Effects of a single intensive harvest event on fish populations inside a customary marine closure

Effects of a single intensive harvest event on fish populations inside a customary marine closure S. D. Jupiter, R. Weeks, A. P. Jenkins, D. P. Egli & A. Cakacaka Coral Reefs Journal of the International Society for Reef Studies ISSN 0722-4028 Volume 31 Number 2 Coral Reefs (2012) 31:321-334 DOI 10.1007/s00338-012-0888-x 1 23 Your article is protected by copyright and all rights are held exclusively by SpringerVerlag. This e-offprint is for personal use only and shall not be self-archived in electronic repositories. If you wish to self-archive your work, please use the accepted author’s version for posting to your own website or your institution’s repository. You may further deposit the accepted author’s version on a funder’s repository at a funder’s request, provided it is not made publicly available until 12 months after publication. 1 23 Author's personal copy Coral Reefs (2012) 31:321–334 DOI 10.1007/s00338-012-0888-x REPORT Effects of a single intensive harvest event on fish populations inside a customary marine closure S. D. Jupiter • R. Weeks • A. P. Jenkins D. P. Egli • A. Cakacaka • Received: 26 October 2011 / Accepted: 1 February 2012 / Published online: 16 February 2012 Ó Springer-Verlag 2012 Abstract In September 2008, the villagers of Kia Island, Fiji, opened their customary managed closure (Cakaulevu tabu) to fishing for a fundraiser that lasted for 5 weeks. We report on opportunistic before-after-control-impact surveys describing changes to coral reef communities both 4 weeks into the harvest and 1 year later compared with pre-harvest conditions. Prior to the harvest, there was a gradient in mean fish abundance and biomass per transect, with highest levels in the north of the closure (250 fish transect-1, 8,145.8 kg ha-1), intermediate levels in the south of the closure (159 fish transect-1, 4,672.1 kg ha-1) and lowest levels in the control area open to fishing (109 fish transect-1, 594.0 kg ha-1). During the harvest, there were extensive depletions in large-bodied, primary targeted fish species, with significant loss in biomass of Acanthuridae and Carangidae in the north and Lutjanidae and Serranidae in the south. We also observed significant increases in Acanthuridae, Lethrinidae and Scaridae in the control, suggesting a ‘‘bail-out’’ effect whereby fish left the closure in response to a rapid increase in fishing pressure. These Communicated by Biology Editor Prof. Philip Munday Electronic supplementary material The online version of this article (doi:10.1007/s00338-012-0888-x) contains supplementary material, which is available to authorized users. S. D. Jupiter (&)
R. Weeks
D. P. Egli
A. Cakacaka Wildlife Conservation Society, Fiji Country Program, 11 Ma’afu Street, Suva, Fiji e-mail: sjupiter@wcs.org R. Weeks ARC Centre of Excellence for Coral Reef Studies, James Cook University, Townsville, Australia A. P. Jenkins Wetlands International-Oceania, Suva, Fiji changes were coupled with a large increase in turf algal cover at all survey areas, despite a large numerical increase in small, roving acanthurids (e.g., Ctenochaetus striatus) and scarids (e.g., Chlorurus sordidus). By 1 year later, fish biomass was significantly lower within the closure than before the harvest, while values in the control returned to pre-harvest levels, suggesting non-compliance with the reinstated fishing ban. We use the lessons learned from this event to suggest recommendations for promoting effective management of periodically harvested customary closures that are a common feature across much of Oceania. Keywords Marine protected areas
Customary management
Tabu areas
Coral reef fisheries
Periodic harvest
Fiji Introduction Attempts to reduce threats to the marine environment in Oceania, particularly from overfishing, have increasingly focused on ecosystem-based approaches to conservation, including the designation of marine protected areas (MPAs) (Jupiter and Egli 2011; Aswani et al. 2012). There is strong evidence that permanent, no-take areas provide fisheries benefits in terms of increased numerical abundance and biomass (Mumby and Steneck 2008; Lester et al. 2009) and growing evidence to support density-dependent adult spillover and larval exchange supported by increased egg production from larger fish (Abesamis and Russ 2005; McCook et al. 2010). However, in many Pacific islands, there are strong cultural and legislative barriers to placing permanent restrictions on access to traditional fishing grounds (Foale and Manele 2004; Ruddle and Hickey 2008; Clarke and Jupiter 2010). Conservation organizations and practitioners 123 Author's personal copy 322 working in this region, such as partners within the Locally Managed Marine Area (LMMA) network, typically advocate an integration of traditional management practices and scientific knowledge. The LMMA network is composed of government and non-government organizations working together with communities to share knowledge and best practice for coastal and marine resource management to achieve local objectives (Govan et al. 2009). Such approaches have resulted in management systems that comprise a multitude of different strategies including no-take closures, temporary closures, size limits, seasonal or species bans and gear restrictions (Ban et al. 2011). Of these, periodically harvested marine closures are the most common form of spatial management (Govan et al. 2009). The relative effectiveness of permanent no-take MPAs versus other management strategies has been the subject of much debate in the literature. There is some evidence that periodically harvested closures can promote short-term recovery of fish abundance and biomass (Cinner et al. 2005; Bartlett et al. 2009a), and compliance may be stronger under these customary management systems compared with western-style permanent no-take areas (Cinner et al. 2006). Yet, positive perceptions by locals of their effectiveness are not always validated by ecological surveys, particularly for fish biomass (Bartlett et al. 2009b). Furthermore, long-term ecological studies and meta-analyses suggest that MPAs must be no-take and permanently closed to achieve sustained fisheries benefits (Russ and Alcala 2004; Di Franco et al. 2009; Lester et al. 2009; Maliao et al. 2009). This is often due to focused fishing efforts within MPA boundaries when they are opened. Permitted periodic harvests within traditional MPAs have been observed to remove marine resources with ‘‘alarming efficiency’’ (Foale and Manele 2004), and fisheries benefits from protection (e.g., increased fish size, abundance and spillover potential) have been rapidly removed where notake status has been revoked (Alcala et al. 2005; Williams et al. 2006). It is clear that while MPAs other than permanent no-take closures can contribute towards achieving objectives for sustainable resource use and conservation, different management strategies are not equally effective (Mills et al. 2011). The effectiveness of a management strategy will vary across species and habitats and in response to the objective being measured. Given that periodically harvested closures are the only feasible management option throughout much of the Pacific, science-based guidelines for the frequency and intensity with which such MPAs can be sustainably harvested are urgently required. Attempts to