIIA (European
Hays. 2006. Developing a simple rapid method for identify-
Regional Development Fund), the Countryside Council for
ing and monitoring jellyfish aggregations from the air.
Wales Species Challenge Fund, and the Marine Conservation
Marine Ecology Progress Series 314:159–170.
Society. Special thanks to David Jones, Vincent, Sean, and
Jackson, J. B. C. et al. 2001. Historical overfishing and recent
Christina Rooney, Jim and Rose Hurley, Kevin McCormack,
collapse of coastal ecosystems. Science 293:629–638.
Eithne Lee, Maria Doyle, Kate Williamson, Irena Kruszona
James, M. C., and T. B. Herman. 2001. Feeding of Dermochelys
and colleagues, Vernon Jones, and Tom Stringell; and to Rod
coriacea on medusae in the Northwest Atlantic. Chelonian
REPORTS

Penrose and Chris Pierpoint of Marine Environmental Mon-

Conservation Biology 4:202–205.
itoring for compiling the TURTLE database.
James, M. C., C. A. Ottensmeyer, and R. A. Myers. 2005.
LITERATURE CITED Identification of high-use habitat and threats to leatherback
sea turtles in northern waters: new directions for conserva-
Arai, M. N. 1988. Interactions of fish and pelagic coelenterates. tion. Ecology Letters 8:195–201.
Canadian Journal of Zoology 66:1913–1927. Lewison, R. L., S. A. Freeman, and L. B. Crowder. 2004.
Arai, M. N. 2005. Predation on pelagic coelenterates: a review. Quantifying the effects of fisheries on threatened species: the
Journal of the Marine Biological Association of the United impact of pelagic longlines on loggerhead and leatherback
Kingdom 85:523–526. sea turtles. Ecology Letters 7:221–231.
Baum, K. A., and W. E. Grant. 2001. Hummingbird foraging Mills, C. E. 2001. Jellyfish blooms: are populations increasing
behavior in different patch types: simulation of alternative globally in response to changing ocean conditions? Hydro-
strategies. Ecological Modeling 137:201–209. biologia 451:55–68.
Boyd, I. L. 1996. Temporal scales of foraging in a marine O’Connor, B. D. S., and D. McGrath. 1978. On the occurrence
predator. Ecology 77:426–434. of the scyphozoan Rhizostoma octopus (L.) around the Irish
Brongersma, L. D. 1972. European Atlantic turtles. Zoologi- coast in 1976. Irish Naturalist Journal 19:261–263.
sche Verhandelingen (Leiden) 121:1–318. Pandolfi, J. M., R. H. Bradbury, E. Sala, T. P. Hughes, K. A.
Buckland, S. T., D. R. Anderson, K. P. Burnham, J. L. Laake, Bjorndal, R. G. Cooke, D. McArdle, L. McClenachan, M. J.
D. L. Borchers, and L. Thomas. 2001. Introduction to H. Newman, G. Paredes, R. R. Warner, and J. B. C. Jackson.
distance sampling: estimating abundance of biological 2003. Global trajectories of long-term decline of coral reef
populations. Oxford University Press, Oxford, UK. ecosystems. Science 301:955–958.
Catry, P., R. A. Phillips, B. Phalan, J. R. D. Silk, and J. P. Russell, F. S. 1970. The medusae of the British Isles. Volume II.
Croxall. 2004. Foraging strategies of grey-headed albatrosses Cambridge University Press, Cambridge, UK.
Thalassarche chrysostoma: integration of movements, activity Sims, D. W., and V. A. Quayle. 1998. Selective foraging
and feeding events. Marine Ecology Progress Series 280:261– behaviour of basking sharks on zooplankton in a small-scale
273. front. Nature 393:460–464.
Davenport, J. 1998. Sustaining endothermy on a diet of cold Spotila, J. R., R. D. Reina, A. C. Steyermark, P. T. Plotkin,
jelly: energetics of the leatherback turtles Dermochelys and F. V. Paladino. 2000. Pacific leatherbacks face ex-
coriacea. British Herpetological Society Bulletin 62:4–8. tinction. Nature 405:529–530.

APPENDIX
A description of the construction of coastal zones, a figure showing construction of the coastal zones, and a figure showing
stranded specimens of Rhizostoma octopus (Ecological Archives E087-123-A1).