Historical abundance of juvenile commercial fish in coastal habitats: Implications for fish habitat management in Canada

An important component of science-based fisheries policy is the provision of habitat adequate for population renewal. In Canada, the Fisheries Act pays little attention to managing fish habitat, and was further weakened by changes enacted in 2012. Specifically, determining the role of fish habitat in contributing to fisheries and fish stock recovery is challenging when many stocks have severely declined and no longer occupy former habitats. This study compared the abundance of juvenile fish in coastal vegetated habitats before and after collapse or decline of groundfish stocks in Atlantic Canada. This comparison was done by compiling past studies that surveyed juvenile Atlantic cod (Gadus morhua) and pollock (Pollachius virens) in vegetated habitats across three provinces. Two studies were repeated, and one that already had post-collapse data was analyzed to quantify long-term changes in juvenile abundance. In all three cases substantial reduction in juvenile abundance coincided with declines in adult stocks. However, juvenile fish still occur in coastal habitats and could aid in adult stock recovery. The current version of the Canadian Fisheries Act requires presence of an ongoing fishery to trigger habitat protection. This is problematic as low fish abundance may lead to lowered habitat protection and potentially habitat degradation, with less or lesser-quality habitat for fish in the future. Thus, recommendations are made to repeal the 2012 Fisheries Act changes and enhance current fish habitat legislation. Using a precautionary approach for coastal fish habitat management, particularly in valuing its potential for fish stock recovery, would strengthen Canadian fisheries management.     

Coral bleaching and habitat effects on colonisation of reef fish assemblages: An experimental study

Degradation and mortality of corals is increasing worldwide and is expected to have significant effects on coral reef fish; hence studies on these effects are essential. In the present study, a field experiment was set up within Mafia Island Marine Park in Tanzania (East Africa) to examine the effects of bleaching and habitat structure on colonisation of coral reef fish assemblages. Live and bleached staghorn coral Acropora formosa was transplanted onto plots in a site dominated by sand and rubble, and the experimental design comprised of three treatments: live coral, bleached coral and eroded coral rubble. There was an immediate increase (within 24 h) in fish abundance and diversity in the two treatments with standing corals. Overall, live and bleached coral plots showed similar effects, but differed from the eroded coral plots which had a much lower abundance and diversity of fish. In general, fish species diversity changed with time over the study period while fish abundance did not. Multivariate analyses showed that while there were differences in fish assemblage structure between standing corals and the eroded coral treatment, there was neither a difference between live and bleached coral treatments nor any temporal effects on fish assemblage structure. Our findings suggest that physical structure and complexity of habitat have stronger effects on colonisation of reef fish assemblages than changes in coral health (such as bleaching) which do not affect coral structure. This may have important implications for appropriate coral reef management.     

Food utilisation by coastal fish assemblages in rocky and soft bottoms on the Swedish west coast: Inference for identification of essential fish habitats

The concept of essential fish habitats (EFHs) is widely accepted for conservational and management purposes. EFHs are often considered as high quality habitats for fisheries species and subsequently of high values for society. In this study, fish and Substrate-Associated Prey (SAP) were sampled during the productive summer season 1998 (fish) and 2003 and 2004 (SAP) in shallow coastal rocky- and soft-bottom habitats on the Swedish west coast. The aim was to study the spatial and monthly variation of SAP as well as abundance and biomass of fish, and to examine if food items found in the diet of the fish assemblage were derived from SAP. We also examined if the diet of Ctenolabrus rupestris, a resident and abundant fish species in the shallow coastal habitats, and the diet of four seasonally abundant and commercially important fish species (Gadus morhua, Pleuronectes platessa, Salmo trutta and Scomber scombrus) were derived from SAP. There were significantly higher mean species number and abundance of the SAP assemblage on rocky compared to soft bottoms and the highest values were found on the rocky bottoms in August and in the shallowest (0–3 m) depth strata. There were no significant differences in number of fish species caught in the two habitats, although mean number of fish and mean biomass were significantly higher on rocky bottoms. Both habitats showed the same seasonal variation and the highest values of number of fish species, abundance and biomass were observed in June. On rocky bottoms, gastropods and amphipods were the most frequent food items in the diet of the entire fish assemblage and these items were also the most abundant SAP in this habitat. The dominant food items of the soft-bottom fish assemblage were decapods and fish, which were not common SAP. However, except for S. scombrus, the diet of the selected fish species showed a strong association to the SAP availability. Gadus morhua displayed the strongest association to SAP on rocky bottoms and P. platessa and C. rupestris to SAP on soft bottoms. Further, for C. rupestris, multivariate statistical analysis showed a significant association to the SAP assemblage on both rocky and soft bottoms. These results provide vital new information for the management and conservation of Essential Fish Habitats on the Swedish west coast.     

