Patterns of Species Rarity as a Driving Mechanism for Species Richness Gradients

Abstract

Broad scale geographic variation in species diversity correlates with environmental variables in most taxa, but a mechanistic understanding of this relationship has remained elusive. More than a half-century ago, F.W. Preston observed that the number of individuals per species in species assemblages is log-normally distributed (with two parameters: the total number of individuals, I, and the number of individuals of the rarest species, m). Here, we show that ϕ, a proxy for m, is correlated with environmental variables in several datasets of trees, birds, fish, and invertebrates. Moreover, variation in species richness is more strongly related to this measure of rarity than to environment. In all the datasets we examined, structural equation models are consistent with the hypothesis that environmental variables affect species richness principally by affecting rarity, which in turn affects richness. We propose that geographic variation in the ability of species to persist at low densities provides a possible unifying explanation for global gradients of species richness. Our findings may have important implications regarding Earth’s biodiversity, highlighting the rarest species as those most at-risk but also important indicators for the ongoing consequences of climate change.