Evolutionary constraints and information entropy in ecology

Ecological communities can assemble in many different ways. Given such mechanistic heterogeneity, probabilistic constraints could be more important in determining general ecological patterns such as the species abundance distribution. The principle of maximum information entropy (MaxEnt) can show us the most probable macrostate of a community given such probabilistic constraints. MaxEnt works in statistical mechanics and often works surprisingly well in ecology. But could the uniquely biological process of evolution lead ecology to systematically deviate from MaxEnt? Using rapidly evolving communities across the Hawaiian Islands I will explore when and why evolution adds extra constraint to ecology.