Interdecadal variability of anchoveta abundance and overcapacity of the fishery in Peru

Paleontological and historical stock abundance estimates indicate that pelagic fish populations inhabiting upwelling ecosystems undergo large interdecadal variations in abundance with amplitudes equal to, if not larger than, the interannual variability. The interdecadal variability is characterized by periods of high and low abundance, termed “pseudo-cycles”, because of their irregular periodicity. Fisheries targeting small pelagic fish suffer from overall overcapitalization, like many other fisheries, but also from an additional overcapitalization problem: a phase displacement between rapid fish population decreases and a slower disinvestment which follows. This lag produces economic hardship.Here we document the fish:fishery relationship for the Peruvian anchoveta. Anchoveta pseudo-cycles of 20 to >100 years are observed, with the present stock abundance most likely located near upper part of the cycle. Fleet overcapacity expressed as the proportion of unused present capacity is estimated at 72% and processing overcapacity at 89%. A simple bio-economic model demonstrates the risks associated with the pseudo-periodicity in fish stock abundance in conjunction with fishery investment, open access, and overcapacity: a timing bomb for the fishing sector. The lag between disinvestment and decrease in fish abundance is quantified. A reduction of the fishing and processing capacity and measures to decrease the investment lag are recommended to limit the social, economical and political tensions that will result from the expected decrease in stock abundance. Finally, some management options to reduce these risks are discussed.     

Effectiveness of conventional management in Mediterranean type artisanal fisheries

Artisanal fisheries throughout the Mediterranean region are managed mostly by applying gear specific regulations. The data about the extent and dynamics by which littoral fish resources respond to commonly proposed changes of such regulations are lacking. Here the results of a 15 year (1995–2009) monitoring programme of littoral fish resources in a pilot region of island Vis aquatorium, central Adriatic Sea, are reported with the scope of investigating whether a more restrictive fishing regime (encompassing an increase in minimum inner layer mesh size of trammel net from 28 to 40 mm and excluding the trammel net from subsistence artisanal fishing) that has progressively been put in place during the study period has been accompanied by expected positive changes of littoral fish resources’ abundance, biomass and structure. Significant increases over time were observed in most of the community indices analyzed (abundance and biomass catch per unit effort, diversity indices) as well as a directional change in abundance and biomass catch composition of littoral fish resources. Positive responses were, however, primarily related with a recovery of Mullus surmuletus stock. Additionally, time-series analysis of ABC curves and their corresponding W index revealed that fish community in the study area was still moderately disturbed. Limited extent of conventional management restorative potential is further exacerbated by social issues – noncompliant behavior mainly among subsistence artisanal fishers and consequent resentment of commercial artisanal fishers. In order to improve further the state of the resources a more comprehensive set of management measures incorporating closed areas and a new approach actively involving fishers in the management process should be adopted.     

Regime shifts in the North Pacific: early indications of the 1976–1977 event

A major environmental change, since dubbed a “regime shift”, occurred in the North Pacific in 1976–1977. However, it was not identified as such until the early 1990s. Recognition was complicated by the presence of higher frequency, interannual variations in both the physical and biological time series. In retrospect some symptoms of the event were apparent within a few years of the shift. These included changes in SST distribution and zooplankton abundance as well as in the recruitment, abundance, and distribution of commercial fishes. Such observations, while not diagnostic at the time, are consistent with the larger scale analyses of recent years.Of particular interest was the demonstrated impact on commercial fish stocks, indicating that early recognition of such events could be useful for fishery management. There is some evidence that appropriate ecosystem data, by integrating the higher frequency variations, when combined with physical data might have advantages for early detection of major environmental changes. While fishery data can clearly indicate that change has occurred, their utility for this purpose is often limited because of inadequate sampling, length of fish life cycles, the complicating effects of fishing and delays in processing of the data. Selected zooplankton species might be the most useful early indicators of environmental change if sampling coverage could be increased and speedy means of identification developed.     