develop such guidelines have been constrained by a lack of empirical evidence for the impact of periodic harvests on fish populations. Here, we report on an opportunistic pre-, during and post-harvest (1 year later) survey of a 123 Coral Reefs (2012) 31:321–334 customary closure (tabu area) in Fiji. We investigate the direct and indirect effects of the harvest event on coral reef fish biomass and abundance, potential fish behavioural responses and reef benthos. Based on these analyses, we provide recommendations to improve the sustainability of local coastal resource management strategies. Methods Study site The Cakaulevu tabu area (15.5 km2) is located on the exposed outer barrier reef west of Kia Island, Macuata Province, on the island of Vanua Levu in Fiji (Fig. 1). In 2005, the high chief of Macuata, together with the Cokovata Qolioli Management Committee, endorsed the establishment of a network of nine community-managed tabu areas, including the Cakaulevu tabu. The MPA network was established with the assistance of conservation partners, led by the WWF South Pacific Programme, with the objectives to protect reef fish stocks for the future and to conserve marine biodiversity. Prior to 2005, the Cakaulevu tabu was informally protected and may have been opened periodically for subsistence use, but to our knowledge was never previously harvested for commercial extraction. The residents of the three villages on Kia Island have traditional fishing rights within their customary fishing ground (qoliqoli) and are heavily dependent upon marine resources for their livelihoods, as the small, steep-sloped island contains little arable land. In September 2008, the communities collectively decided to harvest the Cakaulevu tabu to raise funds to support school, church and provincial fees for island residents. Initially, the goal was to raise FJD$12,000 (USD$7,492 based on exchange rates from 22 September 2008, the first day of the harvest) from the sale of invertebrates (predominantly bêche-de-mer) and fish. This goal was exceeded on the first day, and upon realizing the potential for profit, community members extended the harvest period for 5 weeks. Residents and traditional fishing rights owners living outside of Kia Island who returned for the event fished in shifts over 24 h periods from Mondays through Saturdays. Based on daily amounts paid out by middlemen from three seafood export companies (reported to A. Cakacaka), we estimated a total revenue to community members of over FJD$200,000 (USD$124,871). The majority of the catch was caught with spearguns, Hawaiian slings, and hook and line (S. Jupiter and A. Cakacaka, pers. obs.). Researchers from the Wildlife Conservation Society (WCS) surveyed reefs inside and outside the Cakaulevu tabu in September 2008 as part of a larger project to assess Author's personal copy Coral Reefs (2012) 31:321–334 Tabu Areas Coral Reefs Qoliqoli boundary Survey ey Sites NorthTabu North 1 Control South T Tabu Fig. 1 Map of Kia Island and the Cakaulevu tabu area showing the locations of survey sites inside and outside the MPA. The inset shows the location of the study region in Macuata Province on the island of Vanua Levu, Fiji 323 Cakaulevu Reefs Tabu North 2 South 1 South 2 Kia Island Control 1 Control 2 0 the effectiveness of networks of community MPAs to increase food fish biomass (Jupiter and Egli 2011). Three days after the start of the survey, the communities informed WCS of their decision to open the tabu area. The WCS team then returned to survey the same sites 4 weeks into the harvest (October 2008) to determine the effects of the harvest event on fish populations, and 1 year later (October 2009) to look for evidence of recovery. Fish and benthic surveys Underwater visual census surveys of fish and benthic substrate were conducted on forereef sites inside (n = 4) and outside (n = 2) of the Cakaulevu tabu area (Fig. 1) using the methods described in Jupiter and Egli (2011). Due to notable differences in reef geomorphology (S. Jupiter, pers. obs.) that initially supported distinct fish communities, the survey sites within the tabu area were divided into two regions for analysis. The northern tabu sites are characterized by high current and steep reef slopes/vertical walls, and the southern tabu sites are characterized by less current and a gradual slope with variable topography and channels to 20 m. The control sites, located on a reef to the south of the Cakaulevu tabu, share similar reef geomorphology to the southern tabu sites. The location and replication of survey sites were constrained by the opportunistic nature of the study and are admittedly not ideal for this purpose. Nevertheless, these data permit a before-after-control-impact (BACI) analysis of the effects of opening a customary marine closure. 1 2 3 4 5 Kilometers At each site, trained fish observers from WCS estimated fish size (2–5 cm, then 5 cm classes to 40 cm, and size to the nearest cm above 40 cm) and abundance at ten replicate 5 m 9 50 m belt transects at shallow (5–8 m; n = 5) and deep (12–15 m; n = 5) depths. They recorded fish from the following families: Acanthuridae, Balistidae, Carangidae, Carcharhinidae, Chaetodontidae, Chanidae, Ephippidae, Haemulidae, Kyphosidae, Labridae, Lethrinidae, Lutjanidae, Mullidae, Pomacanthidae, Scaridae, Scombridae, Serranidae (subfamily Epinephelinae only), Siganidae, Sphyraenidae and Zanclidae. We calculated biomass from size class estimates of length (L) and existing published values from FishBase (Froese and Pauly 2009) used in the standard length–weight (L–W) expression W = aLb, where a and b parameter values were preferentially selected from sites closest to Fiji (Jupiter and Egli 2011). As many of the L–W conversions required fork length (FL), a length–length (L–L) conversion factor was obtained from FishBase where necessary to convert from total length (TL) recorded during the surveys to FL before biomass estimation. Because the biomass conversion formula resulted in some grossly overestimated weights for fishes that substantially change morphology as they age, maximum published weights were used for certain species when these fish were sighted above threshold sizes, as per Jupiter and Egli (2011). We additionally assessed which species were most likely to be primary targets based on: market value information from the Fiji Department of Fisheries divisional office in Labasa; maximum length from FishBase; and position typically found in the water column (fish high in the water column are more easily targeted by spearfishers). 