Ocean and Coastal Management 192 (2020) 105210
Contents lists available at ScienceDirect
Ocean and Coastal Management
journal homepage: http://www.elsevier.com/locate/ocecoaman
Reef fish biomass recovery within community-managed no take zones
Hannah Gilchrist a, *, Steve Rocliffe d, Lucy G. Anderson b, Charlotte L.A. Gough a, c
a
Blue Ventures Conservation, The Old Library, Trinity Road, Bristol, BS2 0NW, UK
Independent Researcher, Bath, UK
c
Centre for Ecology and Conservation, College of Life and Environmental Sciences, University of Exeter, Cornwall Campus, Penryn, TR10 9EZ, UK
d
College of Life and Environmental Sciences, University of Exeter, Stocker Road, Exeter, EX4 4QD, UK
b
A R T I C L E I N F O
A B S T R A C T
Keywords:
Western Indian Ocean
Coral reefs
Locally managed marine areas
Conservation
Madagascar
Locally managed marine areas (LMMAs) are increasingly recognised as a key strategy for small-scale fisheries
management across the Indo-Pacific. When effective, LMMAs can encourage responsible fishing, strengthen
compliance and improve adaptive capacity, and may help to safeguard food security, address coastal poverty and
increase resource sustainability. However, evidence that LMMAs can achieve long-term biological goals is
limited. Here, we used a six-year dataset and a before-after-control-impact sampling design to assess the biological effectiveness of five community-managed no take zones (CMNTZs) situated within the Velondriake LMMA
in southwest Madagascar. Linear mixed-effect models revealed that the difference in biomass between control
and reserve sites increased over time. Significant differences in biomass between CMNTZs and controls were only
evident from year two onwards, with 189% more total biomass in CMNTZs than fished control sites by year six.
There was no effect of CMNTZs on the biomass of fish families preferentially targeted by the local fishery,
limiting the long-term fisheries benefits of this reserve network unless individual CMNTZs are made larger to
accommodate the home ranges of fishery targeted families. There were however, reserve effects preventing the
decline of untargeted fish families and species richness.
Importantly, these CMNTZs delivered a conservation benefit that rivals government-run NTZs in the region,
against a backdrop of severe biomass depletion, coastal poverty and human dependence on fishing - illustrating
their suitability as a solution to marine resource depletion in developing tropical countries.
1. Introduction
Marine protected areas (MPAs), places where resource use is regulated, are a widely advocated tool for marine conservation and fisheries
management. Meta-analyses have shown that MPAs, especially those
involving the use of permanent no take zones (areas where all extraction
is prohibited), can increase the average total biomass, density, size and
diversity of species (Halpern, 2003; Lester et al., 2009a). Mounting evidence also indicates that spillover of adults (Gell and Roberts, 2003;
Russ and Alcala, 2011) and export of larvae beyond protected boundaries can benefit surrounding fisheries (Harrison et al., 2012; Pelc et al.,
2010).
Despite their widespread use, MPAs that are managed using a ‘topdown’ approach can often be of little conservation value, existing only as
‘paper parks’ without active management or enforcement (Halpern,
2014; Mora et al., 2006). Though negative assessments of individual
sites are rare in the literature, global (Balmford et al., 2004; Burke et al.,
2011; Gill et al., 2017) and regional (Samoilys and Obura, 2011) evaluations suggest that many MPAs lack adequate management or
financing. Furthermore, the most effective MPAs tend to be fully protected, well enforced, established for more than 10 years, larger than
100 km2 and in isolated locations (Edgar et al., 2014); criteria that are
unlikely to be achievable along populated coastlines, where resource
pressure and degradation are often most acute, and where over a billion
people depend on seafood as their primary protein source (Gutierrez
et al., 2011).
In these contexts, and particularly in low income settings where
human dependence on fishing is high, the participation of communities
in marine resource management may be a preferable approach. Comanagement, where local communities share responsibility for marine
resource management with governments or other partners, is receiving
increasing attention worldwide (Cinner et al., 2012). A recent review of
* Corresponding author.
E-mail addresses: hannah@blueventures.org (H. Gilchrist), s.rocliffe@exeter.ac.uk (S. Rocliffe), anderson.lucyg@gmail.com (L.G. Anderson), charlie@
blueventures.org (C.L.A. Gough).
https://doi.org/10.1016/j.ocecoaman.2020.105210
Received 3 June 2019; Received in revised form 27 January 2020; Accepted 7 April 2020
Available online 24 April 2020
0964-5691/© 2020 Elsevier Ltd. All rights reserved.
H. Gilchrist et al.
Ocean and Coastal Management 192 (2020) 105210
130 initiatives in 44 countries found that co-management could
contribute to the successful management of aquatic resources and
concluded that it was the only practical solution for most of the world’s
fisheries (Gutierrez et al., 2011).
Across the Indo-Pacific, areas where marine resources are comanaged with, or managed solely by communities, are increasingly
termed “locally managed marine areas” (LMMAs) (Govan and Tawake,
2009; Rocliffe et al., 2014). LMMAs often blend local and scientific
knowledge as well as customary and contemporary management systems, with many employing a range of techniques including permanent
no-take zones (NTZs), periodic closures, gear restrictions,
species-specific NTZs, access rights and integrated alternative livelihood
strategies (Jupiter et al., 2014; Mills et al., 2011). No global inventories
exist, but evidence from the Pacific and Western Indian Ocean suggests
that more than 500 communities in 19 countries are directly involved in
managing 24 000 km2 of coastal resources, more than 1000 km2 of
which is afforded no-take protection (Govan and Tawake, 2009; Rocliffe
et al., 2014). Furthermore, there are likely to be thousands more communities informally managing resources, which are not captured in inventories or published literature (Govan and Tawake, 2009; Jupiter
et al., 2014).
Like government-led MPAs, the central objective of many LMMAs is
to improve the long-term sustainability of fisheries. However, LMMAs
are also frequently initiated to meet other goals such as enhancing
livelihoods, strengthening local customs or empowering the community
(Jupiter et al., 2014). When effective, LMMAs can improve compliance
with regulations by resource users, encourage responsible fishing and
eopold et al., 2013;
enhance adaptive capacity (Gutierrez et al., 2011; L�
Levine and Richmond, 2014). They have also been associated with
short-term increases in resource abundance (Cinner et al., 2005, 2012;
Bartlett et al., 2009; Pollnac et al., 2001) and have shown some promise
as a means to safeguard food security and address coastal poverty
(Oliver et al., 2015; Weiant and Aswani, 2006). However, evidence that
LMMAs can achieve long-term biological goals of improving ecosystem
eopold
condition remains scarce and inconclusive (Jupiter et al., 2014; L�
et al., 2013). Upon establishment, LMMAs often face the challenges of
multiple and potentially conflicting objectives (e.g. enhancing
fisheries-dependent livelihoods whilst conserving biodiversity) and
frequently lack the leadership, technical knowledge and financial resources necessary to sustain effective management (Jupiter et al., 2014;
L�eopold et al., 2013; Levine and Richmond, 2014; Guti�errez et al.,
2011). Moreover, the recently-established and informal nature of many
areas has largely precluded the collection of comprehensive, long-term
datasets that can provide the robust evidence that such management
often requires.
Madagascar, among the world’s poorest nations, is highly dependent
on artisanal fisheries for food and income (Le Manach et al., 2012), and
has a strong recent history of LMMA establishment (Rocliffe et al.,
2014). In this study, we used a before-after-control-impact (BACI)
design to assess the biological effectiveness of a long-established
network of community-managed no take zones (CMNTZs) within
Velondriake, one of Madagascar’s oldest LMMAs. Specifically, we aimed
to determine whether the biomass of both fishery-targeted and untargeted fish species are higher within the CMNTZs, than at paired control
sites where fishing is permitted. Based on our findings, we identify
priority management and research actions to improve the effectiveness
of CMNTZs in Velondriake. We also discuss broader implications for
fisheries management and conservation, with a focus on the potential for
LMMAs to be a cost-effective, scalable means to manage tropical
coastlines in developing countries.
Madagascar in 2006 (22 S, 43 E; Fig. 1) with the aim of empowering
coastal fishing communities to take ownership of their marine resources,
and to safeguard vulnerable marine ecosystems and traditional fishing
livelihoods for future generations (Harris, 2011). The creation of the
LMMA followed a successful program of periodic fishery closures which
began in 2004 and saw fishers temporarily refrain from harvesting
Octopus cyanea, a regionally important fishery species that is both
consumed locally and sold for export to international seafood markets,
predominantly in southern Europe (Humber et al., 2006; Moreno, 2011).
The positive outcomes from the closures (Oliver et al., 2015) led to
25 coastal and island villages in the area to come together to form the
Velondriake LMMA, a 640 km2 managed area encompassing a range of
seagrass, coral reef and mangrove habitats (Gardner et al., 2017) along
approximately 45 km of coastline. As of 2015, Velondriake is managed
by an elected association of representatives from 32 villages (the
Velondriake Association) responsible for rule setting and enforcement.
Support and technical backstopping are provided by the marine conservation NGO Blue Ventures.
These communities have the legal right to define and enforce their
own management plans through a system of traditional law called Dina,
which the Malagasy judiciary system recognises. The Dina for the
Velondriake LMMA prohibits commercial fishing, destructive subsistence fishing methods such as beach seining and poison fishing, as well
as the fishing of turtles, whales, dolphins and dugongs. In addition to
these restrictions, all extractive activities are prohibited in seven permanent CMNTZs, spread over five coral reefs and two mangrove forests,
together covering 3.2 km2 within the LMMA (Harris, 2011; Gardner
et al., 2017).
2.2. Research design
To assess biological effectiveness, we used a before-after controlimpact (BACI) experimental design whereby fish biomass was sampled
at five paired control and impact sites in the year running up to reserve
closure, and in all years where data were available following the
establishment of the CMNTZs (Table 1). By attempting to deal with both
temporal and spatial variation, these designs provide the most reliable
measures of protection effects (Osenberg et al., 2006, 2011). Note that
throughout this paper, ‘reserve’ is used as a synonym for CMNTZ.
The five CMNTZs sampled (Table 1) were all located in coral reef
habitat. The two mangrove areas within Velondriake were excluded
from the dataset since their primary management objective was to increase invertebrate biomass (specifically the mud crab Scylla serrata)
rather than fish biomass. Furthermore, fish biomass in mangrove areas
would be incomparable to data derived from surveys of coral reef habitats due to inherent differences in ecosystem function and habitat
structure. All surveys took place between 2009 and 2015.
The long-term reef monitoring programme in Velondriake was
established in 2004, before the placement of the five coral reef CMNTZs.
The monitoring objectives at this time focussed on the health of coral
reef ecosystems and ecological changes over time across the LMMA as a
whole. Surveys took place on 36 reef sites comprising patch, fringing
and barrier reefs of varying depths, spread across the north, central and
southern parts of Velondriake. Unfortunately, monitoring objectives
were not adjusted as CMNTZs were established, meaning that control
sites for each CMNTZ were not chosen before their closure. Part of the
analysis presented here involves selecting fished control sites from the
places available in the pre-existing survey schedule.
To ensure that control sites were as similar to CMNTZs as possible
and to account for habitat variability, we compared habitat complexity
indices (HCIs) (Russ et al., 2005) between reserve and all candidate
control sites in the year running up to CMNTZ closure using one-way
Analysis Of Similarity (ANOSIM) (Clarke, 1993). HCIs extend live
hard coral cover metrics by including additional structural factors as
well as the underlying topography of the reef slope, and are therefore
considered to provide more reliable estimates of similarity than hard
2. Materials and methods
2.1. Study site
The Velondriake LMMA was established by 25 villages in southwest
2
H. Gilchrist et al.
Ocean and Coastal Management 192 (2020) 105210
Fig. 1. Map showing the location of the Velondriake LMMA in southwest Madagascar, as well as the locations of the five community-managed no-take zones and
paired control sites within the Velondriake LMMA that were surveyed as part of this study. (1) Ambatohalaomby, (2) Ankafotiambe, (3) Riva, (4) Agnorondriake,
(5) Tampolove.
