Temperature predicts the maximum tree-species richness and water and frost shape the residual variation

The kinetic hypothesis of biodiversity proposes that temperature is the main driver of variation in species richness, given its exponential effect on biological activity and, potentially, on rates of diversification. However, limited support for this hypothesis has been found to date. I tested the fit of this model on the variation of tree-species richness along a continuous latitudinal gradient in the Americas. I found that the kinetic hypothesis accurately predicts the upper bound of the relationship between the inverse of mean annual temperature (1/\textit{k}K) and the natural logarithm of species richness, at a broad scale. In addition, I found that water availability and the number of days with freezing temperatures organize a part of the residual variation of the upper bound model. The finding of the model fitting on the upper bound rather than on the mean values suggest that the kinetic hypothesis is modeling the variation of the potential maximum species richness per unit of temperature. Likewise, the distribution of the residuals of the upper bound model in function of the number of days with freezing temperatures suggest the importance of environmental thresholds rather than gradual variation driving the observable variation in species richness.