Pre-existing spatial patterns in fish abundances influence species-specific responses in a Caribbean marine reserve

No-take marine reserves have been used to promote biodiversity conservation and have increasingly been implemented with the goal of rebuilding depleted fish stocks. Yet, marine reserves may cause some species to decrease in abundance while others increase, in part because species are not equally impacted by the effort displaced from the reserve into surrounding fished areas. We evaluate temporal patterns in the abundance and distribution of reef fishes at Glover's Atoll, Belize, where a no-take reserve was enforced in 1998. For the set of commercial species investigated in our study, the most important variable in predicting species responses to reserve protection was the pre-reserve spatial distribution of fishes with respect to reserve boundaries. Variation in species responses to protection could not be attributed to benthic habitat degradation, but maximum body length and trophic level were both positively correlated with response to protection. We suggest that reserve implementation caused mortality rates to increase for some fish species that were originally distributed in higher abundances outside reserve boundaries, leading to declines in abundance in these populations. These findings highlight the importance of incorporating knowledge on spatial distributions of fishes when designing and monitoring marine reserves. Our methods can be used by reserve managers to predict the responses of individual species to protection at the time of reserve implementation, thereby facilitating agreement between the expectations of fishery managers and the needs of stakeholders.