Evolutionary rate explanations for latitudinal diversity gradients predict faster speciation and diversification rates in richer, older, and more stable tropical regions (climatic stability hypothesis). Numerous modern lineages have emerged in high latitudes, however, suggesting that climatic oscillations can drive population divergence, at least among extratropical species (glacial refugia hypothesis). This conflicting evidence suggests that geographical patterns of evolutionary rates are more complicated than previously thought.

Here, we reconstructed the complex evolutionary dynamics of a comprehensive dataset of modern mammals, both terrestrial and marine. We performed global and regional regression analyses to investigate how climatic instability could have indirectly influenced contemporary diversity gradients through its effects on evolutionary rates. In particular, we explored global and regional patterns of the relationships between species richness and assemblage-level evolutionary rates and between evolutionary rates and climatic instability.

We found an inverse relationship between evolutionary rates and species richness, especially in the terrestrial domain. Additionally, climatic instability was strongly associated with the highest evolutionary rates at high terrestrial latitudes, supporting the glacial refugia hypothesis there. At low latitudes, evolutionary rates were unrelated to climatic stability.

The inverse relationship between evolutionary rates and the modern latitudinal diversity gradient casts doubt on the idea that higher evolutionary rates in the tropics underlie the current diversity patterns of modern mammals. Alternatively, the longer time spans for diversity to accumulate in the older and more stable tropics (and not high diversification rates) may explain the latitudinal diversity gradient.

After computing evolutionary rates for each species using BAMM, we calculated the mean speciation rate (MS) and mean net diversification rate (MDN) of all species of mammals per grid cell (Table_MS_MDN.csv). Grid cells at a 1° × 1° resolution

We estimated historical climatic instability (CI) as the difference between the mean annual temperature of the last glacial maximum and the present mean annual temperature in each cell, for both landmasses and the sea surface (Table_CI.csv). This dataset excluded Antarctic Ocean information (above latitude 70°S). Grid cells at a 1° × 1° resolution.

All coastal cells as well as cells with less than two species were removed of datafiles.