The influence of an extreme drought event in the fish community of a southern Europe temperate estuary

Between 2003 and 2006, a severe drought occurred throughout the Mondego River catchment's area, inducing lower freshwater flows into the estuary. As a consequence, both 2004 and 2005 were considered as extreme drought events. From June 2003 to June 2006, the fish assemblage of the Mondego Estuary was sampled monthly in five stations during the night, using a 2 m beam trawl. Fish abundance was standardized as the number of individuals per 1000 m² per season and the assemblage was analyzed based on ecological guilds: estuarine residents, marine juveniles, marine adventitious, freshwater, catadromous and marine species that use the estuary as a nursery area. A total of 42 species belonging to 23 families were identified, with estuarine residents and nursery species dominating the fish community. Variations in the fish community were assessed using non-metric MDS, being defined as three distinct periods: summer and autumn 2003, 2004/2005 and winter and summer 2006. The main drought-induced effects detected were the depletion of freshwater species and an increase in marine adventitious in 2004/2005, due to an extended intrusion of seawater inside the estuary and a significant reduction in abundance during the driest period of estuarine resident species. Nevertheless, from the management point of view, it could be stated that although some variations occurred due to environmental stress, the main core of the Mondego Estuary fish community remained relatively unchanged.     

Influence of habitat structure on fish assemblage of an artificial reef in southern Brazil

Habitat complexity strongly influences reef fish community composition. An understanding of the underlying reasons for this relationship is important for evaluating the suitability of artificial reef (AR) habitats as a marine resource management tool. We studied the influence of AR habitat structure on fish assemblage composition off the southern coast of Brazil. We found that reef blocks with greater area and number of holes possessed the greatest fish species richness and abundance. Reef blocks with greater complexity had higher abundance of almost 30% of fish species present. Natural reef (NR) and AR were different in their fish species composition, trophic structure and categories of water column occupancy by fish (spatial categories). Although NR was more diverse and harboured more trophic levels, AR presented the higher abundances and the presence of distinct fish species that underlined their importance at a regional scale. The greater availability of sheltering habitat where hard substrate is scarce, together with their frequent use by economically important species, make AR a useful tool for coastal management when certain ecological conditions are met.     

Dysfunction by design: Consequences of limitations on transferability of catch shares in the Alaska pollock fishery

The Eastern Bering Sea fishery for pollock is one of the largest fisheries in the world and has often been touted as an example of sustainable fisheries management. Yet, sustainability requires more than protection of the biological productivity of the targeted fish species. It requires preservation of the flows of net social and economic benefits of the fishery, and flexible governance and management institutions that allow for adaptation to changes in fish abundance as well as changes in consumer demand and in the prices of key inputs. From an economic perspective, the Alaska pollock fishery was unsustainable before passage of the American Fisheries Act (AFA) in 1998. The AFA permanently divided the pollock Total Allowable Catch (TAC) giving a 10% allocation to the Community Development Quota (CDQ) program and partitioning the remainder among four fishing sectors. It allowed CDQ to be leased to any sector but largely prohibited leasing or sale between other sectors. These limits on intersectoral transfers have reduced the ability of sectors to respond to changes in the abundance and distribution of pollock, implementation of management measures intended to benefit other species, and changes in the prices of input factors and products. This paper explores these consequences.     

Own-price elasticity of open access supply as a long-run measure of fish stock abundance

Extensive criticisms of the management of marine fisheries periodically arise for a variety of reasons. While the complexity of the US fisheries management process is partially to blame, these reasons manifest themselves in statements that include a perceived decline in living marine resource abundance levels, a proposed change in the regulatory infrastructure, or improvements in stock abundance that some feel should allow for new entrants. Following these criticisms are calls to end the federal management process or at the very least the replacement of the fishery management agency leadership. These demands driven by rent seeking behavior use short-term declines in abundance as justification when the long-run abundance measures would have indicated improvements in stock abundance and achievement of management objectives. Empirical, long-run, stock abundance indices are estimated using synthetic demand methodology, based on bioeconomic fishery theory, applied in a general and partial equilibrium theory framework to determine if maximum sustainable yield (MSY) management targets have been achieved and maintained for four fisheries considered to be successfully managed and one characterized as a socio-economic disaster. The abundance of a fish stock is more influence by extra-market and extra-fishery forces than by fisheries managers who base allocation decisions solely on single species stock assessments.     

