Modeling ocelot (Leopardus pardalis) distribution in the southern limits in Brazil

Ocelot is a Neotropical wildcat with a broad geographic distribution. The southernmost region of its distribution is an ecotone zone between biomes Atlantic Forest and Pampa in Brazil. Ocelot is rarely seen in this region, especially in the Pampa, being one of the most endangered wildcats in this region. The objective of this study is to predict the potential distribution of ocelots in the southern limit of it known distribution and to measure the percentage of protected areas coverage. We collected 24 records of the species in the state of Rio Grande do Sul and we modeled the potential distribution using seven distinct algorithms tested using two evaluation metrics (TSS and AUC). The variables with the highest percentage contribution to the models were: BIO8, BIO19, BIO7 (related to temperature and precipitation) and NDVI. Additionally, we observed that of the areas considered most suitable for the species (8,388 km2), only (39 km2 or 0.46%) were included in the protected areas and these were primarily located in the Atlantic Forest biome. By contrast, the most suitable areas for the species in the Pampa biome were not covered by protected areas. We consider this to be a serious issue considering the species is endangered in the state.