Energy and Large-Scale Patterns of Animal- and Plant-Species Richness

Many hypotheses have been proposed to explain the great variation among regions in species richness. These were tested by first examining patterns of species richness of birds, mammals, amphibians, and reptiles in 336 quadrats covering North America. These patterns were then compared with the regional variation of 21 descriptors of the environment suggested by the hypotheses. I found that, in the four vertebrate classes studied, 80%-93% of the variability in species richness could be statistically explained by a monotonically increasing function of a single variable: annual potential evapotranspiration (PET). In contrast, tree richness is more closely related to actual evapotranspiration (AET). Both AET and PET appear to be measures of available environmental energy. The relationships between tree and vertebrate richness are strikingly poor. Species richness in particular orders and families of the Vertebrata is also closely related to PET, but not always monotonically, often resembling a replacement series along an environmental gradient. The present results are consistent with the hypothesis that environmentally available energy limits regional species richness. However, my observations are not completely consistent with earlier species-energy theory. The energy-richness relationship appears to depend on scale, and it is affected differently by variations in area and in areal energy flux.