123 Author's personal copy 324 To determine associations between fish communities and benthos both before and following the harvest, separate observers recorded benthic life-form categories along the same 50 m transects surveyed for fish assemblages at 0.5 m intervals using the point-intercept method. We combined life-form categories into 6 functional strata: fleshy algae/cyanobacteria, including cyanobacteria, algal assemblages and fleshy macroalgae [2 cm height; live hard coral, including Millepora and Tubipora; other soft substrate, including soft corals, sponges, ascidians, anemones and zooanthids; reef matrix, including dead coral, reef pavement and coralline algae; turf algae, defined as algae B2 cm height on reef matrix or dead coral; and unconsolidated substrate, comprised of rubble, sand or silt. Coral Reefs (2012) 31:321–334 between groups (Anderson and Willis 2003). We selected the number of principal coordinate axes (m) that provided the best distinction between groups, following the method described in Anderson and Willis (2003). The CAP procedure provides misclassification errors using a ‘‘leaveone-out’’ method, whereby each observation is removed from the analysis and then placed in the canonical space determined by the rest of the observations (Anderson and Willis 2003). The percentage of correct classifications provides a measure of the goodness of fit. Species’ correlations with the first two CAP axes were plotted to identify those that had greatest influence on site groupings. Results Statistical analyses Total fish abundance and biomass To assess how the harvest affected reef fish populations within and outside of the Cakaulevu tabu, we conducted permutational multivariate analysis of variance (PERMANOVA) analyses with 4,999 permutations using log10 Modified Gower resemblance matrices (Anderson 2001; McArdle and Anderson 2001) of fish abundance and biomass summed at the transect level for total fish and for the following major food fish families: Acanthuridae; Carangidae; Lethrinidae; Lutjanidae; Scaridae; and Serranidae. We first tested the main effects models with three levels of fixed factors: sampling period, area and site nested within area. Samples were pooled across depths for each site. Where a significant interaction term was found between sampling period and area, we conducted separate post hoc pair-wise comparisons within sampling period between each area and within area between each sampling period (Anderson et al. 2008). Total and scarid biomass PERMANOVA analyses were conducted with and without Bolbometopon muricatum records, as the presence of these large fish that tend to occur in groups may mask other impacts from fishing. To assess potential differences in benthic structure that may influence fish community assemblages, we performed the same PERMANOVA model design on a Euclidean distance similarity matrix of percent benthic strata cover for each transect and then conducted post hoc pair-wise comparisons within sampling period between each area and within area between each sampling period. We performed these analyses using PRIMER version 6 software with PERMANOVA extension. We assessed differences in reef fish community composition through time using canonical analysis of principal coordinates (CAP) performed on a Bray–Curtis dissimilarity matrix of species presence–absence data using R software. CAP is a constrained ordination procedure that initially calculates unconstrained principal coordinate (PCO) axes, followed by canonical discriminant analysis on the principal coordinates to maximize separation 123 Under pre-harvest conditions, we observed a strong gradient of mean fish abundance with the highest values in the north (250 fish transect-1), followed by the south (159 fish transect-1) and control (109 fish transect-1) (Fig. 2). There was significant variability in abundance by sampling period, area, and the interactions between sampling period and area, as well as sampling period and site (Table 1a). Each area was significantly different from the others (Table 2a). Mean fish biomass followed the same general pattern (north: 8,145.8 kg ha-1; south: 4,672.1 kg ha-1; control: 594.0 kg ha-), with significant variation across all factors (Table 1b). When Bolbometopon muricatum biomass was excluded, each area was significantly different from the others, but only the north and south were significantly different from the control with B. muricatum (Table 2b, c). Four weeks into the harvest, mean fish abundance significantly increased in the north and control (Fig. 2a). In contrast, mean fish biomass declined in the north and south but significantly increased in the control area (Fig. 2b). Fish abundance was significantly higher in the north (354 fish transect-1) than south (188 fish transect-1), but not higher than the control (256 fish transect-1) (Table 2a). Fish biomass was significantly greater in the north (4,763.6 kg ha-1) than the south (2,246.2 kg ha-1) and control (4,582.6 kg ha-1), which were not significantly different from each other (Table 2b). By 1 year later, fish biomass was significantly lower than pre-harvest levels in the north and south of the tabu, while the control was not significantly different from pre-harvest conditions (Fig. 2b). Although fish abundance was significantly higher in the north (211 fish transect-1) than the control (153 fish transect-1), fish biomass was significantly lower in the north (987.4 kg ha-1) than in the control (1,544.8 kg ha-1). Exclusion of B. muricatum biomass records from the pair-wise comparisons of within-sampling period analyses only changed the outcome of the pre-harvest Author's personal copy 325 200 a a a c 100 a 8000 6000 a Pre−harvest During +1 year b a a 4000 b a ab 2000 a Mean fish biomass (kg ha−1) 300 400 (b) b a b b a 0 0 Mean fish abundance (# per 250 m2) (a) 10000 Coral Reefs (2012) 31:321–334 North South Control North 2 Fig. 2 Total mean abundance (# per 250 m ) (a) and biomass (kg ha-1) (b) per transect of reef fish at sites inside (north and south) and outside (control) the Cakaulevu tabu area before (black bars), during (white bars) and 1 year after (grey bars) a 5-week intensive harvest event. Adjacent different letters indicate that a significant South Control difference exists between sampling periods (e.g., ‘‘a b a’’ indicates that pre-harvest is significantly different from during harvest but not from 1 year later, and ‘‘a ab b’’ indicates that pre-harvest is not significantly different from during harvest but is significantly different from 1 year later). Error bars indicate one standard error Table 1 PERMANOVA main effects results of differences in mean fish (a) abundance per transect (# per 250 m2), (b) biomass (kg ha-1) and (c) biomass (kg ha-1) with Bolbometopon muricatum records excluded Sources SS df MS Pseudo-F P(perm) Permutations (a) Total fish abundance Sampling 2 1.176 0.588 16.359 0.0002 4,988 Area 2 1.344 0.672 18.694 0.0002 4,988 Site (area) 3 0.173 0.058 1.601 0.1910 4,989 Sampling 9 area 4 0.506 0.126 3.520 0.0074 4,987 Sampling 9 site (area) 6 0.783 0.130 3.631 0.0022 4,986 0.036 Residual 162 5.821 Total 179 9.802 (b) Total fish biomass Sampling 2 6.820 3.410 25.290 0.0002 4,985 Area 2 7.107 3.554 26.354 0.0002 4,986 Site (area) 3 7.509 2.503 18.562 0.0002 4,982 Sampling 9 area 4 4.580 1.145 8.492 0.0002 4,987 Sampling 9 site (area) 6 5.765 0.961 7.126 0.0002 4,986 162 179 21.844 53.625 0.135 Residual Total (c) Total fish biomass (no B. muricatum) Sampling 2 5.719 2.859 28.373 0.0002 4,988 Area 2 9.851 4.926 48.876 0.0002 4,989 Site (area) 3 7.626 2.542 25.225 0.0002 4,988 Sampling 9 area 4 2.805 0.701 6.957 0.0002 4,988 Sampling 9 site (area) 6 4.699 0.783 7.771 0.0002 4,984 0.101 Residual 162 16.326 Total 179 47.025 Significant P(perm) values are indicated in bold north–south comparison described above (Table 2c). Exclusion of B. muricatum biomass records from betweensampling period analyses only changed the outcome of the comparison between pre-harvest and during harvest conditions in the south, which became significantly different. 