Influence of spatial extent on habitat suitability models for primate species of Atlantic Forest

Abstract

The Atlantic Forest biome is essential to the conservation of primates, the mammals with the highest number of endangered species. Habitat suitability modeling is extensively used in ecology and conservation, but the influence of the spatial extent used to train and test these models have not yet been assessed. Here we evaluated how the spatial extent definition influences the suitability surfaces and distributions obtained from habitat suitability modeling for the Atlantic Forest primate species. We built habitat suitability models for each species in three different extents. We used extents generally defined when models are built for more than one species (combining the spatial extent for all species or the complete biome) and the extent generally used for single species (encompassing all occurrence records for the species). Although the suitability spatial patterns did not vary among extents, the suitability frequency distributions were different. The extent can also influence the distribution area obtained from these models. Habitat suitability modeling can be beneficial in the environmental sciences and the efficient use of this tool requires the awareness of the effects of model construction decisions.