Long-term trends in the structure of eastern Adriatic littoral fish assemblages: Consequences for fisheries management

Long-term interannual changes in abundance, biomass, diversity and structure of littoral fish assemblages were examined between 1993 and 2009 by experimental trammel net fishing up to six times per year, within the warm period - May to September, at multiple areas along the eastern Adriatic coast with the aim of testing for the consistency of patterns of change across a large spatial scale (∼600 km). The results revealed spatially consistent increasing trends of total fish abundance and biomass growing at an average rate of 15 and 14% per year, respectively. Of the diversity indices analysed, the same pattern of variability was observed for Shannon diversity, while Pielou evenness and average taxonomic distinctness measures Δ^∗ and Δ⁺ showed spatial variability with no obvious temporal trends. Multivariate fish assemblage structure underwent a directional change displaying a similar pattern through time for all the areas. The structural change in fish assemblages generally involved most of the species present in trammel net catches. A large pool of fish species responsible for producing the temporal pattern of assemblage change was relatively different in each of the areas reflecting a large geographic range covered by the study. An analysis of 4 fish species (Symphodus tinca, Pagellus erythrinus, Mullus surmuletus, Scorpaena porcus) common to each of the study areas as the ones driving the temporal change indicated that there were clear increasing trends of their mean catches across the years at all the study areas. A common pattern among time trajectories across the spatial scale studied implies that the factor affecting the littoral fish assemblages is not localised but regional in nature. As an underlying factor having the potential to induce such widespread and consistent improvements in littoral fish assemblages, a more restrictive artisanal fishery management that has progressively been put in place during the study period, is suggested and discussed.     

Size matters: experimental partitioning of the strength of fish herbivory on a fringing coral reef in Moorea,French Polynesia

While the importance of top‐down control by grazers in maintaining tropical reef community structure is well known, the effect of ‘fishing down the food web’, which simultaneously changes the abundance and size of herbivorous fishes, has received less attention. As many fishing practices target the biggest fish and regulations often set minimum size limits, understanding size‐dependency of this controlling force is important. We evaluated the hypothesis that reduction in the abundance and size of fish, regardless of species identity, reduces the role of herbivory in controlling algal abundance on reefs and assessed variation in efficacy of this top‐down control on different types of common macroalgae. We quantified herbivory rates within cages of differing opening sizes on assemblages of four common algal species (Padina boryana, Dictyota bartayresiana, Halimeda opuntia and Galaxaura fasciculata) on two fringing reefs in Moorea, French Polynesia. Small acanthurids (<15 cm length) were the dominant herbivorous fish while other herbivorous fish were rare. For the two most palatable algae, the majority of herbivory occurred in open plots, with herbivory reduced by >50% in the largest opening cages (6 × 6 cm) where the maximum size fish entering was 12 cm in length (mean = 7.6, SE = 0.4). Fish entering medium (4.5 × 4.5 cm openings, maximum fish length = 8 cm, mean = 6.3, SE = 0.4) and small (3 × 3 cm openings, no fish observed entering) cages had herbivory rates approximately equal to the control treatment (1 × 1 cm openings). Consumption varied among algal species, with minimal consumption of physically and chemically defended algae and no pattern across treatments. Our results demonstrate a need for management plans to not only maintain the overall abundance of herbivorous fish but to protect the largest sizes for effective top‐down control of algal communities.     

Relative and combined effects of habitat and fishing on reef fish communities across a limited fishing gradient at Ningaloo

Habitat degradation and fishing are major drivers of temporal and spatial changes in fish communities. The independent effects of these drivers are well documented, but the relative importance and interaction between fishing and habitat shifts is poorly understood, particularly in complex systems such as coral reefs. To assess the combined and relative effects of fishing and habitat we examined the composition of fish communities on patch reefs across a gradient of high to low structural complexity in fished and unfished areas of the Ningaloo Marine Park, Western Australia. Biomass and species richness of fish were positively correlated with structural complexity of reefs and negatively related to macroalgal cover. Total abundance of fish was also positively related to structural complexity, however this relationship was stronger on fished reefs than those where fishing is prohibited. The interaction between habitat condition and fishing pressure is primarily due to the high abundance of small bodied planktivorous fish on fished reefs. However, the influence of management zones on the abundance and biomass of predators and target species is small, implying spatial differences in fishing pressure are low and unlikely to be driving this interaction. Our results emphasise the importance of habitat in structuring reef fish communities on coral reefs especially when gradients in fishing pressure are low. The influence of fishing effort on this relationship may however become more important as fishing pressure increases.     