123 Author's personal copy 326 Coral Reefs (2012) 31:321–334 Table 2 PERMANOVA pair-wise comparisons within sampling period between each area for mean fish (a) abundance per transect (# per 250 m2), (b) biomass (kg ha-1) and (c) biomass (kg ha-1) with Bolbometopon muricatum records excluded (a) Abundance (b) Biomass (c) Biomass (no B. muricatum) Level t P(perm) Level t P(perm) Level t P(perm) North–south * 2.3099 0.0262 NS 1.7506 0.0912 * 2.3365 0.0296 North–control South–control *** * 4.8354 2.3664 0.0002 0.0234 *** *** 8.4888 5.2251 0.0002 0.0002 *** *** 8.4319 4.5654 0.0002 0.0002 North–south ** 4.1252 0.0010 *** 4.5592 0.0006 *** 4.5592 0.0004 North–control NS 1.9563 0.0566 * 2.2941 0.0252 *** 5.2354 0.0002 South–control NS 1.8283 0.0734 NS 1.0254 0.3118 NS 0.4823 0.6374 North–south NS 1.5589 0.1348 NS 0.0731 0.9406 NS 0.0731 0.9382 North–control *** 4.1997 0.0002 * 2.4548 0.0126 *** 5.3798 0.0002 South–control NS 1.2805 0.2046 * 2.2189 0.0254 *** 4.2777 0.0002 Pre-harvest Post-harvest 1-year later Significance levels denoted as: NS not significant, * P(perm) \ 0.05, ** P(perm) \ 0.01, *** P(perm) \ 0.0010 Fish community composition The CAP analysis showed significant shifts in species composition between sampling periods as primary target fish disappeared from the assemblages during the harvest and further by 1 year later (Fig. 3). The first eight PCO axes explained 83.1% of the variability in the dissimilarity matrix, and 94.4% of observations were classified correctly using the leave-one-out procedure (p = 0.005). Species’ correlations with the first two CAP axes are shown in Fig. 3. Species negatively correlated with axis 1 are associated with sites pre-harvest and during the harvest. There was significant variability in both fish abundance and biomass at the family level that varied slightly by taxonomic group (Electronic Supplemental Material, ESM Table S1). While the interaction between sampling period and area was always significant for fish biomass across all fish families analysed, it was not significant for fish abundance of acanthurids, serranids or scarids when B. muricatum numbers were excluded (ESM Table S1a, k, m). Pre-harvest biomass in the north was dominated by acanthurids and carangids, which was significantly higher than in the control (ESM Table S2a, b; Fig. 4). Pre-harvest biomass in the south was dominated by lutjanids and scarids, which was significantly higher than in the control (ESM Table S2d, e; Fig. 4). With respect to changes in fish biomass at the family level, by 4 weeks into the harvest, there were significant decreases inside the Cakaulevu tabu of acanthurids (north), carangids (north), lutjanids (south) and serranids (south), coupled with significant increases in acanthurids, lethrinids and scarids in the control (Fig. 4). These decreases in biomass inside the tabu occurred despite numerical increase in 123 acanthurids and scarids that increased in abundance everywhere (though only significantly for scarids in the south and control; Fig. 5). Excluding B. muricatum biomass from the scarid within-sampling and between-sampling analyses did not substantially affect the outcomes (ESM Table S2f). Within acanthurids and carangids, the largebodied species of Naso unicornis, N. tonganus, N. caesius, Acanthurus fowleri, A. auranticavus, Caranx melampygus and C. papuensis were severely depleted during the harvest, with large increases in the highly mobile N. caesius and A. auranticavus in the control (Table 3). Within lutjanids and scarids, there was a significant decrease in biomass during the harvest due to losses of Macolor macularis, Lutjanus bohar, L. gibbus, L. fulviflamma, B. muricatum and Cetoscarus bicolor, of which L. gibbus, B. muricatum and C. bicolor increased in the control (Table 3). Biomass of Chlorurus bleekeri, C. sordidus, Ctenochaetus striatus, Melichthys vidua and Scarus schlegeli increased across all survey areas during the harvest (Table 3). By 1 year later, the biomass of carangids and serranids was nearly negligible everywhere, and biomass of acanthurids (north only), carangids (north and south), lutjanids (south only) and serranids (south only) was significantly lower than preharvest values, even though numbers of acanthurids (north and south), lutjanids (south) and serranids (south) remained nearly equivalent to pre-harvest levels (Figs. 4, 5). Benthic composition There were significant differences in benthic composition, and these differences varied by sampling period, area, site and the interactions between sampling period and area and Author's personal copy Coral Reefs (2012) 31:321–334 327 Fig. 3 CAP ordination plot of reef fish presence–absence data from survey sites inside and outside the Cakaulevu tabu area at three different times of sampling. Arrows indicate species’ correlations with the first two CAP axes, with primary target species shown in black 0 −5 CAP axis 2 5 Pre−harvest During +1 year −5 0 5 CAP axis 1 sampling periods and site (Table 4). Prior to the harvest, the cover of benthic strata in the north was not significantly different from the south or the control, though the south was significantly different from the control which had less live coral and higher cover of reef matrix, turf algae and other soft substrate (Table 4). During the harvest, benthic cover in the north changed significantly, with mean cover of turf algae increasing from 13.2 to 24.2% and reef matrix declining from 21.9 to 11.8%. Benthic strata cover changed significantly in the south during the harvest and was significantly different from the north and control due to relatively larger increases in turf algae (14.7% to 44.0%) and reductions in live coral (49.7% to 33.3%). Benthic cover also changed significantly in the control, with substantial increase in turf algae (17.6% to 37.4%) and consequent reduction in reef matrix (26.6 to 11.5%). By 1 year later, benthic strata cover in the north, south and control was not significantly different from preharvest conditions as turf algal cover returned to preharvest levels everywhere. Overall benthic community composition was not significantly different between the north and the south or between the south and the control, although the north was again significantly different from the control (Table 4). Discussion Our study demonstrates that a single intensive harvest event can quickly remove almost all positive effects of protection on fish biomass and subsequent reproductive output in a marine protected area. Our results echo other findings, suggesting that substantial benefits to fisheries from closures can be removed in a very short time period through focused fishing efforts (Russ and Alcala 2003; Williams et al. 2006). The main impact of the Cakaulevu tabu harvest was seen in loss of large-bodied fish, whose depletion differed between the north and south and reflected initial differences in community composition and catchability. The high currents and steep walls of the north part of the tabu supported initial high densities of planktivorous and predatory fish, which would have built up during the preceding years of informal and formal protection. The largest losses from the north unsurprisingly were from acanthurids, especially Naso unicornis, N. tonganus, N. caesius, Acanthurus fowleri and A. auranticavus, and from carangids, particularly Caranx melampygus, C. papuensis and Carangoides oblongus (Table 3). Of these, all except A. auranticavus inhabit midto high position in the water column near the reef, making 123 Author's personal copy Coral Reefs (2012) 31:321–334 1000 2000 3000 4000 328 (a) North Tabu a Pre−harvest During +1 year a a b b c a ab ab a c a a a a a a 1000 2000 3000 4000 (b) South Tabu a a a a a a a ab 0 Mean fish biomass (kg ha−1) 0 b a b a ab b b a a b b a a a 1000 2000 3000 4000 b (c) Control c b a 0 a Acanthuridae ab b a a a Carangidae a a Lethrinidae -1 a a Lutjanidae a Scaridae Serranidae Fig. 4 Mean biomass (kg ha ) of targeted food fish families at sites inside (a and b) and outside (c) the Cakaulevu tabu area, before (black bars), during (white bars) and 1 year after (grey bars) a 5-week intensive harvest event. Adjacent different letters indicate that a significant difference exists between sampling periods within each family group (e.g., ‘‘a b a’’ indicates that pre-harvest is significantly different from during harvest but not from 1 year later, and ‘‘a ab b’’ indicates that pre-harvest is not significantly different from during harvest but is significantly different from 1 year later). Error bars indicate one standard error them easy targets for spearfishers. By contrast, the largest loss of biomass in the south was principally from large lutjanids, such as Lutjanus bohar and L. gibbus, plus Macolor macularis and M. niger, which are also easily caught due to their high position in the water column. Large scarids, such as Chlorurus microrhinos, C. bleekeri and Cetoscarus bicolor, in addition to Bolbometopon muricatum, were also reduced, likely because they are easy targets for spearfishers at night when they rest on the reef (e.g., Aswani and Hamilton 2004). During the harvest, fish communities in the Cakaulevu tabu were dominated by acanthurids and scarids, which became more numerically abundant. Similar effects have been seen in previous studies, where scarids increased in abundance following fishing (Russ and Alcala 1998) or where acanthurid and siganid numbers increased with linear distance away from a no-take zone (Ashworth and Ormond 2005). We have also observed this pattern elsewhere in Fiji, where fish communities on reefs under heavy fishing pressure are dominated by Ctenochaetus striatus, Chlorurus sordidus, Scarus schlegeli and other non-targeted species (Jupiter and Egli 2011). The prevalence of high numbers of small herbivores and detritivores could possibly be a consequence of a decline in territorial aggression from the removal of large species (Vine 1974; Robertson et al. 1979; Choat and Bellwood 1985) or a combination of various ecological processes. Regardless of the mechanism, although absolute numbers increased inside the tabu during the harvest, mean biomass of these predominantly herbivorous acanthurids and scarids declined. These large declines, coupled with large increases in turf algae, suggest that by 4 weeks into the harvest grazing capacity may have been reduced. The interpretation is supported by herbivore exclusion manipulations on the Great Barrier Reef that demonstrate the strong role herbivores play in shaping spatial variation in algal turfs (Bonaldo and Bellwood 2011). We are not aware of any other study that documents a turf algal bloom 123 Author's personal copy (a) 329 North Tabu Pre−harvest During +1 year 100 150 Coral Reefs (2012) 31:321–334 a 50 a a b a ab a a a b 150 South Tabu 50 100 (b) a ab b a ba b a a a 150 0 a a a (c) Control 100 Mean fish abundance (# per 250 m2) 0 a a 50 b b b ab a a a a a a c 0 a Acanthuridae Carangidae Lethrinidae Lutjanidae Scaridae Serranidae Fig. 5 Mean abundance (# per 250 m2) of targeted food fish families at sites inside (a and b) and outside (c) the Cakaulevu tabu area, before (black bars), during (white bars) and 1 year after (grey bars) a 5-week intensive harvest event. Adjacent different letters indicate that a significant difference exists between sampling periods within each family group (e.g., ‘‘a b a’’ indicates that pre-harvest is significantly different from during harvest but not from 1 year later, and ‘‘a ab b’’ indicates that pre-harvest is not significantly different from during harvest but is significantly different from 1 year later). Error bars indicate one standard error associated with a massive fish harvest, likely because there are few opportunities to conduct BACI surveys so soon following the start of an intensive fishing event. However, because there were also equivalent increases in turf algae in the control area and considerable variability in benthic cover at the site level, we cannot unequivocally attribute the change in benthic cover to fishing alone and we cannot discount that the results may partially be attributed to the placement of transects in different locations when we conducted surveys during the harvest and 1 year later. Our results further suggest that surveys using fish abundance data alone, without size estimates or complementary benthic composition data, need to be interpreted with caution. In our study, reliance solely on abundance data for interpretation could have led to the quite different conclusion that harvesting a protected area can be beneficial, as mean fish abundance in the north during the harvest was significantly greater than pre-harvest levels. This observation is important as partner organizations within the Fiji LMMA network generally only train community monitors to record fish abundance (e.g., Tawake et al. 2004). Significant increases in surgeonfish abundances recorded over time by Fijian communities (Seidel 2009) may thus reflect increased abundance of small opportunistic species (e.g., C. striatus) responding to indirect effects of fishing pressure rather than positive fishery gains from management. During the harvest, we observed significant increases in both fish abundance and biomass (with and without B. muricatum) in the control area, suggesting a possible behaviourally driven ‘‘bail-out’’ effect of fish moving from the tabu area to the adjacent control in response to the reversal in patterns of fishing effort. This is also supported by the breakdown of the north–south–control biomass gradient during the fishing event and the levelling off in the following year. Short-term behavioural responses of exploited species to fishing (e.g., Gotanda et al. 2009) can potentially magnify the perceived effectiveness of 123 Species North South Control Feeding Water column Group size 330 123 Table 3 Fish species that exhibited large changes in biomass following 4 weeks of intensive harvest listed with feeding guild, position most often found in the water column, group size and likelihood to move in response to fishing pressure Mobility (a) Large depletion in tabu, no increase in control Naso unicornis -1,303.3 (-15.7) -2.4 (-1.5) 0.0 (0.0) Planktivore High Small groups Low Caranx melampygus -1,014.4 (-14.3) -36.7 (-1.5) 0.0 (0.0) Predator Mid Solitary/small groups High -425.9 (-17.8) -36.2 (-6.8) 0.0 (0.0) Grazer Mid Solitary Low Acanthurus fowleri Naso tonganus -315.5 (-8.5) 0.0 (0.0) 0.0 (0.0) Planktivore High Small groups Low Lethrinus olivaceus -209.0 (-8.8) -63.8 (-12.0) 0.0 (0.0) Predator Mid Solitary Moderate Caranx papuensis -187.7 (7.0) Macolor macularis -105.1 (-1.7) 0.0 (0.0) -681.9 (-8.1) 0.0 (0.0) Predator Mid Solitary/small groups High -23.3 (0.2) Predator Mid Small groups Moderate -62.5 (-38.0) 0.0 (0.0) 0.0 (0.0) Invertivore Benthic-assoc. Solitary Low -45.1 (-0.4) -27.6 (0.0) 0.0 (0.0) Invertivore Mid Solitary Moderate Lutjanus fulviflamma -44.8 (-31.5) -19.4 (-3.3) 0.0 (0.0) Predator Mid Small groups Moderate Carangoides