Uso do habitat e ocupação por carnívoros em uma reserva de uso sustentável na Amazônia Central, Brasil

Abstract

Studies focusing on habitat use and species occurrence provide important information about the ecology, distribution, and population dynamics of species, enabling the creation of improved conservation planning strategies. Carnivore mammals deserve special attention due to their vulnerability to habitat loss and sensitivity to anthropic impacts, as well as their role in the regulation of populations of species at lower trophic levels. We evaluated habitat use and site occupancy by five carnivores using camera-trap data obtained in Piagaçu-Purus Sustainable Development Reserve (PP-SDR), central Amazonia, Brazil. Camera-traps were installed along 24 trails in terra firme forest in two sampling sessions: one at the Aiapuá sector of PP-SDR, from March to May 2011, and the other at Uauaçu sector, from May to June 2011. Final number of sampling points was 107, within a study area of approximately 1380 km². At each point, we characterized the habitat, using environmental variables obtained on site and through remotely sensed data. We evaluated site occupancy probabilities for carnivores and their relationships with environmental variables in two spatial scales, site (each camera-trap, n = 107; ~50 ha) and local (each trail, n = 24; ~1100 ha), using maximumlikelihood occupancy modeling techniques implemented in the software PRESENCE 3.1. Model selection by Akaike s Information Criteria (AIC) showed that understory density (vegetation between 0.5 m and 1.5 m from soil level) positively affected the occurrence of ocelots (Leopardus pardalis) at site scale and negatively affected short-eared dogs (Atelocynus microtis) at local scale. Short-eared dog s occurrence was strongly related to the areas of Brazil-nut extraction activity (Bertholletia excelsa), with 94% of its registers occurring in these areas. Pumas (Puma concolor) occurred preferentially in points with low drainage density (meters of streams/hectare), possibly due to the inundation of the areas with high stream density during the wet season, when camera-trapping occurred. The presence of tayras (Eira barbara) and jaguars (Panthera onca) were weakly related to the variables analyzed in either scale, suggesting they are not determinant of these species occurrence at the study area. Detectability of ocelots and pumas had a positive relationship with river level, probably reflecting their movement to dry terra firme forests during the flooding season, when rivers levels are high. These results indicate that castanhais are important areas for short-eared dog conservation, and Brazil-nut extraction activities must be considered in the conservation planning strategies for the species. Furthermore, ocelot and puma dependence on terra firme forests indicates that conservation units designed to protect wetlands in the Amazon must also include large tracts of terra firme, to where such species may move during the flooding season. Distance to human communities was not related to occupancy rates for any carnivore species, suggesting that human activities in PP-SDR do not negatively impact carnivore site occupancy. This suggests that PP-SDR has so far accomplished its objective of conciliating the use of natural resources by human communities with species conservation, at least for carnivore mammals.