Rapoport's rule and determinants of species range size in snakes

Aim: Understanding determinants of species' range size is paramount to explaining global ecological patterns and estimating extinction risk of species. Here, we examined whether a sample of 536 snake species exhibits a latitudinal gradient of range size in support of Rapoport's rule, and determined predictors of range size from a set of environmental and biological factors. Location: Global. Methods: Based on a priori hypotheses about the effects of latitude, environmental and biological factors on species' range, we calculated mid-latitudes of species ranges, and collected data on environmental factors (altitude, temperature, precipitation, size and number of ecoregions occupied) and biological traits (body size, fecundity, habitat breadth and species age) to construct multivariate models of snake range size. We used a recently published dated consensus phylogeny to determine minimum adequate models of range size using phylogenetic generalized least squares models and establish correlations between range size and time since species' description. Results: Range size increased significantly with latitude, consistent with Rapoport's rule, especially across mid- and high latitudes in the Northern Hemisphere. Habitat breadth, body size and altitudinal range had a significant positive effect on range size, with minor negative effects on range size from mean altitude and reproductive output. Biological variables explained more variation in range size than environmental variables. Species' range size had a significant effect on species' description, with larger-ranged species having been described earlier. Main conclusions: Prediction of range size in lesser-known species such as snakes relies on a suite of factors. Species with restricted habitat breadth, small body size and at high altitudes generally have smaller ranges, and are thus likely to have higher extinction risk. [Correction added on 10 October 2017, after first online publication: The preceding sentence has been rephrased for clarity.] Our work illustrates that it is these species we are likely to under-report in extinction risk assessments.