oblongus -44.7 (-8.0) 0.0 (0.0) 0.0 (0.0) Predator High Solitary High Chanos chanos -42.9 (-2.0) 0.0 (0.0) 0.0 (0.0) Predator Mid–High Schooling High Caesio caerulaurea -42.4 (-35.6) -11.5 (-25.7) 0.0 (0.0) Planktivore High Schooling High Platax teira -19.4 (-1.5) 0.0 (0.0) Generalist High Schooling Low High -106.5 (-0.5) Author's personal copy Scolopsis trilineatus Cheilinus undulatus (b) Large depletion in tabu, slight increase in control -323.1 (-4.1) -458.1 (-6.7) 9.1 (1.0) Predator Mid Small groups Epinephelus malabaricus -26.7 (0.0) -45.2 (-1.0) 1.3 (1.0) Predator Benthic-assoc. Solitary Low Epinephelus polyphekadion -11.6 (-3.0) -28.3 (-0.8) 0.7 (1.0) Predator Benthic-assoc. Solitary Low -102.4 (-4.4) 0.0 (0.0) 2.2 (3.5) Generalist Benthic-assoc. Small groups Moderate 64.4 (-4.3) Planktivore High Small groups High Excavator Mid Schooling High Lutjanus bohar Parupeneus crassilabris (c) Large depletion in tabu, large increase in control Naso caesius -399.6 (-8.8) Bolbometopon muricatum -63.9 (-2.0) Acanthurus auranticavus Balistoides viridescens 0.7 (4.0) -868.6 (-5.3) 2,801.3 (21.7) -55.1 (-13.5) -41.4 (-4.1) 104.0 (38.5) -8.1 (-0.3) -110.5 (-0.8) 42.6 (0.2) Grazer/Detritivore Mid Schooling High Invertivore Mid Solitary Low 178.3 (15.4) -99.6 (-6.9) 239.6 (20.3) Predator Mid Schooling High 234.8 (3.6) -221.4 (-3.5) 259.1 (8.3) Invertivore Mid Schooling Low 34.1 (1.4) -44.5 (-3.7) 26.0 (0.6) Excavator Benthic-assoc. Solitary Moderate 60.6 (6.5) 47.4 (0.8) 59.6 (3.6) Scraper/small excavator Benthic-assoc. Solitary High High Cetoscarus bicolor (d) Large increase everywhere Chlorurus bleekeri 89.2 (5.6) 57.1 (2.0) 58.4 (2.2) Scraper/small excavator Benthic-assoc. Solitary 126.6 (11.1) 64.1 (7.0) 78.8 (8.3) Detritivore Benthic-assoc. Small groups High Melichthys vidua 88.4 (1.9) 18.7 (0.4) 26.4 (0.3) Omnivore Mid Small groups Low Scarus schlegeli 49.5 (6.8) 26.1 (1.9) 22.5 (2.6) Scraper/small excavator Benthic-assoc. Small groups High Chlorurus sordidus Ctenochaetus striatus Values for north, south and control indicate change in mean kg ha abundance (# per 250 m2) -1 biomass from 4 weeks into the harvest compared with pre-harvest surveys. Values in parenthesis indicate mean change in Coral Reefs (2012) 31:321–334 Lutjanus gibbus Monotaxis grandoculis Author's personal copy Coral Reefs (2012) 31:321–334 protected areas. For example, open areas with high intensity of fishing pressure can cause ‘‘spill-in’’ of exploited species into reserves (Eggleston and Parsons 2008). Consequently, the reverse pattern is likely to occur when periodic closures are opened to fishing, resulting in a ‘‘bailout’’ of targeted species. A small portion of this trend observed from the Kia Island data may be simply due to instantaneous variation in fish community assemblages (McClanahan et al. 2007a), especially from schooling species. This may explain, for example, why the sum of the mean increase in lethrinid biomass during the harvest in the control (?281.4 kg ha-1) and the north (?85.5 kg ha-1) was in fact slightly more than the mean decrease from the south (-234.4 kg ha-1). However, the consistent pattern of loss of large-bodied acanthurids, carangids, lethrinids, lutjanids and scarids in the tabu during the harvest, coupled with a corresponding increase in numbers and biomass of most of these taxa in the control, suggests that many of the targeted fish that did not get caught moved. Some of the larger individuals of vulnerable species in the Cakaulevu tabu may have escaped predation by swimming deeper (S. Jupiter and A. Cakacaka, pers. obs.), as has been observed in other studies of Fijian reef fish communities from fished locations (Goetze et al. 2011). Others that are capable of moving large distances may have fled away from the focus of predation (Table 3), which is supported by the work of Januchowski-Hartley et al. (2011) who demonstrate that body size is a strong factor associated with flight initiation distance at higher levels of fishing pressure. One year after the harvest event, total fish biomass had not recovered inside the Cakaulevu tabu, and the biomass of piscivores such as carangids and serranids was nearly exhausted everywhere. As there were no significant differences in benthic composition inside the tabu compared with pre-harvest conditions, it is likely that the fish communities were still impacted by ongoing fishing. Theoretical and empirical studies suggest that even light levels of fishing can prevent any potential recovery to fish populations and eliminate any benefits to surrounding areas (Westera et al. 2003; Denny and Babcock 2004; Byers and Noonburg 2007). Although the residents of Kia Island claimed to have reinstated the ban on fishing after 5 weeks of harvesting in 2008, several community members stated that they regularly saw boats fishing in the tabu area (S. Jupiter, pers. comm.). Once the communities established contacts with middlemen from local seafood companies during the harvest, the ease of market access may have eroded compliance with local management rules (Cinner et al. 2007; Clarke and Jupiter 2010). Additionally, if outsiders perceive that the tabu is no longer enforced and they will not be detected, they may be more likely to poach (Sutinen and Kuperan 1999), particularly given the light penalties for non-compliance under the Fiji Fisheries Act (Clarke and Jupiter 2010). 331 Studies from elsewhere in Fiji and from across Oceania demonstrate that customary management can effectively conserve marine resources under certain conditions (e.g., Cinner et al. 2005; McClanahan et al. 2006; Jupiter and Egli 2011). These conditions include secure community fishing rights or customary marine tenure, respect for the management authority from within and outside the community, broad awareness of management rules and boundaries, and limits on the duration and intensity with which managed areas may be harvested (Aswani and Hamilton 2004; Bartlett et al. 2009a; Clarke and Jupiter 2010). Indeed, the highly significant differences in fish abundance and biomass inside versus outside of the Cakaulevu tabu prior to the harvest confirm anecdotal reports of strong compliance with customary management rules since 2005. These conditions are additionally aided by the placement of customary fishing closures in naturally productive locations (Jupiter and Egli 2011). Customary management is less likely to succeed in areas with high population density and commercialization of marine resources (Cinner and Aswani 2007). Given that hundreds of periodically harvested closures have already been established in Oceania and that the practice is rapidly expanding (Bartlett et al. 2009b), we offer the following recommendations to improve knowledge and planning for management success: 1. 2. 