Table 1
Properties and location of each CMNTZ. Habitat complexity index and hard coral cover are means (�SE) for all surveys at each site.
CMNTZ
Date of
Closure
Coordinates
Reef
Type
Habitat Complexity
Index
Hard Coral Cover
(%)
Average depth of reef top
(m)
Max depth
(m)
Reserve area
(km2)
Agnorondriake
Sep 2009
Patch
14.5 (�0.2)
41.8 (�1.1)
11
27.5
0.18
Ambatohalaomby
Apr 2014
Barrier
7.0 (�0.2)
17.1 (�3.3)
9.5
26.5
1.00
Ankafotiambe
Nov 2011
Barrier
5.2 (�0.1)
30.5 (�1.8)
11
29.5
0.12
Riva
Nov 2011
Patch
12.9 (�0.2)
8.0 (�1.7)
9
26
0.12
Tampolove
Feb 2014
22.12 S
43.20 E
21.88 S
43.25 E
22.95 S
43.20 E
22.01 S
43.20 E
22.21 S
43.23 E
Patch
7.2 (�0.2)
19.7 (�3.0)
12
29.5
0.19
coral cover alone (Alvarez-Filip et al., 2011; Graham et al., 2006; Wilson
et al., 2006) The site with the highest p value (p > 0.05) in pairwise
comparisons between potential control sites and each CMNTZ was
chosen as the control. Each CMNTZ was paired with a unique control
site. Candidate control sites were all located more than 1 km from
CMNTZs, as spillover effects have been detected up to this distance for
some fish families in CMNTZs elsewhere in the Western Indian Ocean
region (Da Silva et al., 2015).
Habitat complexity indices were calculated as in Russ et al. (2005),
where
0–4, where 0 was horizontal/low rugosity, and 4 was a vertical reef
wall/structurally complex (Russ et al., 2005).
2.3. Data collection
From the year running up to each CMNTZ’s closure (Table 1) to
2015, we sampled each CMNTZ site and its paired, fished control site
using underwater visual census methods at least once a year, where
possible within the same quarter to account for seasonal variation. We
conducted all surveys between an hour after sunrise and 12:30 p.m. to
account for changes in fish abundances and community composition at
different times of day (Birt et al., 2012; Mallet et al., 2016).
To quantify fish assemblages, trained field researchers sampled each
site using six 20 m x 5m x 5 m belt transects. Three transects were placed
HCI¼ (proportion of hard coral cover þ 1) x (steepness þ 1) x (rugosity þ 1)
We estimated steepness and rugosity semi-quantitatively on scales of
3
H. Gilchrist et al.
Ocean and Coastal Management 192 (2020) 105210
along the outermost edge of the reef (13–17 m depth) and three on top of
the reef (5–13 m depth), running parallel at least 10 m apart (Halford
and Thompson, 1996; Guidetti et al., 2005). Surveys were only conducted when visibility was >5 m.
After laying down the transect line but before starting a count, the
two-person survey team swam away for 3 min to allow repopulation of
the belt (Harmelin-Vivien et al., 1985). One observer counted and
identified all non-cryptic, diurnally active reef fish to species level and
estimated length to the nearest 5 cm category (Gust et al., 2001). The
second individual’s role was to provide safety support, by following
closely behind the surveyor with a surface marker buoy.
Hard coral cover was surveyed by trained volunteer data collectors
using point intercept transects (Facon et al., 2016; Hill and Wilkinson,
2004). Each survey comprised eight transects of 10 m in length, running
parallel at least 5 m apart, on each site (Jokiel et al., 2005). The transects
were orientated north-south or east-west depending on the reef’s aspect
relative to the shore. Staff dive leaders from Blue Ventures identified the
starting locations of each transect and the direction in which they ran on
each dive to avoid overlap or replication. The substrate or organism that
intersected the transect line every 20 cm was recorded. The benthos was
classified into five main categories; hard coral, sand/rubble, soft coral,
algae, and other sessile invertebrates.
the true size and all estimates not significantly different from the pipes’
actual lengths (measured using paired Student’s t-tests) (English et al.,
1997). Finally, each team member was required to conduct three supervised fish surveys following the fish biomass survey protocol. The
identification and abundance data collected had to be at least 90%
similar to those recorded by the supervising researcher.
2.6. Data analysis
In order to examine the influence of reserve protection on fish
biomass under different levels of fishing pressure, we calculated the
natural logarithm of the Response Ratio (lnRR) for total biomass in years
since reserve closure, then for families that were targeted and untargeted by the local fishery, and finally species richness (Russ et al., 2015).
Here, we define the response ratio as the ratio of mean fish biomass or
species richness in control and reserve sites, where values greater than
zero indicate higher biomass or species richness in the reserve and, ratios less than zero show greater biomass or species richness on control
sites (Samoilys et al., 2007; Hamilton et al., 2010). Response ratios were
calculated for each of the five CMNTZs using the pooled means of total
biomass on transects from the final two years of surveying (2014 and
2015) and were back-transformed to be expressed as percentage differences. Ratios were interpreted as significantly different from zero
when the 95% confidence interval of the grand mean did not include
^t�e et al., 2001).
zero (Co
The effects of reserve status (CMNTZ vs. fished control) and years
established on (log-transformed) fish biomass were explored with linear
mixed effects models using a Maximum Likelihood fitting method.
Models were repeated for i) all fish families; i) targeted fish families and
iii) untargeted fish families. We also ran a linear mixed effect model
exploring the same effects on species richness. Site status (control/
reserve), years since reserve closure, transect depth (m), hard coral
cover (%) and distance from the nearest mainland (km; a proxy for
distance to landing site and fishing pressure) were included as fixed
factors in all models, while site was included as a random factor to account for the paired experimental design. We considered that a significant interactive effect between years since established and reserve status
demonstrated a reserve effect, as long as fish biomass increased in the
reserve sites but not in the control sites over time (Russ et al., 2015).
We accounted for habitat by matching control and CMNTZ sites
using HCI, but this was not included in our models because we did not
have values for the rugosity and steepness of sites for all years. Instead
we included hard coral cover and depth as fixed factors in our models
and distance from shore was also included as a proxy for fishing pressure
differences between control sites (Mazor et al., 2014; Thiault et al.,
2017).
We identified the most parsimonious model by dropping the least
influential factors (low estimate, non-significant) and comparing the
AIC values of the before/after models, as well as using likelihood ratio
tests (ANOVA) to determine whether a simplified model with a lower
AIC value was a significant improvement on the former model (Crawley,
2007). Likelihood ratio tests were also used to test the significance of the
final model versus a null model (response variable ~ 1 (þrandom
factors)).
We assessed changes in fish community structure between sites and
years using a PERMANOVA with 999 permutations to test the interactive
effect of reserve status and years since closure on community composition. This test was run twice, once using mean family-level biomass and
once using species-level biomass.
Note that CMNTZs were established in different years (Table 2). For
Tampolove and Ambatohalaomby we have data for the two years
following CMNTZ closure, Riva and Ankafotiambe we have data for four
years following reserve closure, and six years for Agnorondriake. This
highlights an inherent flaw in our study design whereby mean patterns
in fish biomass three to six years following CMNTZ closure respond more
to site-level changes than years one and two, with years five and six
2.4. Data preparation
Using the mid-point of each size category, we converted fish lengths
from the underwater surveys to biomass using the allometric lengthweight conversion W ¼ a TLb, where W is weight in grams, TL is total
length in millimetres and a and b are species-specific constants. We
obtained the length-weightfitting parameters a and b for each species or
closest congener from Fishbase (Froese and Pauly, 2016) and standardised resulting estimates to kilograms per hectare (kg ha 1) to facilitate comparison with other studies.
Here, ‘total biomass’ refers to all identified fish families apart from
Caesionidae (fusiliers). We excluded Caesionidae from all analyses since
their widespread nature and large total biomass can mask communitylevel trends (Ackerman and Bellwood, 2000).
In order to examine the influence of reserve protection on fish
biomass under different levels of fishing pressure, we selected fish
families that were ‘targeted’ (Acanthuridae, Lethrinidae, Lutjanidae,
Siganidae), and ‘untargeted’ (Balistidae, Chaetodontidae, Labridae,
Pomacanthidae) by the local fishery (Russ et al., 2015). At this stage it is
important to note that the coral reef fishery in Velondriake is opportunistic; families classed as ‘untargeted’ are still landed for subsistence and
sale, but they are not actively targeted during fishing. Instead these
‘untargeted’ families are caught as bycatch or through opportunistic
fishing practices (Gough et al., 2009). The ‘targeted’ fish families in this
study are actively sought by fishers in the region. The fish families for
both categories were chosen using catch data collected between 2010
and 2015 in the Velondriake LMMA, with ‘targeted’ fish together making up over 60% of all landings, and ‘untargeted’ fish families accounting for <10% of landings (Blue Ventures. Unpublishe, 2015).
2.5. Quality assurance
The six-year time period of the study meant that several surveyors
were involved in data collection. To minimise observer bias and ensure
consistent and accurate data collection, we established several quality
assurance procedures. All researchers involved in data collection were
qualified marine biologists. Before conducting a survey, each member of
the team had to learn over 150 common reef fish species. To demonstrate their skills, they were required to identify 50 randomly selected
fish in-water on a local SCUBA dive and an additional 50 from images on
a screen with 100% accuracy. Researchers were also required to estimate the lengths of 42 randomly sized PVC pipes from at least 5 m away
(Bell et al., 1985) with 95% of estimates required to fall within 5 cm of
4
H. Gilchrist et al.
Ocean and Coastal Management 192 (2020) 105210
Table 2
Results of one-way ANOSIM pairwise tests between CMNTZ and paired
control site. Analysis (999 permutations) was conducted for the year running
up to reserve closure.
CMNTZ
Year of closure
R value
p value
Agnorondriake
Ambatohalaomby
Ankafotiambe
Riva
Tampolove
2009
2014
2011
2011
2014
0.109
0.003
0.089
0.082
0.092
0.92
0.40
0.15
0.06
0.93
Table 3
Model-averaged coefficient estimates, standard errors, T-values, and associated
p-values for predictors of fish biomass for all fish families across the five coral
reef CMNTZs in the Velondriake LMMA. CMNTZ status effects are relative to
fished (control) sites. The explanatory variables presented are those remaining
in the minimum adequate (Linear Mixed Effect) model. Non-significant predictors that were dropped from the model have also been listed for reference (see
NS).
representing fish biomass only at Agnorondriake.
Our analyses were performed using PRIMER 7 (Clarke and Gorley,
2015) and R Version 3.1.1 (R Development Core Team, 2014) with the
‘metafoR’ package (Viechtbauer, 2010) for response ratio calculations
and ‘lme4’ for the generalised linear mixed effect models (Bates et al.,
2015).
Response variable
Explanatory
variable
Estimate
(SE)
T
value
Total fish biomass
(all sites)
Intercept
Reserve status
Years
Years*Status
Depth
Distance to landing
site
Hard coral cover
(%)
5.28 (0.21)
0.17 (0.15)
0.05 (0.05)
0.16 (0.06)
0.04 (0.02)
NS
25.13
1.16
1.09
2.68
2.01
p value
<0.001
0.25
0.28
<0.001
<0.05
NS
3. Results
3.1. Habitat matching between control-impact pairs
The results of our ANOSIM allowed us to identify fished control sites
which had similar habitat characteristics to each CMNTZs in the year
running up to each CMNTZ’s closure (Table 2), using a minimum
required value exceeding p ¼ 0.05. The CMNTZ at Riva was relatively
unique in terms of habitat structure when compared to all other survey
sites. In the case of Riva, we concluded that HCI is no different between
the control site chosen and the CMNTZ, with a p value of 0.06 (Table 2).
We recognise the high risk of type two error; however, as this is the
highest p value out of all candidate control sites for Riva, we chose to
accept this as an unavoidable potential flaw in our study design.