Does small-scale vertical distribution of juvenile schooling fish affect prey availability to surface-feeding seabirds in the Wadden Sea?

Food availability is a key variable influencing breeding performance and demography of marine top predators. Due to methodological problems, proportionality between fish abundance and availability is often assumed without being explicitly tested. More specifically, better breeding performance of surface-feeding seabirds at times of large prey stocks suggests that prey availability is also a function of prey abundance. Using vertically resolved stow net sampling we tested whether local abundance and length composition of pelagic fish are reliable predictors of the availability of these fish to surface-feeding Common Terns (Sterna hirundo) breeding in the German Wadden Sea. Prey fish were found to concentrate below the maximum diving depth of the terns. Individuals caught close to the surface were in most cases smaller than conspecifics caught at greater depth. Correlations between fish abundance within and out of reach of the terns appeared to be both species- and site-specific rather than driven by overall fish abundance. Vertical distribution patterns of the terns' main prey fish could be explained as anti-predator behavior, reducing prey availability to the terns. In 2007, when breeding performance was much better than in 2006, herring and whiting were much more abundant, suggesting that overall prey abundance may also increase prey availability in habitats other than those represented by the stow net sampling.     

Abundance,biomass and species composition of nearshore fish assemblages in Ghana,West Africa

Many of the world's fish populations are overexploited, including Ghana's fish resources. This study examines spatio-temporal trends in fish species composition in relation to biotic, abiotic and anthropogenic factors, towards achieving better-informed management of the beach-seine fisheries. Fishery-dependent data were collected between November 1999 and October 2001 from 94 beach-seine hauls fished at two stations along the Ghanaian coast. The catch consisted of fish, crustaceans, other invertebrates, and macroalgae. Generally, juveniles of species that are exploited by offshore fisheries were found in the catches. Species abundance and their occurrence generally peaked between November and January. Duration of solar radiation and tide level appeared to be important predictors of fish biomass. It is important to exploit the nearshore fish assemblage sustainably because of its nursery role. A co-managed (fishers and government) three-month ban on beach-seining (between May and July) is recommended as the most appropriate control measure towards the sustainability of Ghanaian fish stocks.     

Potential impacts of alternative regulatory interventions on the recreational value of angling on the Breede River estuary,South Africa

This study investigated the support for, and potential impacts of, alternative management measures to address declining fish stocks in the Breede River estuary. A survey of residents and visitors was carried out during 2011–2012. Angling was the most important attraction of the estuary, and accounted for at least R 25 million in annual expenditure in the area. The majority of recreational anglers were concerned about the status of fish stocks and felt that additional regulations were necessary. There was greatest support for banning the trolling of artificial lures, and majority support for a closed area and an upper size limit for one of the main species targeted – the dusky kob Argyrosomus japonicus. The options of a night ban on fishing and a closed season received little support. Ratings of alternative scenarios suggested that anglers were more sensitive to fish abundance than to regulations. Modelled results showed that regulations resulting in increased abundance of fish, and large A. japonicus in particular, would have a positive impact on angling value of the estuary, whereas further declines in fish stocks would result in a decrease in expenditure in the area.     

The role of archaeological and cultural-historical records in long-range coastal fisheries resources management strategies and policies in the Pacific Islands

The usefulness of conventional fisheries science for long-term fisheries management and policies in the resource-poor islands of the Pacific is very limited. Fishery managers can, however, make use of such alternative sources of information as archaeological and historical data to develop fishery management initiatives. Most Pacific Islands were settled by humans over the last three millennia, although human settlement in Western Melanesia dates from the late Pleistocene, a period of about 40 millennia. Archaeological studies over the past half-century contain information on the long-term subsistence exploitation of fish and invertebrates from nearshore coral reefs and lagoons. Molluscs appear to have been a very important food source for early human colonists in the Pacific Islands, but declines in abundance through harvest pressure and environmental changes forced a greater reliance on fin-fish capture and on agriculture. There is no firm evidence from the archaeological record to suggest that long-term subsistence exploitation of reef fish populations has had any serious negative effects on abundance or structure of reef fish communities. Sources for more recent historical information are also exemplified and briefly examined for their usefulness in assessing the impacts of commercial fishing. It is argued that fishery managers who adopt a data-less approach to fishery management could profitably use these sources of information to enrich their assessment of the impacts of various management scenarios.     