3. To maintain sustainable fisheries benefits for the future, harvests must be controlled. Control measures can include restrictions on effort, gear, duration of opening, access and size of catch. Moreover, during a harvest, the amount extracted from a closure has to be lower than the amount of build-up since initiation of protection, and in many cases, substantial build-up may take years to decades to occur (McClanahan and Graham 2005; Babcock et al. 2010). Given that different species have very different catchability and recovery rates (e.g., McClanahan et al. 2007b; Babcock et al. 2010), species- or guild-specific take guidelines may need to be developed (Cohen and Foale 2011). Catch should be monitored to ensure that some breeding stocks remain to provide for local replenishment as well as replenishment of adjacent fishing grounds and neighbouring closures. Larger reef fish generally have greater egg production and tend to produce larger eggs, which may be more viable (e.g., Evans et al. 2008). Managers should undertake broad consultation to reach consensus for a verbal or written plan that details the process by which openings are authorized and the frequency, location, duration and gear type with which they can occur. This has been done for certain areas in Fiji (e.g., Kubulau District; WCS 2009), and generally results in fewer uncontrolled harvests. Because the Kia 123 Author's personal copy 332 Coral Reefs (2012) 31:321–334 Table 4 (a) Mean percent cover of each benthic strata. (b) PERMANOVA main effects results of differences in mean benthic cover. (c) PERMANOVA pair-wise comparisons of differences in benthic FAC LC OT North 0.8 46.6 South Control 0.4 0.8 49.7 36.0 North 0.6 South 0.5 Control cover within sampling period between each area. (d) PERMANOVA pair-wise comparisons of differences in benthic cover within area between each sampling period RM TA US 6.0 21.9 13.3 10.2 7.4 10.8 20.9 26.6 14.7 17.6 6.8 8.0 47.9 5.7 11.8 24.3 9.8 33.3 5.8 6.3 44.0 10.2 1.1 36.9 8.2 11.5 37.4 5.1 North 3.3 48.2 4.4 29.3 9.5 5.4 South 2.5 43.2 5.6 26.2 14.5 8.2 Control 3.0 40.2 5.5 27.0 17.3 7.2 (a) Benthic cover Pre-harvest During harvest 1 year later Source df SS MS Pseudo-F P(perm) Permutations (b) PERMANOVA main effects Sampling 2 28,779.0 14,390.0 29.2 0.0002 4,978 Area 2 6,739.9 3,369.9 6.8 0.0002 4,987 Site (area) 3 24,431.0 8,143.6 16.5 0.0002 4,986 Sampling 9 area 4 4,864.2 1,216.1 2.5 0.0064 4,979 4.9 0.0002 4,977 6 14,630.0 2,438.3 Residual Sampling 9 site (area) 162 79,948.0 493.5 Total 179 159,390.0 Pair-wise comparisons (c) Between area (d) Between sampling Level t P(perm) North–south NS 0.628 0.7458 Pre-During ** 2.120 0.0074 North–control NS 1.425 0.1248 Pre-1 year NS 1.254 0.1944 South–control * 1.767 0.0392 During-1 year *** 3.815 0.0002 Pre-harvest Level t P(perm) North During harvest South North–south *** 5.472 0.0002 Pre-During *** 6.045 0.0002 North–control South–control *** ** 3.420 2.102 0.0002 0.0040 Pre-1 year During-1 year NS *** 1.304 6.726 0.1616 0.0002 NS 1.213 0.2130 Pre-During *** 3.099 0.0008 1 year later North–south Control North–control * 1.672 0.0440 Pre-1 year NS 0.846 0.4796 South–control NS 0.666 0.7072 During-1 year *** 4.207 0.0002 Significant P(perm) values are indicated in bold FAC, fleshy algae/cyanobacteria; LC, live coral; OT, other soft substrate; RM, reef matrix; TA, turf algae; US, unconsolidated substrate For (c) and (d), significance level denoted as: NS not significant, * P(perm) \ 0.05, ** P(perm) \ 0.01, *** P(perm) \ 0.001 4. communities did not have a well-developed management plan, there was no impediment to extending their harvest until resources were exhausted. Finally, management success from the perspective of Pacific communities is most often related to achieving 123 objectives of providing for immediate food or income to meet cultural or social obligations (Foale et al. 2011). Conservation practitioners and resource managers need to realize this difference in objectives and work with fishers to raise awareness of the risk of stock collapse of Author's personal copy Coral Reefs (2012) 31:321–334 vulnerable species if openings are too frequent or harvests are too intense. If not regulated, even a single opening may cause a highly significant decline in fish biomass, requiring a long period for recovery. Acknowledgements The authors gratefully acknowledge the support from the chiefs and communities of Kia Island, staff of the WWF South Pacific Programme for assisting with management support, and the WCS staff and volunteers who assisted with field surveys and data entry: S. Dulunaqio, F. Januchowski-Hartley, U. Mara, W. Moy, A. Murphy, W. Naisilisili, Y. Nand, A. Patrick, T. Tui and P. Veileqe. We thank participants of the symposium on the effectiveness of community-managed MPAs (co-led by R. Weeks and S. Jupiter) at the Society for Conservation Biology’s 1st Marine Think Tank for insightful discussions to develop guidelines for periodic harvests, and we are grateful to the three reviewers and topic editor whose comments improved the manuscript. Funding for this work was generously provided by the David and Lucile Packard Foundation (2007-31847) and the Gordon and Betty Moore Foundation (540.01). References Abesamis RA, Russ GR (2005) Density-dependent spillover from a marine reserve: long-term evidence. Ecol Appl 15:1798–1812 Alcala A, Russ GR, Maypa AP, Calumpong HP (2005) A long-term, spatially replicated experimental test of the effect of marine reserves on local fish yields. Can J Fish Aquat Sci 62:98–108 Anderson MJ (2001) A new method for non-parametric multivariate analysis of variance. Austral Ecol 26:32–46 Anderson MJ, Willis TJ (2003) Canonical analysis of principal coordinates: A useful method of constrained ordination for ecology. Ecology 84:511–525 Anderson MJ, Gorley RN, Clarke KR (2008) PERMANOVA? for PRIMER: guide to software and statistical methods. PRIMER-E, Plymouth Ashworth JS, Ormond RFG (2005) Effects of fishing pressure and trophic group on abundance and spillover across boundaries of a no-take zone. Biol Conserv 121:333–344 Aswani S, Hamilton RJ (2004) Integrating indigenous ecological knowledge and customary sea tenure with marine and social science for conservation of bumphead parrotfish (Bolbometopon muricatum) in the Roviana Lagoon, Solomon Islands. Environ Conserv 31:69–83 Aswani S, Christie P, Muthiga NA, Mahon R, Primavera JH, Cramer LA, Barbier EB, Granek EF, Kennedy CJ, Wolanski E, Hacker SD (2012) The way forward with ecosystem-based management in tropical contexts: Reconciling with existing management systems. Mar Policy 36:1–10 Babcock RC, Shears NT, Alcala AC, Barrett NS, Edgar GJ, Lafferty KD, McClanahan TR, Russ GR (2010) Decadal trends in marine reserves reveal differential rates of change in direct and indirect effects. Proc Natl Acad Sci USA 43:18256–18261 Ban NC, Adams V, Almany GR, Ban S, Cinner J, McCook LJ, Mills M, Pressey RL, White A (2011) Designing, implementing and managing marine protected areas: emerging trends and opportunities for coral reef nations. J Exp Mar Biol Ecol 408:21–31 Bartlett CY, Manua C, Cinner JE, Sutton S, Jimmy R, South R, Nilsson J, Raina J (2009a) Comparison of outcomes of permanently closed and periodically harvested coral reef reserves. Conserv Biol 23:1475–1484 Bartlett CY, Pakoa K, Manua C (2009b) Marine reserve phenomenon in the Pacific islands. Mar Policy 33:673–678 333 Bonaldo RM, Bellwood DR (2011) Spatial variation in the effects of grazing on epilithic algal turfs on the Great Barrier Reef, Australia. Coral Reefs 30:381–390 Byers JE, Noonburg EG (2007) Poaching, enforcement, and the efficacy of marine reserves. Ecol Appl 17:1851–1856 Choat H, Bellwood DR (1985) Interactions amongst herbivorous fishes on a coral reef: influence of spatial variation. Mar Biol 89:221–234 Cinner J, Aswani S (2007) Integrating customary management into the conservation of coral reef fisheries in the Indo-Pacific. Biol Conserv 140:201–216 Cinner JE, Marnane MJ, McClanahan TR (2005) Conservation and community benefits from