3.2. General fish community characteristics
During this study, we counted a total of 77 906 fishes belonging to
165 species and 32 families over 577 transects. Average species richness
and abundance differed greatly between sites, varying from 13 species/
100 m3 at Riva Control to 20 species/100 m3 at Ankafotiambe control,
and from 5888 fish/ha at Agnorondriake control to 23 258 fish/ha at
Riva reserve. Across all sites, the three most commonly observed families were Acanthuridae (surgeonfish), Pomacentridae (damselfish) and
Pempheridae (sweepers), which together accounted for 52% of total fish
biomass, with the acanthurids comprising 43% of all biomass. Lutjanidae
(snappers) and Siganidae (rabbitfish) made the second and third largest
contributions to total biomass at 10% and 6% respectively.
Fig. 2. Response ratios (mean and 95% confidence intervals) calculated
using paired CMNTZs and controls and total fish biomass. Positive values
show biomass is higher inside the CMNTZ while negative values show biomass
is higher in fished control sites. Confidence intervals that overlap the vertical
line at zero are not significantly different from zero (i.e. there is no significant
reserve effect).
a minimum of 251 kg/ha (CI95 � 102) in year four to 665 kg/ha (CI95
� 465) in year five (Fig. 3), excluding an anomalous peak of 1404 kg/ha
(CI95 � 469) in year three. However, biomass inside reserves increased
over time from a minimum of 372 kg/ha (CI95 � 75) in year one to a
peak of 2238 kg/ha (CI95 � 406) in year fieve (Fig. 3).
In 2015, taking all five CMNTZs into account there was 219% (CI95
� 58) more biomass in CMNTZs than control sites. At the site level, Riva
and Agnorondriake both had significant positive effects on total fish
biomass (Table 4), containing 555% (CI95 � 376, year four of closure)
and 189% (CI95 � 102, year six of closure) more biomass inside the
reserve than outside in 2015, respectively. None of the other CMNTZs
had a significant effect on total fish biomass.
3.3. Reserve effects on overall fish biomass
Across all families and all sites, while neither reserve status nor time
had an independent effect on total fish biomass, the interaction of
reserve status and time since closure had a significant positive effect
(Table 3). Depth was also a significant predictor of total fish biomass
(Table 3). There was no significant difference in biomass between
CMNTZs and controls in year zero, but 322% (CI95 � 66) more biomass
in reserves than controls in year four (n ¼ 3 CMNTZs) and 189% (CI95 �
102) more biomass in CMNTZs in year six (n ¼ 1 CMNTZ) (Fig. 2). In
years four and six the difference in mean biomass between reserves and
controls was 567 kg/ha and 809 kg/ha respectively (Fig. 3). Russ et al.
(2015) considered a significant time � reserve status interaction to
demonstrate a reserve effect if fish density increased over time in the
reserve, but not in the control site over time (Russ et al., 2015). As total
biomass follows this pattern in our analysis, we consider CMNTZs to
have a significant positive influence on total fish biomass.
Total biomass at control sites remained relatively level, ranging from
3.4. Reserve effects on fishery targeted and untargeted biomass
Across all sites, both time and reserve status were independently
significant predictors of targeted fish biomass (Table 5). However, these
effects were not interactive i.e. any differences in targeted biomass
5
H. Gilchrist et al.
Ocean and Coastal Management 192 (2020) 105210
Fig. 3. Mean total biomass inside reserves and outside reserves (in controls) in years since reserve closure. All reserve and control sites are aggregated in this
figure. Error bars are 95% confidence interval about the mean.
Table 4
Model-averaged coefficient estimates, standard errors, T-values, and
associated p-values for site-level linear mixed effects models to identify
whether reserve effect had a significant influence on total fish biomass.
Reserve status was included as a fixed factor, transect number and year as
random factors. Significant predictors are highlighted in bold.
Site
Explanatory variable
Estimate (SE)
t value
p value
Riva
Agnorondriake
Ankafotiambe
Tampolove
Ambatohalaomby
Reserve
Reserve
Reserve
Reserve
Reserve
1.18 (0.28)
0.83 (0.17)
0.14 (0.14)
0.26 (0.33)
0.47 (0.30)
4.08
4.89
1.02
0.77
1.57
<0.001
<0.001
0.31
0.44
0.13
status
status
status
status
status
Table 5
Model-averaged coefficient estimates, standard errors, T-values, and associated
p-values for predictors of fish biomass for targeted families (Acanthuridae,
Lethrinidae, Lutjanidae, Siganidae) and untargeted families (Balistidae, Chaetodontidae, Labridae, Pomacanthidae) across the five coral reef CMNTZs in the
Velondriake LMMA. CMNTZ status effects are relative to fished (control) sites.
The explanatory variables presented are those remaining in the minimum
adequate (linear mixed effects) model. Non-significant predictors of fish biomass
that were dropped from the model are also listed for reference (see NS).
between reserves and controls did not change over time, suggesting a
lack of reserve effect on targeted fish biomass and an inherent difference
in targeted biomass between CMNTZs and control sites. Fig. 4 illustrates
this result, with significantly higher biomass of targeted fish families
inside CMNTZs than outside in years zero (n ¼ 5 CMNTZs; 110% � 86),
four (n ¼ 3 CMNTZs; 197% � 118) and five (n ¼ 1 CMNTZ; 274% � 175)
but not in year six (n ¼ 1 CMNTZ; 48% � 175). This is further supported
in Fig. 5, where raw biomass values in year six are comparable at 318
kg/ha (CI95 � 221) and 216 kg/ha (CI95 � 158) inside and outside
reserves respectively.
At the site level, two CMNTZs had a higher biomass of targeted
families when compared with their controls in 2014 and 2015 (Fig. 6).
These CMNTZs were Tampolove (20.92% � 17.35 higher than control
site) and Agnorondriake (16.18% � 12.75 higher than in control site).
On the other hand, CMNTZ protection and years since CMNTZ
establishment had no independent effect on untargeted fish biomass but,
did have an interactive effect (Table 5). Depth was also a significant
predictor of untargeted fish biomass, but not targeted fish biomass
(Table 5). The relative difference in biomass of untargeted fish families
between CMNTZs and control sites differed depending on the year, with
no difference in year zero (Fig. 4), but significantly more untargeted
biomass inside CMNTZs than control sites in years four (86% � 65) and
six (322% � 110), implying an effect of CMNTZs on untargeted fish
families (Russ et al., 2015). Raw values in figure five suggest that the
biomass of untargeted fish in controls declined from 71.5 kg/ha (CI95 �
Response variable
Explanatory variable
Estimate
(SE)
T
value
Targeted Fish
Biomass
Intercept
Status
34.69
2.32
<0.0001
<0.05
7.89
<0.0001
Untargeted Fish
Biomass
Years since
established
Hard coral cover (%)
Distance to landing
site (km)
Depth (m)
Intercept
Status
4.34 (0.13)
0.39
(0.17)
0.23
(0.03)
NS
NS
21.58
1.15
<0.0001
0.28
0.04
0.96
2.01
<0.05
2.09
<0.05
Years
Years*Status
(interaction)
Depth
Distance to landing
site (km)
Hard coral cover (%)
NS
4.56 (0.21)
0.18
(0.16)
0.002
(0.05)
0.12
(0.06)
0.04
(0.02)
NS
P value
NS
23.2) in year zero to 21.5 kg/ha (CI95 � 10.9) in year six whereas
biomass of untargeted families remained relatively stable in reserves at
74.3 kg/ha (CI95 � 22.6) in year zero and 91.0 kg/ha (CI95 � 49.6) in
year six.
In 2014 and 2015, the mean untargeted fish biomass was significantly higher in CMNTZs than controls at Ambatohalaomby and Ankafotiambe at 46% � 17 and 31% � 25 respectively (Fig. 6).
6
H. Gilchrist et al.
Ocean and Coastal Management 192 (2020) 105210
Fig. 4. Response ratios (mean and 95%
confidence intervals) calculated using
biomass of untargeted and targeted fish
families for each year. Ratios were calculated using data across all CMNTZs, positive
values show biomass higher inside the
CMNTZ while negative values show that
biomass is higher in control sites. Confidence
intervals that overlap the vertical line at zero
are not significantly different from zero (i.e.
there is no significant reserve effect). Years
zero and one represent all reserve-control
pairs, years two-four are representative of
three reserves (Agnorondriake, Ankafotiambe and Riva) and, only data from
Agnorondriake is used to calculate response
ratio in years five and six. This is due to the
different closure dates of each CMNTZ
(Table 1).
Fig. 5. Mean un-targeted (left) and targeted (right) fish family biomass inside and outside reserves in years since reserve closure. All control and reserve
sites are aggregated and error bars are 95% confidence interval about the mean.
3.5. Reserve effects on fish species richness and community composition
4. Discussion
PERMANOVA showed no interactive effect of reserve status and time
on family-level fish community structure (Table 6), nor species-level
community structure (Table 7). There was also no independent influence of time and reserve status (Tables 6 and 7).
However, there was a significant, positive interactive effect of
reserve status and time on species richness, alongside depth and hard
coral cover (Table 8). In year zero there were 17.3% (CI95 � 16.2) more
species in control sites than in reserves. Two years after reserve closure,
reserves had 43.3% (CI95 � 16.2) more species than controls, however
in year six there was no difference (Fig. 7) in species richness between
reserves and controls.
Despite the positive reserve effect, there was a significant decline in
species richness over time across all sites (Table 8), with reserves
dropping from 17.1 species (CI95 � 1.3) in year zero to 15.3 species
(CI95 � 1.1) in year six, and controls from 20.7 species (CI95 � 2.6) in
year zero to 17.0 species (CI95 � 3.7) in year six (Fig. 7).
4.1. Effectiveness of the CMNTZs
4.1.1. Effects at a network-level
Overall, this study provides empirical evidence of the long-term
biological effectiveness of community managed no-take zones on
broader reef fish assemblages. Across the network of five CMNTZs
studied within southwest Madagascar’s Velondriake LMMA, total
biomass increased over time when compared with paired control sites,
where biomass remained relatively stable over time. In 2015, the most
recent year for which data were available, overall biomass was 219 �
58% higher within CMNTZs when compared with control sites. The first
significant differences in total biomass between CMNTZs and controls
emerged in year two of protection. By year four, overall biomass was
322 � 66% higher in reserves than in controls, and 189 � 101% in year
six. These differences were due to increases in biomass in CMNTZs and
relatively stable total biomass in controls over time.
In contrast with previous research, the biomass of fish targeted by the
^t�e et al., 2001;
local fishery was not influenced by CMNTZ protection (Co
Mosquera et al., 2000; Tetreault and Ambrose, 2007; Babcock et al.,
2010). The biomass of targeted fish was 110% � 86% higher within
CMNTZs than on control sites at the time of protection, showing no
7
H. Gilchrist et al.
Ocean and Coastal Management 192 (2020) 105210
Fig. 6. Response ratios (mean and 95% confidence intervals) calculated using biomass of targeted and untargeted fish families for each
reserve-control pair. Ratios were calculated using
pooled data from 2014 to 2015. Positive values show
biomass is higher inside the CMNTZ while negative
values show biomass is higher in control sites. Confidence intervals that overlap the vertical line at zero
are not significantly different from zero (i.e. there is
no significant reserve effect).
Table 6
Results of PERMANOVA for family-level biomass community structure.
Response variable
Df
F statistic
R2
P value
Status
Years
Status*Years
Residuals
Total
1
6
6
28
41
1.94808
0.98682
0.76824
0.04813
0.14627
0.11387
0.69173
1.00000
0.051
0.480
0.871
Table 8
Model averaged coefficient estimates, standard errors, T-values and associated
p-values for predictors of fish species richness for all fish families across the five
coral reef CMNTZs in the Velondriake LMMA. CMNTZ status effects are relative
to fished (control) sites. The explanatory variables presented are those remaining in the minimum adequate model. The non-significant predictors (NS)
dropped from the most parsimonious (Linear Mixed Effects) model are also listed
for reference.
Table 7
Results of PERMANOVA for species-level biomass community structure.
