Protected area evaluation for the conservation of endangered Amazon river dolphins (Inia geoffrensis)

The potential value of protected areas for the conservation of cetaceans is widely recognized; however, few evaluation methods exist to assess their effectiveness. In this study, a modeling approach based on long-term mark-recapture/resight data was used to assess the effectiveness of a Brazilian reserve in protecting endangered Amazon River dolphins or boto (Inia geoffrensis), a species killed for use as fish bait. We built an annual discrete-time model with subdivisions based on hydrological periods and age classes. It included transition probabilities in and out of the reserve that were estimated utilizing multi-state mark-recapture models. To evaluate five reserve configurations, we re-estimated the transition probabilities to represent changes in the reserve boundaries. Model predictions showed that four scenarios, including one representing the existing boundaries, would be insufficient to protect the local boto population (n(0) = 528) and a steep decline in abundance would occur in the next 50 years (0 <= n(50) <= 108). However, one reserve configuration, encompassing both flooded forests and adjacent river habitats, and including beaches and channel/floodplain entrances, resulted in a nearly stable population in the same time frame (n(50) = 515). These findings suggest that, with careful design and efficient management, protected areas could be an effective conservation tool for boto populations that exhibit site fidelity to Amazonian floodplains. With appropriate mark-recapture/resight data, the modeling framework employed could be extended to evaluate and design protected areas for populations of other species in other systems.