traditional coral reef management at Ahus Island, Papua New Guinea. Conserv Biol 19:1714–1723 Cinner JE, Marnane MJ, McClanahan TR, Almany GR (2006) Periodic closures as adaptive coral reef management in the Indo-Pacific. Ecology and Society 11: 31 http://www.ecologyandsociety.org/ vol11/iss31/art31 Cinner JE, Sutton SG, Bond TG (2007) Socioeconomic thresholds that affect use of customary fisheries management tools. Conserv Biol 21:1603–1611 Clarke P, Jupiter SD (2010) Law, custom and community-based natural resource management in Kubulau District (Fiji). Environ Conserv 37:98–106 Cohen P, Foale S (2011) Fishing taboos: Securing Pacific fisheries for the future? SPC Trad Mar Res Manage Knowl Inf Bull 28:3–13 Denny CM, Babcock RC (2004) Do partial marine reserves protect reef fish assemblages? Biol Conserv 116:119–129 Di Franco A, Bussotti S, Navone A, Panzalis P, Guidetti P (2009) Evaluating effects of total and partial restrictions to fishing on Mediterranean rocky-reef fish assemblages. Mar Ecol Prog Ser 387:275–285 Eggleston DB, Parsons DM (2008) Disturbance-induced ‘spill-in’ of Caribbean spiny lobster to marine reserves. Mar Ecol Prog Ser 371:213–220 Evans RD, Russ GR, Kritzer JP (2008) Batch fecundity of Lutjanus carponotatus (Lutjanidae) and implications of no-take marine reserves on the Great Barrier Reef, Australia. Coral Reefs 27:179–189 Foale S, Manele B (2004) Social and political barriers to the use of Marine Protected Areas for conservation and fishery management in Melanesia. Asia Pac Viewpoint 45:373–386 Foale S, Cohen P, Januchowski-Hartley S, Wenger A, Macintyre M (2011) Tenure and taboos: origins and implications for fisheries in the Pacific. Fish Fish 12:357–369 Froese R, Pauly D (2009) FishBase. http://www.fishbase.org Goetze JS, Langlois TJ, Egli DP, Harvey ES (2011) Evidence of artisanal fishing impacts and depth refuge in assemblages of Fijian reef fish. Coral Reefs 30:507–517 Gotanda KM, Turgeon K, Kramer DL (2009) Body size and reserve protection affect flight initiation distance in parrotfishes. Behav Ecol Sociobiol 63:1563–1572 Govan H, Tawake A, Tabunakawai K, Jenkins A, Lasgorceix A, Schwarz A-M, Aalbersberg B, Manele B, Vieux C, Notere D, Afzal D, Techera E, Rasalato ET, Sykes H, Walton H, Tafea H, Korovulavula I, Comley J, Kinch J, Feehely J, Petit J, Heaps L, Anderson P, Cohen P, Ifopo P, Vave R, Hills R, Tawakelevu S, Alefaio S, Meo S, Troniak S, Malimali S, Kukuian S, George S, Tauaefa T, Obed T (2009) Status and potential of locallymanaged marine areas in the South Pacific: meeting nature conservation and sustainable livelihood targets through widespread implementation of LMMAs. SPREP/WWF/WorldFishReefbase/CRISP, Suva, Fiji. (http://www.spc.int/DigitalLibrary/ Doc/FAME/Reports/Govan_09_Status_LMMAs.pdf) Januchowski-Hartley FA, Graham NAJ, Feary DA, Morove T, Cinner JE (2011) Fear of fishers: human predation explains behavioural 123 Author's personal copy 334 changes in coral reef fishes. PLoS ONE 6:e22761. doi: 22710.21371/journal.pone.0022761 Jupiter SD, Egli DP (2011) Ecosystem-based management in Fiji: successes and challenges after five years of implementation. J Mar Biol. doi:940710.941155/942011/940765 Lester SE, Halpern BS, Grorud-Colvert K, Lubchenco J, Ruttenberg BI, Gaines SD, Airame S, Warner RR (2009) Biological effects within no-take marine reserves: a global synthesis. Mar Ecol Prog Ser 384:33–46 Maliao RJ, White AT, Maypa AP, Turingan RG (2009) Trajectories and magnitude of change in coral reef fish populations in Philippine marine reserves: a meta-analysis. Coral Reefs 28: 809–822 McArdle BH, Anderson MJ (2001) Fitting multivariate models to community data: a comment on distance-based redundancy analysis. Ecology 82:290–297 McClanahan TR, Graham NAJ (2005) Recovery trajectories of coral reef fish assemblages within Kenyan marine protected areas. Mar Ecol Prog Ser 294:241–248 McClanahan TR, Marnane MJ, Cinner JE, Kiene WE (2006) A comparison of marine protected areas and alternate approaches to coral-reef management. Curr Biol 16:1408–1413 McClanahan TR, Graham NAJ, Calnan JM, MacNeil MA (2007a) Toward pristine biomass: reef fish recovery in coral reef marine protected areas in Kenya. Ecol Appl 17:1055–1067 McClanahan TR, Graham NAJ, Maina J, Chabanet P, Bruggemann JH, Polunin NVC (2007b) Influence of instantaneous variation on estimates of coral reef fish populations and communities. Mar Ecol Progr Ser 340:221–234 McCook LJ, Ayling T, Cappo M, Choat H, Evans RD, DeFrietas DM, Heupel M, Hughes TP, Jones GP, Mapstone B, Marsh H, Mills M, Molloy FJ, Pitcher CR, Pressey RL, Russ GR, Sutton S, Sweatman H, Tobin R, Wachenfeld DR, Williamson DH (2010) Adaptive management of the Great Barrier Reef: A globally significant demonstration of the benefits of networks of marine reserves. Proc Natl Acad Sci USA 107:18278–18285 Mills M, Jupiter SD, Pressey RL, Ban NC, Comley J (2011) Incorporating effectiveness of community-based management strategies in a national marine gap analysis for Fiji. Conserv Biol 25:1155–1164 Mumby PJ, Steneck RS (2008) Coral reef management and conservation in light of rapidly evolving ecological paradigms. Trends Ecol Evol 23:555–563 123 Coral Reefs (2012) 31:321–334 Robertson DR, Polunin NVC, Leighton K (1979) The behavioural ecology of three Indian Ocean surgeonfishes (Acanthurus lineatus, A. leucosternon and Zebrasoma scopas): their feeding strategies, and social and mating systems. Environ Biol Fish 4:125–170 Ruddle K, Hickey F (2008) Accounting for the mismanagement of tropical nearshore fisheries. Environ Dev Sustain 10:565–589 Russ GR, Alcala AC (1998) Natural fishing experiments in marine reserves 1983–1993: roles of life history and fishing intensity in family responses. Coral Reefs 17:399–416 Russ GR, Alcala AC (2003) Marine reserves: rates and patterns of recovery and decline of predatory fish, 1983–2000. Ecol Appl 13:1553–1565 Russ GR, Alcala AC (2004) Marine reserves: long-term protection is required for full recovery of predatory fish populations. Oecologia 138:1432–1939 Seidel H (2009) Evaluating the role of science in Community Based Adaptive Management of coastal resources in Fiji. University of Bremen, 101 pp (http://www.reefbase.org/download/download. aspx?type=10&docid=A0000005290_1) Sutinen JG, Kuperan K (1999) A socio-economic theory of regulatory compliance. Int J Soc Econ 26:174–193 Tawake A, Meo S, Cakacaka A, Aalbersberg B (2004) Community based biological monitoring training guide. Application: FLMMA project sites at Tagaqe Village, Tikina Korolevu-i-wai, Nadroga. University of the South Pacific, Suva, Fiji (http://www.reefbase.org/ download/open_download.aspx?type=10&docid=C0000000054_1) Vine PJ (1974) Effects of algal grazing and aggressive behaviour of the fishes Pomacentrus lividus and Acanthurus sohal on coralreef ecology. Mar Biol 24:131–136 WCS (2009) Ecosystem-Based Management Plan: Kubulau District, Vanua Levu, Fiji. Wildlife Conservation Society, Suva, Fiji (http:// www.wcsfiji.org/Resources/ManagementPlans/tabid/3431/ Default.aspx) Westera M, Lavery P, Hyndes G (2003) Differences in recreationally targeted fishes between protected and fished areas of a coral reef marine park. J Exp Mar Biol Ecol 294:145–168 Williams ID, Walsh WJ, Miyasaka A, Friedlander AM (2006) Effects of rotational closure on coral reef fishes in Waikiki-Diamond Head Fishery Management Area, Oahu, Hawaii. Mar Ecol Prog Ser 310:139–149