Response variable
Df
F statistic
R2
P value
Status
Years
Status*Years
Residuals
Total
1
6
6
28
41
1.75752
0.74029
0.77260
0.04526
0.11438
0.11937
0.72100
1.00000
0.063
0.949
0.920
Response
variable
Explanatory variable
Estimate
(SE)
T value
p value
Species
richness
Intercept
14.83
(1.49)
3.83
(1.78)
¡1.02
(0.31)
1.49 (0.35)
9.94
<0.0001
9.55 (1.80)
0.20 (0.07)
NS
Reserve status
Years
Years*Status
(interaction)
Hard coral cover (%)
Depth
Distance to shore
increase over time as a result of protection. This implies that the differences between CMNTZs and controls in terms of fishery-targeted
biomass are due to site-specific differences unrelated to fishing pressure (Russ et al., 2015; Wilson et al., 2012; Williamson et al., 2014).
The lack of expected improvement in the biomass of targeted families
could be due to the size of the areas (<2 km2) under protection (Edgar
et al., 2014; Green et al., 2015; Claudet et al., 2008). All four targeted
families (Acanthuridae, Lethrinidae, Lutjanidae, Siganidae) have recommended minimum reserves sizes between 0.2 and 10 km (in linear distance of widest point), with all lethrinid species reviewed in Green et al.
(2015) recommended to need a 10 km minimum reserve length (Green
et al., 2015; Chateau and Wantiez, 2009; Kaunda-Arara and Rose, 2004),
and most lutjanid species having a recommended minimum reserve
length of 2 or 6 km (Addis et al., 2008; Luo et al., 2009; McMahon et al.,
2012; Dunlop and Mann, 2012). Acanthurids and siganids have greater
variability between species in terms of minimum recommended reserve
size; between 0.2 and 10 km (Kaunda-Arara and Rose, 2004; Abesamis
2.15
0.06
¡3.29
<0.01
4.24
<0.0001
5.32
2.93
<0.0001
<0.01
and Russ, 2005a; Colin, 2012; Marshell et al., 2011; Samoilys et al.,
2013). Most of these sizes for protection exceed the dimensions of all
five of Velondriake’s CMNTZs. Although Velondriake’s CMNTZs are
large enough to benefit fish populations as a whole (as evidenced by the
increase in total biomass over time), they are not big enough to benefit
species with large home ranges, which coincidentally are also families
preferentially targeted by the fishery (Blue Ventures. Unpublishe, 2015).
Fish families under reduced fishing pressure - Balistidae, Chaetodontidae, Labridae, Pomacanthidae – did exhibit a response to CMNTZ
protection, with no difference at a network level between CMNTZs and
controls in year zero, but 86% (�65%) higher biomass in CMNTZs in
year four, and 322% (�110%) in year six. CMNTZ areas in the case of
these families are large enough to provide protection however, these
untargeted families did not see any improvement in biomass, instead
their biomass remains stable in CMNTZs and drops in controls. Responses to protection in untargeted fish families usually take 13 years
8
H. Gilchrist et al.
Ocean and Coastal Management 192 (2020) 105210
southwest Madagascar, a region that is particularly vulnerable to
climate impacts such as rising sea surface temperatures (Hobday and
Pecl, 2013; Lough, 2012), and increases in strength and frequency of
cyclones (Jury et al., 1999; Emanuel, 2007; Knutson et al., 2010).
However, the idea that protection from fishing could improve resilience
of reef communities to climate impacts remains controversial (Bates
et al., 2019; Bruno et al., 2019; McClanahan et al., 2012; Graham et al.,
2013) therefore, this hypothesis would have to be explored further,
supplemented with cyclone impact or sea surface temperature data, to
draw any firm conclusions.
4.1.2. Site-level effects of reserve protection
While the effects of protection at the network level were generally
positive for total biomass, the picture at the site level was more mixed,
with CMNTZ-specific factors such as years established, compliance,
location, size and enforcement likely contributing to overall effectiveness. Estimates of the time required for NTZs to improve fish biomass
differ, with benefits first observed anywhere between one and five years
after closure (Babcock et al., 2010; Halpern and Warner, 2002; Russ and
Alcala, 2003). Two of the five CMNTZs studied (Tampolove and
Ambatohalaomby) had been closed for less than 2 years at the time of
data analysis. The Tampolove site showed a significant reserve effect on
total biomass, whilst Ambatohalaomby did not; this site may need a
longer protection period before any effects on total biomass are detected. Of the three remaining reserves (all of which have been in existence
for more than two years), two showed significantly positive responses in
total biomass to protection, with total biomass 555% higher at Riva and
189% times higher at Agnorondriake than at controls.
Ankafotiambe had no effect on total biomass. The lack of reserve
effect is likely to be due to a combination of enforcement issues and
biogeography. While the effective Riva and Agnorondriake CMNTZs
protect entire patch reefs, Ankafotiambe covers a small area of barrier
reef that continues for several kilometres south of the reserve boundary.
In this continuous habitat, motile species may be more likely to cross
into fished areas leading to a reduced ability to detect any reserve effect
on biomass (Mora et al., 2006; Edgar et al., 2014; Edgar, 2011). Additionally, the site is close to an island regularly visited by migrating
fishers from settlements elsewhere along the coast (Cripps and Gardner,
2016) who are unlikely to be familiar with local protection laws or
CMNTZ boundaries (Ferse et al., 2010). If compliance with the local laws
can be improved at Ankafotiambe, the resulting benefits that accrue to
fishers may be higher, since the continuous nature of the habitat could
create a larger spillover area (Da Silva et al., 2015; Forcada et al., 2009).
Fig. 7. Response ratios (mean and 95% confidence intervals) calculated
using species richness for each year, and grand means for each year in
controls and reserves in number of species per transect. Ratios were
calculated using data across all CMNTZs, positive values show species richness
higher inside the CMNTZ while negative values show that species richness is
higher in control sites. Confidence intervals that overlap the vertical line at zero
are not significantly different from zero (i.e. there is no significant reserve effect). Years 0 and 1 represent all reserve-control pairs, years two-four are
representative of three reserves (Agnorondriake, Ankafotiambe and Riva) and,
only data from Agnorondriake is used to calculate response ratio in years five
and six. This is due to the different closure dates of each CMNTZ (Table 1).
(�2) to accrue through trophic cascading or the recovery of habitat in
the absence of destructive fishing methods (Babcock et al., 2010; Molloy
et al., 2009). The CMNTZs in Velondriake have been protected for a
maximum of six years, therefore improvement in biomass of untargeted
families should not be expected for another four to 5 years at least
(Babcock et al., 2010), although it is likely that at present the positive
contribution of the CMNTZs is to afford protection to these fish families
from the declines observed in other nearby fished areas or control sites
(Ahmadia et al., 2015; Selig and Bruno, 2010).
There was no change in reef fish community composition as a result
of reserve protection despite increases in total biomass, this corroborates
findings in previous research (Nagelkerken et al., 2012; Mellin et al.,
2016; Emslie et al., 2015). As this study shows, changes in biomass can
occur within a few years of reserve protection (Cinner et al., 2005;
Babcock et al., 2010). However, similar to the biomass of untargeted fish
families, shifts in community composition happen on longer timescales,
sometimes taking decades and going through multiple phases before
stabilising (Babcock et al., 2010; Micheli et al., 2004; Mcclanahan and
Graham, 2015). Therefore, the reserves in this study, with a maximum
age of six years, are too young to show any change in community
composition.
Species richness declined over time in both reserves and controls.
The strong correlation of species richness with hard coral cover suggests
that this decline is due to losses in hard coral cover (Williamson et al.,
2014; McClanahan and Arthur, 2001; Jones et al., 2004). Reserve effects
on species richness are reported in the literature (Russ and Alcala, 2011;
McClanahan and Arthur, 2001; McClanahan et al., 2007), however
species richness did not recover, but did remain higher than species
richness in control sites. Species richness is a proven predictor of fish
biomass (McClanahan et al., 2007, 2011; Duffy et al., 2016), suggesting
that the effect of protection from fishing on species richness may have
contributed towards the increase in total biomass at the same sites.
Furthermore, the higher species richness in CMNTZs compared with
controls implies that reserve protection is helping to prevent a loss in
functional redundancy across species, therefore mitigating declines in
resilience to climate impacts (Duffy et al., 2016; Wilson et al., 2009;
Micheli et al., 2012). This could be good news for coral reefs in
4.1.3. Environmental drivers of biomass in Velondriake
Across all sites, depth was a significant predictor of total fish biomass
and the biomass of untargeted fish families, but not targeted families.
Depth is a widely recognised driver of fish biomass (Guidetti et al., 2005;
Hind et al., 2010; Cohen et al., 2014; Harding et al., 2006; Almany et al.,
2009). In this case, increasing fish biomass with depth can be explained
by reduced wave action and swell effects at depth. The larger body size
and swimming ability of targeted families allow them to inhabit the high
water velocity environment nearer the surface (Fulton and Bellwood,
2004; Friedlander et al., 2003; Fulton et al., 2005).
Hard coral cover was not a significant predictor of biomass in any
models, in contrast with the literature (Nash et al., 2013; Gratwicke and
Speight, 2005; Daw et al., 2011; Emslie et al., 2014). This is particularly
surprising for untargeted fish families (e.g. cheatodontids and pomacanthids) that typically have a strong reliance on coral cover (Pratchett
€
and Rajasuriya, 1998), though
et al., 2008; Coker et al., 2014; Ohman
less so for the larger-bodied targeted families that are more likely to
respond to structural complexity or predator abundance instead of hard
coral cover (Wilson et al., 2006, 2009). We did not have structural
complexity data available at a high enough resolution to explore the
interaction of fish biomass and habitat any further. Research to disentangle environmental drivers of fish biomass in southwest Madagascar is
9
H. Gilchrist et al.
Ocean and Coastal Management 192 (2020) 105210
recommended, with a particular focus on habitat complexity and
benthic composition.
We suggest that Ankafotiambe reserve be expanded to cover at least
2 km2; the majority of this site’s reef area (Harrison et al., 2012; Almany
et al., 2009; Shanks et al., 2003). Additionally, the Velondriake association should consider increasing the area protected by other CMNTZs in
the network in line with minimum areas recommended to protect key
fishery species. These key species or families could be identified by the
Velondriake LMMA Association with the support of ongoing fishery
landings data.
Furthermore, although this paper shows promising signs of improved
total biomass inside CMNTZs, we do not have the information necessary
to draw conclusions on whether this higher total biomass can contribute
to long-term fishery benefits through exportable biomass. Therefore, the
monitoring programme should be adjusted to assess if there is any
spillover of fishable biomass outside the CMNTZ boundaries. We also
suggest that this analysis is revisited every 5 years to regularly evaluate
reserve effectiveness and advise the Velondriake management association on how best to adaptively manage the CMNTZ network.
4.2. How do these CMNTZs compare to other protected areas?
It is challenging to put these findings into context within the broader
western Indian Ocean (WIO) region. In a regional analysis, Rocliffe et al.
(2014) identified 62 LMMAs covering 11 329 km2 but cautioned that the
recently established and informal nature of many initiatives meant that
evidence in support of biological outcomes was scarce (Rocliffe et al.,
2014). At least one of the LMMAs identified by that assessment has
shown a positive effect within a CMNTZ subsequently. Da Silva et al.
(2015) reported significantly higher abundance inside the CMNTZ at
Vamizi island, Mozambique than at control sites (Da Silva et al., 2015).
However, the study does not quantify the magnitude of difference,
making direct comparison difficult. There is also some mention in the
grey literature of an increase in fish abundance and coral cover at the
Kuruwitu LMMA in Kenya, though a lack of detailed, peer-reviewed
information makes further assessment challenging (Nelson, 2012).
Outside the WIO, some authors have found that LMMAs can increase
the biomass of the overall species assemblage (Bartlett et al., 2009;
Cinner et al., 2005; Bonaldo et al., 2017; Clements et al., 2012; Jupiter
and Egli, 2011). In Fiji, for example, Bonaldo et al. (2017) and Clements
et al. (2012) assessed the biological effectiveness of small CMNTZs, both
found that the reserve areas exhibited greater biomass when compared
with control sites (Bonaldo et al., 2017; Clements et al., 2012). However,
these findings are based on single-point-in-time control-impact (CI) assessments, and as such, suffer from pseudoreplication (Hurlbert, 1984,
2009). While the sites being compared may be alike, they will not be
identical, and it is very probable that there would have been statistically
significant differences between them prior to the CMNTZs being established (Osenberg et al., 2006). Such a design therefore lacks the replication necessary to logically disentangle the effects of the CMNTZ from
other sources of spatial variation, and risks the effects being overestimated (Osenberg et al., 2011; Claudet and Guidetti, 2010).
