Combining environmental suitability and habitat connectivity to map rare or Data Deficient species in the Tropics

The IUCN Red List is a widely accepted system for classifying species’ risk of extinction, based on quantitative criteria. Although IUCN discourages the liberal use of the category “Data Deficient” (DD), most assessed groups have a large number of their species assigned to this category, especially in the Tropics. Therefore, DD species can introduce considerable uncertainty into estimates of proportions of threatened species, and research focused on elucidating the true status of those species should be a priority. Here we propose a simple method to gather information on geographic distribution and guide the search for new populations of rare, small-ranged, forest species, using the literature, online data, and standard GIS procedures. The method involves: (i) creating a geographic distribution model; (ii) selecting the environmentally suitable sites from that model; (iii) removing sites that have lost natural vegetation; and (iv) removing habitat networks that are too small and/or isolated, based on thresholds established from known occurrence records and the literature for ecologically similar species. As a case study, we use Lonchophylla peracchii, a recently described forest-dependent bat endemic to southeastern Brazil. We found that environmentally suitable sites for L. peracchii are already heavily deforested, confirming habitat loss as a major threat. Importantly, we identified five priority sites to search for the species outside of its currently known distribution. From that, we discuss its likely status based on IUCNʹs Criterion B2 (Extent of Occurrence). This method could be useful for other poorly known forest species, especially in the Tropics where most of these species are, and funding for research and fieldwork is scarcest. Currently there are 1910 terrestrial vertebrates in tropical forest worldwide classified as DD that could be evaluated using this method, provided that they have at least 5–10 occurrence records.