Species distribution modelling as a macroecological tool: a case study using New World amphibians

Although species distribution modelling (SDM) is widely accepted among the scientific community and is increasingly used in ecology, conservation biology and biogeography, methodological limitations generate potential problems for its application in macroecology. Using amphibian species richness in North and South America, we compare species richness patterns derived from SDM maps and ‘expert’ maps to evaluate if: 1) richness patterns derived from SDM are biased toward climate‐based explanations for diversity when compared to expert maps, since SDM methods are typically based on climatic variables; and 2) SDM is a reliable tool for generating richness maps in hyperrich regions where point occurrence data are limited for many species. We found that although three widely used SDM methods overestimated amphibian species richness in grid cells when compared to expert richness maps in both North and South America due to systematic overestimation of range sizes, diversity gradients were reasonably robust at broad scales. Further, climatic variables statistically explained patterns of richness at similar levels among the different richness sources, although climatic relationships were stronger in the much better known North America than in South America. We conclude that in the face of the high deforestation rates coupled with incomplete data on species distributions, especially in the tropics, SDM represents a useful macroecological tool for investigating broad‐scale richness patterns and the dynamics between species richness and climate.