Current felid (Carnivora: Felidae) distribution, spatial bias, and occurrence predictability: testing the reliability of a global dataset for macroecological studies

The lack of information about species distribution, also known as Wallacean shortfall, targets most species, even charismatic animals like felids, limiting the performance of scientific study and conservation planning. The knowledge about species distribution (specifically occurrence data) is also affected by restriction of data availability and dispersion of existing information in many data sources. Therefore, we created and tested the reliability and validity of Felidae occurrence database at spatial resolution of 0.5 decimal degree by: (i) investigating knowledge bias generated by charismatic attributes, species detectability, and spatial data aggregation; and (ii) evaluating climatic niche predictability and species-niche patterns mediated by ecological or phylogenetic, which is the main concern of database utilization in macroecological studies. We found 3,549 spatially unique felid occurrence records around the world, considering a resolution scale of 0.5 decimal degree. Numbers of occurrences were positively related to body mass and distribution area (R-2 = 0.2, p < 0.001), revealing that charismatic attributes influences the choice of study species and making the Wallacean shortfall more critical for small-bodied than large-bodied felids. All species had systematic or random point patterns, which can efficiently capture climatic variation into species distribution. Climatic variables efficiently described the distributions of felids' niches, through two general proprieties, marginality and tolerance. Niche marginality has spatial and evolutionary signature, since it relates to time of divergence, body mass, and distribution area. Therefore, this database can support macroecological studies due to the non-spatial bias of occurrences and the species predictability of climatic variables, despite the species bias.