Recreational fishing in a time of rapid ocean change

Fishing is an important recreational activity for many Australians, with one in every four people participating every year. There are however many different pressures exerted on Australian fish stocks, including climate-related changes that drive changes in local fish abundances. It is inevitable that recreational fishers will need to adapt to these changes. When resource abundance alters substantially, user adaptation to the new situation is required and policies and incentives may need to be developed to encourage behaviour change. It is important to correctly anticipate fisher's response to these policies and incentives as much as possible. Improved understanding of recreational fisher's likely adaptation decisions and the nature and timing of these decisions can help avoid unintended consequences of management decisions. Based on a survey of recreational fishers in the south-east Australian climate hotspot, we identify 4 relevant dimensions to recreational fisher's behavioural adaptation. There are differences in adaptation timing (early, late, and non-adaptors). Non-adaptors are characterised by greater cultural attachment to fishing and stronger perceptions of the factors that influence abundance change. The fisher's preferred adaptation responses and the timing of the behavioural response differs between decreasing versus increasing fish abundance. Insight into perspectives and expectations on how recreational fishers might adapt to changes is useful to develop a set of behavioural incentives that appeal to different groups but remain efficient and effective in their implementation. Such knowledge can create new pathways to achieve meaningful and targeted adaptation responses for different types of recreational fishers.     

Composition, structural characteristics and temporal patterns of fish assemblages in non-tidal Mediterranean lagoons: A case study

The importance of transitional water ecosystems as nursery habitats and feeding grounds for fish species is well-known. Detailed studies of colonization patterns of fish guilds in response to biotic and abiotic drivers are however unevenly distributed among ecosystem types. We address here the temporal variability of fish assemblages in small non-tidal lagoons in the Mediterranean basin. The study was carried out at the Acquatina lagoon (Lecce, Italy) where four stations, situated in two habitat types along a confinement gradient, were sampled twice per month for one year with fyke nets. Forty-five taxa ranging across 20 families were collected, with the most abundant species, Atherina boyeri, accounting for more than 95% of total abundance. Pooling all species together (excluding sand smelt), the structural features of the assemblage, relative abundance of families, and abundance of individual species all showed significant temporal patterns. Mean abundance peaked in Summer and Autumn and fell in Winter, whereas taxonomic richness and diversity were highest in Summer and lowest in Spring. Within the fish assemblage, multivariate ordination showed temporal segregation of species belonging to the same family or genus and expected to be functionally similar, suggesting that they avoid competition for space and resources by timing inward migration and peak occurrence differently. Of the environmental driving forces, which also showed temporal patterns of variation, salinity was the main factor affecting the distribution of individuals and species. The catch of young individuals of several marine species confirmed the role of this small lagoon as a nursery and feeding area, and emphasized the need for further studies.     

Mangrove fish: a comparison of community structure between forested and cleared habitats

The fish communities of mangrove and cleared sites were investigated in Gazi Bay, Kenya. Five forested sites were compared with paired sites that had been cleared of mangroves by human activity. Forested sites included plantations and natural stands of Sonneratia alba and natural Rhizophora mucronata stands. Two methods of stake netting were used to take quantitative samples; method one used a single 100-m-long, 18-mm mesh net, method two used paired 24-m-long, 1-mm mesh nets—samples were taken during seven different months in 2002. Mean abundances of fish found in mangrove and cleared sites, respectively, were 0.004 m⁻² and 0.014 m⁻² (method 1) and 0.21 m⁻² and 0.25 m⁻² (method 2). Thirty species were sampled, 12 of which were found exclusively in mangrove habitats and 10 of which were limited to cleared sites. The most abundant species in mangrove plots was Atherina afra (although it was only found in two, large catches); the most abundant in cleared plots was Gerres oyena (found frequently). Mean abundance (using data pooled for all sites) was significantly higher in cleared, compared with forested, sites, and multivariate analysis showed significantly different community structures in the two habitat types. There was large variation in catch rates between dates and sites, with one forested site recording no catches at all. These results do not support the predator refuge hypothesis (which predicts higher abundance of juvenile fish inside mangroves). The low abundance of fish recorded in the mangrove sites may have been due to site-specific factors determining fish abundance within mangrove forests, to the sampling techniques used or to relatively high turbidities at these sites.