Reserve effects and natural variation in coral reef communities

1. No-take reserves are a common tool for fisheries management and biodiversity conservation in marine ecosystems. Despite much discussion of their benefits, data documenting many reserve effects are surprisingly scarce. Several studies have also been criticized for a lack of rigour so that changes within reserves cannot be separated from underlying natural variation and attributed unequivocally to protection. 2. We sampled both benthic (video quadrats) and associated fish communities (underwater visual censuses) in a well-enforced reserve in The Bahamas. Sampling was explicitly stratified by habitat ('Montastraea reef' and 'gorgonian plain'). To distinguish reserve effects from natural variation, we compared changes inside and outside the reserve with those seen at equivalent spatial scales in other reef systems in the Bahamian archipelago that lack reserves. Reserve-level differences in benthic or fish communities not documented in other reef systems are categorized as 'robust' effects. 3. Robust reserve effects were limited to Montastraea reefs. The reserve supported an average of O 15% more fish species per site compared to outside the reserve. This pattern was particularly driven by more large-bodied grouper, damselfish, and butterflyfish species inside the reserve. Increases in fish biomass and differences in community structure inside the reserve were limited to large-bodied groupers. Increased grazing pressure by parrotfishes in the reserve has lowered macroalgal cover, and caused previously undocumented changes in benthic community structure compared to sites outside the reserve. 4. Some reserve-level differences in fish communities were categorized as 'misleading' because equivalent differences were seen in other reef systems, and are likely to be caused by natural intra-habitat variation. Separation of robust and misleading results was only possible because of archipelago-scale sampling. 5. Synthesis and applications. The Bahamas represents a relatively lightly fished system within the Caribbean. However, cessation of fishing has still increased the mean number of species, the abundance of the most highly prized fishes and, through trophic cascades, altered benthic community structure. In certain habitats, reserves are clearly important for conserving fisheries and biodiversity. However, reserve effects must be explicitly separated from confounding variables to ensure conservation benefits are accurately identified and reported, and not oversold to managers and local stakeholders.