The most robust evidence for the long-term effectiveness of local or
co-management initiatives comes from the Philippines, where multiple
studies over several decades have demonstrated that biomass and density of reef fish within the CMNTZ at Apo Island are significantly higher
than at control sites (Russ and Alcala, 1996, 1998, 2003, 2011; Russ
et al., 2004, 2005; Abesamis and Russ, 2005b; Walmsley and White,
2003; White, 1988). However, while Apo Island was a CMNTZ until the
mid-1990s, it has since switched to a top-down government-led model,
meaning that no recent data on CMNTZ effectiveness exists at this site
(Hind et al., 2010). These examples notwithstanding, despite the estimated thousands of established LMMAs across the WIO and Western and
Central Pacific, there is little empirical evidence that they can achieve
long-term biological goals (Cohen et al., 2014).
4.4. Implications for broader fisheries management and conservation
Deprived of their community-managed context, the core findings of
this study – that a network of no-take zones experienced an increase in
total fish biomass over time relative to control sites with statistically
similar habitats – are hardly novel. Several studies and meta analyses
have shown that such areas can increase fish biomass and abundance
(Halpern, 2003; Claudet et al., 2008; Molloy et al., 2009; Lester et al.,
2009b; Stewart et al., 2009).
However, our findings present novel evidence that community-led
marine management can drive fish biomass recovery to levels comparable with NTZs that employ other governance mechanisms. For
example, Gill et al. (2017) found that biomass within MPAs was on
average 82% higher than at control sites, vs. a peak of 322% across three
of Velondriake’s CMNTZs after four years of protection (Gill et al.,
2017). The mean biomass levels of Velondriake’s CMNTZs (786 � 247
kg/ha) are similar to those in other NTZs within the WIO (mean 731
kg/ha; range 270 kg/ha – 1180 kg/ha), and are over the estimated
biomass needed to maintain species diversity, fish life histories and,
maximum sustainable yields of fish stocks (McClanahan et al., 2011;
Graham and Mcclanahan, 2013). This is despite a substantially lower
biomass on fished reefs in Velondriake (247 � 33 kg/ha) than both the
documented threshold indicating a degraded ecological state (300 kg/ha
(McClanahan et al., 2011)), and the WIO mean for fished reefs (581
kg/ha, range 51 kg/ha-1463 kg/ha (Graham and Mcclanahan, 2013)).
From an economic standpoint, comparisons with MPAs are favourable to Velondriake. In a global analysis of MPA costs, Balmford et al.
(2004) found that median expenditure per unit area was USD$2698
(Balmford et al., 2004). Adjusting for inflation, this figure is approximately 10 times higher than Blue Ventures’ most generous internal cost
estimates for Velondriake (Blue Ventures unpublished data). However,
although costs of running the LMMA are relatively low compared to
top-down governance, they are predominantly borne by Blue Ventures,
ultimately limiting the long-term sustainability of Velondriake unless
steps are taken to improve the capacity of the LMMA association to seek
and manage funding. Currently, Blue Ventures and the Velondriake
Association are exploring ways to grow their financial autonomy
through multiple avenues, including grants and sustainable local
enterprise.
Figures for coastal MPA coverage lag far behind those for offshore
areas at a global level (Maire et al., 2016; Thomas et al., 2014). In isolated parts of the tropics, such as southwest Madagascar, a high degree
of coastal poverty results in strong dependence on fishing as a primary or
sole source of food and income (Walmsley et al., 2006; Barnes-Mauthe
et al., 2013). In this context, fishers rarely perceive the short-term costs
of marine management to outweigh future benefits, particularly as these
benefits are viewed as uncertain (Buxton et al., 2014), diffuse (Hanna,
2004), slow to accrue (Russ and Alcala, 2011), and potentially shared
4.3. Priority management and research actions for velondriake
While the findings of our study suggest that the CMNTZ network in
Velondriake is effective at increasing total biomass of fish, CMNTZs are
not large enough to protect families more prone to fishing pressure, and
responses in total biomass to protection varied considerably between
sites, with some showing no benefits from CMNTZ protection. In light of
this, Velondriake’s CMNTZ network should be reviewed to consider
including CMNTZs that better reflect the variation in size and biogeography of the LMMA’s reefs (Harding et al., 2006; Almany et al., 2009)
and, to increase the area covered by existing CMNTZs to better protect
fish families that are preferentially targeted by the fishery (Green et al.,
2015). Without increasing the area of these CMNTZs, the Velondriake
LMMA is unlikely to see significant fishery benefits coming about as a
result of the CMNTZ network (Edgar et al., 2014; Green et al., 2015), and
could risk falling short of Velondriake’s objective to safeguard traditional fishing livelihoods (Harris, 2011).
10
H. Gilchrist et al.
Ocean and Coastal Management 192 (2020) 105210
Acknowledgements
with others who may not have invested in protection efforts (Oliver
et al., 2015; Pomeroy et al., 2007). Considering these perceptions, marine protection efforts are unlikely to succeed without the involvement
of fishers and community members. If we are to act on calls to increase
area and improve effectiveness of coastal management efforts, the
establishment of LMMAs is arguably the only viable option along
errez et al., 2011).
low-income coastlines such as these (Guti�
We are grateful to the many field scientists and fishers who aided in
the collection of data for this study in addition to all staff and volunteers
involved in Blue Ventures Expeditions programme, without whom the
monitoring and our final dataset would not have been possible.
Specific thanks go to the Velondriake LMMA Management Association for allowing us to survey inside their CMNTZs, Bic Manahira and
James Paul who were involved in counting the majority of the 77,906
fishes included in our analysis, and boat captains through the years
whose invaluable knowledge of the southwest coast of Madagascar has
kept our colleagues safe in the water. And finally, Alasdair Harris for his
constructive comments and review of the manuscript in its final stages.
This research did not receive any specific grant from funding
agencies in the public, commercial, or not-for-profit sectors.
5. Conclusion
LMMAs are an increasingly important component of coastal marine
resource management strategies across the Indo-Pacific. However,
empirical evidence of their long-term biological effectiveness has been
lacking. This is the case particularly in the WIO, where an establishment
and spread of LMMAs has occurred in the last 12 years. Our findings
suggest that such community-driven initiatives can result in long-term
improvements in overall fish biomass from a severely depleted state
and, may even offer the opportunity to catalyse broader conservation
and management at the local level.
As pressures on marine resources intensify with growing human
populations and increasing commercialisation of traditional coastal
fisheries through seafood markets, the potential of LMMAs to offer a
win-win approach that benefits people and nature alike is such that they
are receiving increasing attention from international agencies, NGOs
and the donor community (personal observation SR). Yet the growing
trend in support of LMMAs among the donor community and environmental sector is in notable contrast to available empirical evidence of
LMMA effectiveness, and co-management has seen only scant scrutiny in
the academic literature (e.g. (Cinner et al., 2012; Sen and Nielsen, 1996;
Stafford, 2018)). If the effectiveness of LMMAs is to be defended and
their use to continue to rise, then arguments as to their effectiveness (or
otherwise) need to be based on the soundest possible evidence. As
Lawton (1996) has pointed out, evidence from research that involves
experimental falsification of hypotheses tends to be much more
persuasive to others than modelling studies, observations, logical
argument and anecdote (Lawton, 1996). Experimental falsification demands high standards of both data collection and experimental design if
such evidence is to stand up to scrutiny. However, achieving those high
standards becomes challenging in the informal, data-poor and
low-capacity contexts in which many LMMAs exist, where the resources
necessary to sustain both effective management and the collection and
evaluation of robust evidence vital to such adaptive management are
often lacking. Furthermore, the recent nature of many initiatives,
particularly in the WIO, has also largely precluded the collection of
comprehensive, long-term datasets that can separate claimed effects of
LMMAs from other confounding variables (like habitat) convincingly.
Undoubtedly, there is an urgent need for better multidisciplinary
evidence of the effects of LMMAs, particularly the nature of the relationship between the structure, objectives and management of an LMMA
and its outcomes, as well as the factors that cause such initiatives to
spread to other communities. But while it is certainly desirable to
improve our understanding in these areas, lack of formal proof of effects
should not be used to delay or excuse co-management of inshore fisheries and marine environments in pursuit of sustainability objectives. In
data-poor, informal and increasingly threatened systems, suffering from
the challenges of poverty, acute marine resource dependence, extreme
climate vulnerability, and weak institutions, the perils of inaction are
likely to outweigh the costs of action.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.ocecoaman.2020.105210.
References
Abesamis, R.A., Russ, G.R., 2005a. Density-dependent spillover from a marine reserve:
long-term evidence. Ecol. Appl. 15, 1798–1812. https://doi.org/10.1890/05-0174.
Abesamis, R.A., Russ, G.R., 2005b. Density-dependent spillover from a marine reserve:
long-term evidence. Ecol. Appl. https://doi.org/10.1890/05-0174.
Ackerman, J.L., Bellwood, D.R., 2000. Reef fish assemblages: a re-evaluation using
enclosed rotenone stations. Mar. Ecol. Prog. Ser. 206, 227–237.
Addis, D.T., Patterson, W.F., Dance, M.A., 2008. Site fidelity and movement of reef fishes
tagged at unreported artificial reef sites off NW Florida. In: Proc 60th Gulf Caribb
Fish Inst.
Ahmadia, G.N., Glew, L., Provost, M., Gill, D., Hidayat, N.I., Mangubhai, S., et al., 2015.
Integrating impact evaluation in the design and implementation of monitoring
marine protected areas. Philos Trans R Soc B Biol Sci. https://doi.org/10.1098/
rstb.2014.0275.
Almany, G.R., Connolly, S.R., Heath, D.D., Hogan, J.D., Jones, G.P., McCook, L.J., et al.,
2009. Connectivity, biodiversity conservation and the design of marine reserve
networks for coral reefs. Coral Reefs 28, 339–351. Available: http://link.springer.
com/content/pdf/10.1007%2Fs00338-009-0484-x.pdf.
Alvarez-Filip, L., Gill, J.A., Dulvy, N.K., Perry, A.L., Watkinson, A.R., C^
ot�
e, I.M., 2011.
Drivers of region-wide declines in architectural complexity on Caribbean reefs. Coral
Reefs 30, 1051–1060. https://doi.org/10.1007/s00338-011-0795-6.
Babcock, R.C., Shears, N.T., Alcala, A.C., Barrett, N.S., Edgar, G.J., Lafferty, K.D., et al.,
2010. Decadal trends in marine reserves reveal differential rates of change in direct
and indirect effects. Proc. Natl. Acad. Sci. U. S. A. 107, 18256–18261. Available:
http://www.pnas.org/cgi/content/long/107/43/18256.
Balmford, A., Gravestock, P., Hockley, N., McClean, C.J., Roberts, C.M., 2004. The
worldwide costs of marine protected areas. Proc. Natl. Acad. Sci. U. S. A. 101,
9694–9697. https://doi.org/10.1073/pnas.0403239101.
Barnes-Mauthe, M., Oleson, K.L.L., Zafindrasilivonona, B., 2013. The total economic
value of small-scale fisheries with a characterization of post-landing trends: an
application in Madagascar with global relevance. Fish. Res. 147, 175–185.
Bartlett, C.Y., Manua, C., Cinner, J., Sutton, S., Jimmy, R., South, R., et al., 2009.
Comparison of outcomes of permanently closed and periodically harvested coral reef
reserves. Conserv. Biol. 23, 1475–1484. https://doi.org/10.1111/j.15231739.2009.01293.x.
Bates, D., M€
achler, M., Bolker, B.M., Walker, S.C., 2015. Fitting linear mixed-effects
models using lme4. J. Stat. Software 67, 1–48. Available: http://cran.r-project.org/p
ackage¼lme4%5Cnhttp://www.jstatsoft.org/index.php/jss/article/view/v067i01/v
67i01.pdf.
