From correlation to causation: The rarest drives biodiversity gradients

Background/Question/Methods

The geographical variation in the distribution of species and their abundances has motivated ecologists to look for first-order commonalities to understand the processes giving rise to such patterns.  This is particularly true for biodiversity gradients.  Nearly a half-century ago, Preston developed a mathematical model based on the canonical log normal distribution of species abundances to explain why some regions have more species compared to others.  Preston’s model predicted that species richness depends on individual density within the community as well as the number of individuals of the rarest species.  Subsequent studies focused almost entirely on geographical variation in the total number of individuals, but that proved to explain relatively little of the geographical variation in richness.  Here, we use Preston’s model to test if the mechanism underlying geographic gradients of species richness might involve geographic variation in the number of individuals of the rarest species.  We used the relationship between temperature, precipitation, and the number of individuals of the rarest tree species from 11 altitudinal sites to predict tree species richness across the world. 

Results/Conclusions

We were able to predict from 66% - 83% of the variation in species richness across the world from the 11-altitudinal transect sites. Our results support the hypothesis that gradients in diversity depend on gradients of the number of individuals of the rarest species, which depends on gradients of temperature and precipitation. Moreover, the ability to use data from an altitudinal transect to accurately predict species richness across different latitudes suggests that communities are assembled via similar processes independent of their geological and evolutionary histories.