Fig. 1: Covariates of genetic diversity in terrestrial mammals at two spatial scales.

Independent contribution (IC; hierarchical partitioning) and the sum of Akaike weights (SAW; multimodel inference) of biodiversity dimensions and climate variables in explaining the global distribution of genetic diversity (average number of mutations per base pair and across taxa) for cytb and co1. Spatial scales: a grid cells (cytb: n = 185; co1: n = 76), and b zoogeographic regions (cytb: n = 34; co1: n = 30). Inferences at the grid cell scale are based only on data-rich cells (see Fig. 2 and Methods). Plus (+) and minus (−) signs after each explanatory variable indicate positive and negative significant association respectively between the explanatory variable and genetic diversity. Note that all explanatory variables for co1 at the zoogeographic scale were insignificant, reflecting the highly unbalanced distribution of available co1 sequences across regions (Supplementary Figs. 1–3).