Bates, A.E., Cooke, R.S.C., Duncan, M.I., Edgar, G.J., Bruno, J.F., Benedetti-Cecchi, L.,
et al., 2019. Climate resilience in marine protected areas and the ‘Protection
Paradox.’. Biol. Conserv. https://doi.org/10.1016/j.biocon.2019.05.005.
Bell, J.D., Craik, G.J.S., Pollard, D.A., Russell, B.C., 1985. Estimating length frequency
distributions of large reef fish underwater. In: Coral Reefs, vol. 4. Springer-Verlag,
pp. 41–44.
Birt, M.J., Harvey, E.S., Langlois, T.J., 2012. Within and between day variability in
temperate reef fish assemblages: learned response to baited video. J. Exp. Mar. Biol.
Ecol. 416–417, 92–100. https://doi.org/10.1016/j.jembe.2012.02.011.
Blue Ventures. Unpublished Data. London, 2015.
Bonaldo, R.M., Pires, M.M., Guimar~
aes, P.R., Hoey, A.S., Hay, M.E., 2017. Small marine
protected areas in Fiji provide refuge for reef fish assemblages, feeding groups, and
corals. PloS One 12. https://doi.org/10.1371/journal.pone.0170638.
Bruno, J.F., C^
ot�
e, I.M., Toth, L.T., 2019. Climate change, coral loss, and the curious case
of the parrotfish paradigm: why don’t marine protected areas improve reef
resilience? Ann Rev Mar Sci. https://doi.org/10.1146/annurev-marine-010318095300.
Declaration of competing interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
11
H. Gilchrist et al.
Ocean and Coastal Management 192 (2020) 105210
Burke, L., Reytar, K., Spalding, M., Perry, A., 2011. Reefs at Risk Revisited. World
Resources Institute, Washington, D.C.
Buxton, C.D., Hartmann, K., Kearney, R., Gardner, C., 2014. When is spillover from
marine reserves likely to benefit fisheries? PloS One 9, e107032. https://doi.org/
10.1371/journal.pone.0107032.
Chateau, O., Wantiez, L., 2009. Movement patterns of four coral reef fish species in a
fragmented habitat in New Caledonia: implications for the design of marine
protected area networks. ICES (Int. Counc. Explor. Sea) J. Mar. Sci. https://doi.org/
10.1093/icesjms/fsn165.
Cinner, J.E., Marnane, M.J., Mcclanahan, T.R., 2005. Conservation and community
benefits from traditional coral reef management at Ahus Island, Papua New Guinea.
Conserv. Biol. 19, 1714–1723. https://doi.org/10.1111/j.1523-1739.2005.00209.xi1.
Cinner, J.E., McClanahan, T.R., MacNeil, M.A., Graham, N.A.J., Daw, T.M.,
Mukminin, A., et al., 2012. Comanagement of coral reef social-ecological systems.
Proc Natl Acad Sci. Proceedings of the National Academy of Sciences 109,
5219–5222. https://doi.org/10.1073/pnas.1121215109.
Clarke, K.R., 1993. Non-parametric multivariate analyses of changes in community
structure. Aust. J. Ecol. 18, 117–143. https://doi.org/10.1111/j.1442-9993.1993.
tb00438.x.
Clarke, R.K., Gorley, R.N., 2015. PRIMER V7: User Manual/tutorial. PRIMER-E,
Plymouth.
Claudet, J., Guidetti, P., 2010. Improving assessments of marine protected areas. Aquat.
Conserv. Mar. Freshw. Ecosyst. 20, 239–242. https://doi.org/10.1002/aqc.1087.
Claudet, J., Osenberg, C.W., Benedetti-Cecchi, L., Domenici, P., García-Charton, J.,
� et al., 2008. Marine reserves: size and age do matter. Ecol. Lett. 11,
P�
erez-Ruzafa, A.,
481–489.
Clements, C., Bonito, V., Grober-Dunsmore, R., Sobey, M., 2012. Effects of small, Fijian
community-based marine protected areas on exploited reef fishes. Mar. Ecol. Prog.
Ser. 449, 233–243. https://doi.org/10.3354/meps09554.
Cohen, P.J., Jupiter, S.D., Weeks, R., Tawake, A., Govan, H., 2014. Is community-based
fisheries management realising multiple objectives? Examining evidence from the
literature. SPC Tradit Mar Resour Manag Knowl Inf Bull 34, 3–12.
Coker, D.J., Wilson, S.K., Pratchett, M.S., 2014. Importance of live coral habitat for reef
fishes. Rev. Fish Biol. Fish. 89–126.
Colin, P.L., 2012. Timing and location of aggregation and spawning in reef fishes. In:
Reef Fish Spawning Aggregations: Biology, Research and Management. https://doi.
org/10.1007/978-94-007-1980-4_5.
C^
ot�
e, I.M., Mosqueira, I., Reynolds, J.D., 2001. Effects of marine reserve characteristics
on the protection of fish populations: a meta-analysis. J. Fish. Biol. 59, 178–189.
https://doi.org/10.1111/j.1095-8649.2001.tb01385.x.
Crawley, M.J., 2007. The R Book. The R Book. John Wiley and Sons.
Cripps, G., Gardner, C.J., 2016. Human migration and marine protected areas: insights
from Vezo Fishers in Madagascar. Geoforum 74, 49–62.
Da Silva, I.M., Hill, N., Shimadzu, H., Soares, A.M.V.M., Dornelas, M., 2015. Spillover
effects of a community-managed marine reserve. PloS One 10. https://doi.org/
10.1371/journal.pone.0111774.
Daw, T.M., Cinner, J.E., McClanahan, T.R., Graham, N.A.J., Wilson, S.K., 2011. Design
factors and socioeconomic variables associated with ecological responses to fishery
closures in the western Indian ocean. Coast. Manag. 39, 412–424. Available: http://
www.tandfonline.com/doi/abs/10.1080/08920753.2011.589224%5Cnhttp://
ezproxy.library.nyu.edu:2121/doi/full/10.1080/08920753.2011.589224%5Cnh
ttp://ezproxy.library.nyu.edu:2121/doi/pdf/10.1080/08920753.2011.589224.
Duffy, J.E., Lefcheck, J.S., Stuart-Smith, R.D., Navarrete, S.A., Edgar, G.J., 2016.
Biodiversity enhances reef fish biomass and resistance to climate change. Proc. Natl.
Acad. Sci. U. S. A. https://doi.org/10.1073/pnas.1524465113.
Dunlop, S.W., Mann, B.Q., 2012. Summary of Tag and Recapture Data for Coral Reef
Associated and Tropical Pelagic Game Fishes Caught along the Southern African
Coastline: 1984-2011. Data Report No 4. Oceanographic Research Institute (ORI),
Durban.
Edgar, G.J., 2011. Does the global network of marine protected areas provide an
adequate safety net for marine biodiversity? Aquat. Conserv. Mar. Freshw. Ecosyst.
21, 313–316.
Edgar, G.J., Stuart-Smith, R.D., Willis, T.J., Kininmonth, S., Baker, S.C., Banks, S., et al.,
2014. Global conservation outcomes depend on marine protected areas with five key
features. Nature 506, 216–220. https://doi.org/10.1038/nature13022.
Emanuel, K., 2007. Environmental factors affecting tropical cyclone power dissipation.
J. Clim. 20, 5497–5509. https://doi.org/10.1175/2007JCLI1571.1.
Emslie, M.J., Cheal, A.J., Johns, K.A., 2014. Retention of habitat complexity minimizes
disassembly of reef fish communities following disturbance: a large-scale natural
experiment. PloS One 9, 105384. https://doi.org/10.1371/journal.pone.0105384.
Emslie, M.J., Logan, M., Williamson, D.H., Ayling, A.M., MacNeil, M.A., Ceccarelli, D.M.,
2015. Expectations and outcomes of reserve network performance following Rezoning of the great barrier reef marine park. Curr. Biol. 25, 983–992.
English, S.A., Baker, V.J., Wilkinson, C.R., 1997. Survey Manual for Tropical Marine
Resources. Australian Institute of Marine Science Townsville.
Facon, M., Pinault, M., Obura, D., Pioch, S., Pothin, K., Bigot, L., et al., 2016.
A comparative study of the accuracy and effectiveness of Line and Point Intercept
Transect methods for coral reef monitoring in the southwestern Indian Ocean
islands. Ecol. Indicat. 60, 1045–1055. https://doi.org/10.1016/j.
ecolind.2015.09.005.
~ ez Costa, M., M�
~ ez, K.S., Adhuri, D.S., Glaser, M., 2010. Allies, not
Ferse, SC.a., M�
an
an
aliens: increasing the role of local communities in marine protected area
implementation. Environ. Conserv. 37, 23–34. https://doi.org/10.1017/
S0376892910000172.
Forcada, A., Valle, C., Bonhomme, P., Criquet, G., Cadiou, G., Lenfant, P., et al., 2009.
Effects of habitat on spillover from marine protected areas to artisanal fisheries. Mar.
Ecol. Prog. Ser. 379, 197–211. https://doi.org/10.3354/meps07892.
Friedlander, A.M., Brown, E.K., Jokiel, P.L., Smith, W.R., Rodgers, K.S., 2003. Effects of
habitat, wave exposure, and marine protected area status on coral reef fish
assemblages in the Hawaiian archipelago. Coral Reefs 22, 291–305. https://doi.org/
10.1007/s00338-003-0317-2.
Froese, R., Pauly, D., 2016. FishBase. Internet. World Wide Web electronic publication
[cited 6 Feb 2016]. Available: www.fishbase.org.
Fulton, C.J., Bellwood, D.R., 2004. Wave exposure, swimming performance, and the
structure of tropical and temperate reef fish assemblages. Mar. Biol. https://doi.org/
10.1007/s00227-003-1216-3.
Fulton, C.J., Bellwood, D.R., Wainwright, P.C., 2005. Wave energy and swimming
performance shape coral reef fish assemblages. Proc R Soc B Biol Sci. https://doi.
org/10.1098/rspb.2004.3029.
Gardner, C.J., Rocliffe, S., Gough, C., Levrel, A., Singleton, R.L., Vincke, X., et al., 2017.
Value chain challenges in two community-managed fisheries in western Madagascar:
insights for the small-scale fisheries guidelines. In: Jentoft, S., Chuenpagdee, R.,
Barrag�
an-Paladines, M.J., Franz, N. (Eds.), The Small-Scale Fisheries Guidelines.
Springer International Publishing, Cham, pp. 335–354.
Gell, F., Roberts, C.M., 2003. Benefits beyond boundaries: the fishery effects of marine
reserves. Trends Ecol. Evol. 18, 448–455. https://doi.org/10.1016/S0169-5347(03)
00189-7.
Gill, D.A., Mascia, M.B., Ahmadia, G.N., Glew, L., Lester, S.E., Barnes, M., et al., 2017.
Capacity shortfalls hinder the performance of marine protected areas globally.
Nature 543, 665–669. https://doi.org/10.1038/nature21708.
Gough, C., Thomas, T., Humber, F., Harris, A., Cripps, G., Peabody, S., 2009. Vezo
Fishing: an Introduction to the Methods Used by Fishers in Andavadoaka Southwest
Madagascar. London.
Govan, H., Tawake, A., 2009. Status and Potential of Locally-Managed Marine Areas in
the South Pacific: Meeting Nature Conservation and Sustainable Livelihood Targets
through Wide-Spread Implementation of LMMAs. Coral Reef Initiatives for the
Pacific, New Calendonia.
Graham, N. a J., Mcclanahan, T.R., 2013. The last call for marine wilderness? Bioscience
63, 397–402. Available: http://bioscience.oxfordjournals.org/cgi/doi/10.1525/bio
.2013.63.5.13.
Graham, N.A.J., Wilson, S.K., Jennings, S., Polunin, N.V.C., Bijoux, J.P., Robinson, J.,
2006. Dynamic fragility of oceanic coral reef ecosystems. Proc. Natl. Acad. Sci. Unit.
States Am. 103, 8425–8429. https://doi.org/10.1073/pnas.0600693103.
Graham, N.A.J., Bellwood, D.R., Cinner, J.E., Hughes, T.P., Norstr€
om, A.V., Nystr€
om, M.,
2013. Managing resilience to reverse phase shifts in coral reefs. Front. Ecol. Environ.
541–548. https://doi.org/10.1890/120305.
Gratwicke, B., Speight, M.R., 2005. Effects of habitat complexity on Caribbean marine
fish assemblages. Mar Ecol Ser 292, 301–310. https://doi.org/10.3354/
meps292301.
Green, A.L., Maypa, A.P., Almany, G.R., Rhodes, K.L., Weeks, R., Abesamis, R.A., et al.,
2015. Larval dispersal and movement patterns of coral reef fishes, and implications
for marine reserve network design. Biol. Rev. https://doi.org/10.1111/brv.12155.
Guidetti, P., Bussotti, S., Boero, F., 2005. Evaluating the effects of protection on fish
predators and sea urchins in shallow artificial rocky habitats: a case study in the
northern Adriatic Sea. Mar. Environ. Res. 59, 333–348. https://doi.org/10.1016/j.
marenvres.2004.05.008.
Gust, N., Choat, J.H., McCormick, M.I., 2001. Spatial variability in reef fish distribution,
abundance, size and biomass: a multi-scale analysis. Mar. Ecol. Prog. Ser. 214,
237–251. https://doi.org/10.3354/meps214237.
Gutierrez, N.L., Hilborn, R., Defeo, O., 2011. Leadership, social capital and incentives
promote successful fisheries. Nature 470, 386–389.
Guti�errez, N.L., Hilborn, R., Defeo, O., 2011. Leadership, social capital and incentives
promote successful fisheries. Nature 470, 386–389. https://doi.org/10.1038/
nature09689.
Halford, A.R., Thompson, A.A., 1996. Visual Census Surveys of Reef Fish. Australian
Institute of Marine Science Townsville, Australia.
Halpern, B.S., 2003. The impact of marine reserves: do reserves work and does reserve
size matter? Ecol. Appl. 13, 117–137.
Halpern, B.S., 2014. Making marine protected areas work. Nature 506, 167–168. https://
doi.org/10.1038/nature13053.
Halpern, B.S., Warner, R.R., 2002. Marine reserves have rapid and lasting effects. Ecol.
Lett. 5, 361–366.
Hamilton, S.L., Caselle, J.E., Malone, D.P., Carr, M.H., 2010. Incorporating biogeography
into evaluations of the Channel Islands marine reserve network. Proc. Natl. Acad.
Sci. U. S. A. 107, 18272–18277. https://doi.org/10.1073/pnas.0908091107.
Hanna, S., 2004. The economics of protected areas in marine fisheries management: an
overview of the issues. In: American Fisheries Society Symposium, pp. 259–265.
Harding, S., Randriamanantsoa, B., Hardy, T., Curd, A., 2006. Coral Reef Monitoring and
Biodiversity Assessment to Support the Planning of a Proposed MPA at Andavadoaka
[Internet]. Toliara, Madagascar. Available: http://blueventures.org/publicatio
n/coral-reef-monitoring-and-biodiversity-assessment-to-support-the-planning-o
f-a-proposed-mpa/.
Harmelin-Vivien, M., Harmelin, J.G., Chauvet, C., Duval, C., Galzin, R., Lejeune, P., et al.,
1985. Evaluation visuelle des peuplements et populations de Poissons. M�ethodes et
probl�emes. Rev dEcoloie Terre vie. 40, 467–539. Available: http://cat.inist.fr/?
aModele¼afficheN&cpsidt¼8512979.
Harris, A.R., 2011. Out of sight but no longer out of mind: a climate of change for marine
conservation in Madagascar. Madagascar Conserv Dev 6.
12
H. Gilchrist et al.
Ocean and Coastal Management 192 (2020) 105210
Harrison, H.B., Williamson, D.H., Evans, R.D., Almany, G.R., Thorrold, S.R., Russ, G.R.,
et al., 2012. Larval export from marine reserves and the recruitment benefit for fish
and fisheries. Curr. Biol. 22, 1023–1028.
Hill, J., Wilkinson, C., 2004. Methods for ecological monitoring of coral reefs. Aust Inst
Mar Sci Townsv 117. https://doi.org/10.1017/CBO9781107415324.004.
Hind, E.J., Hiponia, M.C., Gray, T.S., 2010. From community-based to centralised
national management-A wrong turning for the governance of the marine protected
area in Apo Island, Philippines? Mar. Pol. 34, 54–62.
Hobday, A.J., Pecl, G.T., 2013. Identification of global marine hotspots: sentinels for
change and vanguards for adaptation action. Rev. Fish Biol. Fish. 24, 415–425.
https://doi.org/10.1007/s11160-013-9326-6.
Humber, F., Harris, A., Raberinary, D., Nadon, M., 2006. Seasonal Closures of No-Take
Zones to Promote a Sustainable Fishery for octopus Cyanea (Gray) in Southwest
Madagascar. BlueVentures, London.
Hurlbert, S.H., 1984. Pseudoreplication and the design of ecological field experiments.
Ecol. Monogr. 54, 187–211.
Hurlbert, S.H., 2009. The ancient black art and transdisciplinary extent of
pseudoreplication. J. Comp. Psychol. 123, 434–443. https://doi.org/10.1037/
a0016221.
Jokiel, P.L., Rodgers, K.S., Brown, E.K., Kenyon, J.C., Aeby, G.S., Smith, W.R., et al.,
2005. Comparison of methods used to estimate coral cover in the Hawaiian islands.
Program 1–22. https://doi.org/10.7717/peerj.954.
Jones, G.P., McCormick, M.I., Srinivasan, M., Eagle, J.V., 2004. Coral decline threatens
fish biodiversity in marine reserves. Proc. Natl. Acad. Sci. Unit. States Am. 101,
8251–8253. Available: http://www.pnas.org/cgi/content/abstract/101/21/8251.
Jupiter, S.D., Egli, D.P., 2011. Ecosystem-based management in Fiji: successes and
challenges after five years of implementation. J Mar Biol 2011, 1–14. https://doi.
org/10.1155/2011/940765.
Jupiter, S.D., Cohen, P.J., Weeks, R., Tawake, A., Govan, H., 2014. Locally-managed
marine areas: multiple objectives and diverse strategies. Pac. Conserv. Biol. 20,
165–179. Surrey Beatty and Sons Pty Ltd.
Jury, M.R., Pathack, B., Parker, B., 1999. Climatic determinants and statistical prediction
of tropical cyclone days in the southwest Indian Ocean. J. Clim. 12, 1738–1746.
https://doi.org/10.1175/1520-0442(1999)012<1738:CDASPO>2.0.CO;2.
Kaunda-Arara, B., Rose, G.A., 2004. Long-distance movements of coral reef fishes. Coral
Reefs. https://doi.org/10.1007/s00338-004-0409-7.
Knutson, T.R., McBride, J.L., Chan, J., Emanuel, K., Holland, G., Landsea, C., et al., 2010.
Tropical cyclones and climate change. Nat. Geosci. 3, 157–163. https://doi.org/
10.1038/ngeo779.
Lawton, J.H., 1996. Patterns in ecology. Oikos 75, 145. https://doi.org/10.2307/
3546237.
Le Manach, F., Gough, C., Harris, A., Humber, F., Harper, S., Zeller, D., 2012. Unreported
fishing, hungry people and political turmoil: the recipe for a food security crisis in
Madagascar? Mar. Pol. 36, 218–225. https://doi.org/10.1016/j.
marpol.2011.05.007.
L�
eopold, M., Beckensteiner, J., Kaltavara, J., Raubani, J., Caillon, S., 2013. Communitybased management of near-shore fisheries in Vanuatu: what works? Mar. Pol. 42,
167–176. https://doi.org/10.1016/j.marpol.2013.02.013.
Lester, S.E., Halpern, B.S., Grorud-Colvert, K., Lubchenco, J., Ruttenberg, B.I., Gaines, S.
D., et al., 2009a. Biological effects within no-take marine reserves: a global synthesis.
Mar. Ecol. Prog. Ser. 384, 33–46.
Lester, S., Halpern, B., Grorud-Colvert, K., Lubchenco, J., Ruttenberg, B., Gaines, S.,
et al., 2009b. Biological effects within no-take marine reserves: a global synthesis.
Mar. Ecol. Prog. Ser. 384, 33–46. https://doi.org/10.3354/meps08029.
Levine, A.S., Richmond, L.S., 2014. Examining enabling conditions for community-based
fisheries comanagement: comparing efforts in Hawai’i and American Samoa. Ecol.
Soc. 19 https://doi.org/10.5751/ES-06191-190124.
Lough, J.M., 2012. Small change, big difference: sea surface temperature distributions
for tropical coral reef ecosystems, 1950-2011. J Geophys Res Ocean. https://doi.org/
10.1029/2012JC008199.
Luo, J., Serafy, J.E., Sponaugle, S., Teare, P.B., Kieckbusch, D., 2009. Movement of gray
snapper Lutjanus griseus among subtropical seagrass, mangrove, and coral reef
habitats. Mar. Ecol. Prog. Ser. https://doi.org/10.3354/meps07911.
Maire, E., Cinner, J., Velez, L., Huchery, C., Mora, C., Dagata, S., et al., 2016. How
accessible are coral reefs to people? A global assessment based on travel time. In:
Holyoak, M. (Ed.), Ecol. Lett. 19, 351–360. https://doi.org/10.1111/ele.12577.
Mallet, D., Vigliola, L., Wantiez, L., Pelletier, D., 2016. Diurnal temporal patterns of the
diversity and the abundance of reef fishes in a branching coral patch in New
Caledonia. Austral Ecol. https://doi.org/10.1111/aec.12360.
Marshell, A., Mills, J.S., Rhodes, K.L., McIlwain, J., 2011. Passive acoustic telemetry
reveals highly variable home range and movement patterns among unicornfish
within a marine reserve. Coral Reefs. https://doi.org/10.1007/s00338-011-0770-2.
Mazor, T., Possingham, H.P., Edelist, D., Brokovich, E., Kark, S., 2014. The crowded sea:
incorporating multiple marine activities in conservation plans can significantly alter
spatial priorities. PloS One. https://doi.org/10.1371/journal.pone.0104489.
McClanahan, T.R., Arthur, R., 2001. The effect of marine reserves and habitat on
populations of East African coral reef fishes. Ecol. Appl. 11, 559–569.
Mcclanahan, T.R., Graham, N.A.J., 2015. Marine reserve recovery rates towards a
baseline are slower for reef fish community life histories than biomass. Proc R Soc B
Biol Sci 282, 20151938.
McClanahan, T.R., Graham, N.A.J., Calnan, J.M., MacNeil, M.A., 2007. Toward pristine
biomass: reef fish recovery in coral reef marine protected areas in Kenya. Ecol. Appl.
17, 1055–1067.
McClanahan, T.R., Graham, N.A.J., MacNeil, M.A., Muthiga, N.A., Cinner, J.E.,
Bruggemann, J.H., et al., 2011. Critical thresholds and tangible targets for
ecosystem-based management of coral reef fisheries. Proc. Natl. Acad. Sci. Unit.
States Am. 108, 17230–17233. https://doi.org/10.1073/pnas.1106861108.
McClanahan, T.R., Donner, S.D., Maynard, J.A., MacNeil, M.A., Graham, N.A.J.,
Maina, J., et al., 2012. Prioritizing key resilience indicators to support coral reef
management in a changing climate. PloS One. https://doi.org/10.1371/journal.
pone.0042884.
McMahon, K.W., Berumen, M.L., Thorrold, S.R., 2012. Linking habitat mosaics and
connectivity in a coral reef seascape. Proc. Natl. Acad. Sci. Unit. States Am. https://
doi.org/10.1073/pnas.1206378109.
Mellin, C., Aaron Macneil, M., Cheal, A.J., Emslie, M.J., Julian Caley, M., 2016. Marine
protected areas increase resilience among coral reef communities. Ecol. Lett. 19,
629–637. https://doi.org/10.1111/ele.12598.
Micheli, F., Halpern, B.S., Botsford, L.W., Warner, R.R., 2004. Trajectories and correlates
of community change in no-take marine reserves. Ecol. Appl. 14, 1709–1723.
Micheli, F., Saenz-Arroyo, A., Greenley, A., Vazquez, L., Espinoza Montes, J.A.,
Rossetto, M., et al., 2012. Evidence that marine reserves enhance resilience to
climatic impacts. PloS One. https://doi.org/10.1371/journal.pone.0040832.
Mills, M., Jupiter, S.D., Pressey, R.L., Ban, N.C., Comley, J., 2011. Incorporating
effectiveness of community-based management in a national marine gap analysis for
Fiji. Conserv. Biol. 25, 1155–1164. https://doi.org/10.1111/j.15231739.2011.01749.x.
Molloy, P.P., McLean, I.B., C^
ot�e, I.M., 2009. Effects of marine reserve age on fish
populations: a global meta-analysis. J. Appl. Ecol. 46, 743–751. https://doi.org/
10.1111/j.1365-2664.2009.01662.x.
Mora, C., Andr�efou€et, S., Costello, M.J., Kranenburg, C., Rollo, A., Veron, J., et al., 2006.
Coral reefs and the global network of marine protected areas. Science (80) 312,
1750–1751. Available: http://www.sciencemag.org/content/312/5781/1750%
5Cnhttp://www.ncbi.nlm.nih.gov/pubmed/16794065%5Cnhttp://www.scienc
emag.org/content/312/5781/1750.summary?sid¼4a65a15c-ee54-418b-8b35-5e1
ed85f4a44.
Moreno, G., 2011. Octopus (Octopus cyanea) from Madagascar and Tanzania in
international markets. In: Survey about Market Potential for MSC-Certified octopus
in International Markets. Marine Stewardship Council.
Mosquera, I., C^
ot�e, I.M., Jennings, S., Reynolds, J.D., 2000. Conservation benefits of
marine reserves for fish populations. Anim. Conserv. 3, 321–332.
Nagelkerken, I., Grol, M.G.G., Mumby, P.J., 2012. Effects of marine reserves versus
nursery habitat availability on structure of reef fish communities. PloS One. https://
doi.org/10.1371/journal.pone.0036906.
Nash, K.L., Graham, N.A.J., Wilson, S.K., Bellwood, D.R., 2013. Cross-scale habitat
structure drives fish body size distributions on coral reefs. Ecosystems 16, 478–490.
https://doi.org/10.1007/s10021-012-9625-0.
Nelson, F., 2012. Recognition and support of ICCAs in Kenya. Recognising and
supporting territories and areas conserved by indigenous peoples and local
communities: global overview and national case studies. In: Montreal, Canada:
Secretariat of the Convention on Biological Diversity. ICCA Consortium, Kalpavriksh,
and Natural Justice.
€
Ohman,
M.C., Rajasuriya, A., 1998. Relationships between habitat structure and fish
communities on coral and sandstone reefs. Environ. Biol. Fish. 53, 19–31.
Oliver, T., Oleson, K., Ratsimbazafy, H., Raberinary, D., Benbow, S., Harris, A., 2015.
Positive catch & economic benefits of periodic octopus fishery closures: do effective,
narrowly targeted actions “catalyze” broader management? PloS One.
Osenberg, C.W., Bolker, B.M., White, J.-S.S., Mary, C.M.S., Shima, J.S., 2006. Statistical
issues and study design in ecological restorations: lessons learned from marine
reserves. Found Restor Ecol 280.
Osenberg, C.W., Shima, J.S., Miller, S.L., Stier, A.C., 2011. Assessing effects of marine
protected areas: confounding in space and possible solutions. In: Claudet, J. (Ed.),
Marine Protected Areas: A Multidisciplinary Approach. Cambridge University Press.
Pelc, R.A., Warner, R.R., Gaines, S.D., Paris, C.B., 2010. Detecting larval export from
marine reserves. Proc. Natl. Acad. Sci. U. S. A. 107, 18266–18271. https://doi.org/
10.1073/pnas.0907368107.
Pollnac, R.B., Crawford, B.R., Gorospe, M.L.G., 2001. Discovering factors that influence
the success of community-based marine protected areas in the Visayas, Philippines.
Ocean Coast Manag. 44, 683–710.
Pomeroy, R.S., Mascia, M.B., Pollnac, R.B., 2007. Marine protected areas, the social
dimension. In: Report and Documentation of the Expert Workshop on Marine
Protected Areas and Fisheries Management: Review of Issues and Considerations,
pp. 149–181.
Pratchett, M.S., Berumen, M.L., Marnane, M.J., Eagle, J.V., Pratchett, D.J., 2008. Habitat
associations of juvenile versus adult butterflyfishes. Coral Reefs 27, 541–551.
https://doi.org/10.1007/s00338-008-0357-8.
R Development Core Team, 2014. R: A Language and Environment for Statistical
Computing [Internet]. R Foundation for Statistical Computing Vienna Austria, ISBN
3-900051-07-0. https://doi.org/10.1038/sj.hdy.6800737. http://www.R-project.or
g.
Rocliffe, S., Peabody, S., Samoilys, M., Hawkins, J.P., 2014. Towards a network of locally
managed marine areas (LMMAs) in the Western Indian Ocean. PloS One 9. Public
Library of Science.
Russ, G.R., Alcala, A.C., 1996. Marine reserves: rates and patterns of recovery and
decline of large predatory fish. Ecol. Appl. 6, 947–961. https://doi.org/10.2307/
2269497.
Russ, G.R., Alcala, A.C., 1998. Natural fishing experiments in marine reserves 19831993: community and trophic responses. Coral Reefs 17, 383–397. https://doi.org/
10.1007/s003380050144.
Russ, G.R., Alcala, A.C., 2003. Marine reserves: rates and patterns of recovery and
decline of predatory fish, 1983–2000. Ecol. Appl. 13, 1553–1565. https://doi.org/
10.1890/01-5341.
13
H. Gilchrist et al.
Ocean and Coastal Management 192 (2020) 105210
Russ, G.R., Alcala, A.C., 2011. Enhanced biodiversity beyond marine reserve boundaries:
the cup spillith over. Ecol. Appl. 21, 241–250. https://doi.org/10.1890/09-1197.1.
Russ, G.R., Alcala, A.C., Maypa, A.P., Calumpong, H.P., White, A.T., 2004. Marine
reserve benefits local fisheries. Ecol. Appl. 14, 597–606.
Russ, G.R., Stockwell, B., Alcala, A.C., 2005. Inferring versus measuring rates of recovery
in no-take marine reserves. Mar. Ecol. Prog. Ser. 292, 1–12.
Russ, G., Miller, K., Rizzari, J., Alcala, A., 2015. Long-term no-take marine reserve and
benthic habitat effects on coral reef fishes. Mar. Ecol. Prog. Ser. 529, 233–248.
Available: http://www.int-res.com/abstracts/meps/v529/p233-248/.
Samoilys, M.A., Obura, D.O., 2011. Marine Conservation Successes in Eastern Africa.
Cordio East Africa.
Samoilys, M.A., Martin-Smith, K.M., Giles, B.G., Cabrera, B., Anticamara, J.A., Brunio, E.
O., et al., 2007. Effectiveness of five small Philippines’ coral reef reserves for fish
populations depends on site-specific factors, particularly enforcement history. Biol.
Conserv. 136, 584–601. https://doi.org/10.1016/j.biocon.2007.01.003.
Samoilys, M.A., Kanyange, N., Macharia, D., Maina, G.W., 2013. Dynamics of rabbitfish
(Siganus sutor) spawning aggregations in southern Kenya. In: Robinson, J.,
Samoilys, M.A. (Eds.), Reef Fish Spawning Aggregations in the Western Indian
Ocean: Research for Management, vol. 13. Western Indian Ocean Marine Science
Association (WIOMSA), Zanzibar, pp. 33–45.
Selig, E.R., Bruno, J.F., 2010. A global analysis of the effectiveness of marine protected
areas in preventing coral loss. PloS One 5.
Sen, S., Nielsen, J.R., 1996. Fisheries co-management: a comparative analysis. Mar. Pol.
20, 405–418.
Shanks, A.L., Grantham, B.A., Carr, M.H., 2003. Propagule dispersal distance and the size
and spacing of marine reserves. Ecol. Appl. 13.
Stafford, R., 2018. Lack of evidence that governance structures provide real ecological
benefits in marine protected areas. Ocean Coast Manag. 152, 57–61. https://doi.org/
10.1016/j.ocecoaman.2017.11.013.
Stewart, G.B., Kaiser, M.J., C^
ot�e, I.M., Halpern, B.S., Lester, S.E., Bayliss, H.R., et al.,
2009. Temperate marine reserves: global ecological effects and guidelines for future
networks. Conserv Lett 2, 243–253. https://doi.org/10.1111/j.1755263X.2009.00074.x.
Tetreault, I., Ambrose, R.F., 2007. Temperate marine reserves enhance targeted but not
untargeted fishes in multiple no-take MPAS. Ecol. Appl. https://doi.org/10.1890/
06-0161.1.
Thiault, L., Collin, A., Chlous, F., Gelcich, S., Claudet, J., 2017. Combining participatory
and socioeconomic approaches to map fishing effort in smallscale fisheries. PloS
One. https://doi.org/10.1371/journal.pone.0176862.
Thomas, H.L., Macsharry, B., Morgan, L., Kingston, N., Moffitt, R., Stanwell-Smith, D.,
et al., 2014. Evaluating official marine protected area coverage for Aichi Target 11:
appraising the data and methods that define our progress: evaluating official marine
protected area coverage for aichi target 11. Aquat. Conserv. Mar. Freshw. Ecosyst.
24, 8–23. https://doi.org/10.1002/aqc.2511.
Viechtbauer, W., 2010. Conducting meta-analyses in R with the metafor package. J. Stat.
Software 36, 1–48. https://doi.org/10.1103/PhysRevB.91.121108.
Walmsley, S.f., White, A.t., 2003. Influence of social, management and enforcement
factors on the long-term ecological effects of marine sanctuaries. Environ. Conserv.
30, 388–407. https://doi.org/10.1017/S0376892903000407.
Walmsley, S., Purvis, J., Ninnes, C., 2006. The role of small-scale fisheries management
in the poverty reduction strategies in the Western Indian Ocean region. Ocean Coast
Manag. 49, 812–833. https://doi.org/10.1016/j.ocecoaman.2006.08.006.
Weiant, P., Aswani, S., 2006. Early effects of a community-based marine protected area
on the food security of participating households. SPC Tradit Mar Resour Manag
Knowl Inf Bull 19, 16–31. https://doi.org/10.1016/j.ptsp.2011.03.003.
White, A., 1988. The effect of community-managed marine reserves in the Philippines on
their associated coral reef fish population. Asian Fish Sci. 2, 27–41.
Williamson, D.H., Ceccarelli, D.M., Evans, R.D., Jones, G.P., Russ, G.R., 2014. Habitat
dynamics, marine reserve status, and the decline and recovery of coral reef fish
communities. Ecol Evol 4, 337–354.
Wilson, S.K., Graham, N.A.J., Pratchett, M.S., Jones, G.P., Polunin, N.V.C., 2006.
Multiple disturbances and the global degradation of coral reefs: are reef fishes at risk
or resilient? Global Change Biol. 2220–2234. https://doi.org/10.1111/j.13652486.2006.01252.x.
Wilson, S.K., Dolman, A.M., Cheal, A.J., Emslie, M.J., Pratchett, M.S., Sweatman, H.P.A.,
2009. Maintenance of fish diversity on disturbed coral reefs. Coral Reefs. https://doi.
org/10.1007/s00338-008-0431-2.
Wilson, S.K., Babcock, R.C., Fisher, R., Holmes, T.H., Moore, J.A.Y., Thomson, D.P.,
2012. Relative and combined effects of habitat and fishing on reef fish communities
across a limited fishing gradient at Ningaloo. Mar. Environ. Res. 81, 1–11. https://
doi.org/10.1016/j.marenvres.2012.08.002.
14