Integrating ecological niche modeling and rates of evolution to model geographic regions of mimetic color pattern selection

Abstract

AbstractGeographic variation in natural selection derived from biotic sources is an important driver of trait evolution. The evolution of Müllerian mimicry is governed by dual biotic forces of frequency-dependent predator selection and densities of prey populations consisting of conspecifics or congeners. Difficulties in quantifying these biotic forces can lead to difficulties in delimiting and studying phenomena such as mimicry evolution. We explore the spatial distribution of morphotypes and identify areas of high mimetic selection using a novel combination of methods to generate maps of mimetic phenotype prevalence in Ranitomeya poison frogs, a group of frogs characterized by great phenotypic variation and multiple putative Müllerian mimic pairs. We categorized representative populations of all species into four major recurring color patterns observed in Ranitomeya: striped, spotted, redhead, and banded morphs. We calculated rates of phenotypic evolution for each of the 4 morphs separately and generated ecological niche models (ENMs) for all species. We then split our species-level ENMs on the basis of intraspecific variation in color pattern categorization, and weighted ENM layers by relative evolutionary rate to produce mimicry maps. Our phenotypic evolutionary rate analyses identified multiple significant shifts in rates of evolution for the spotted, redhead, and banded phenotypes. Our mimicry maps successfully identify all suspected and known areas of Müllerian mimicry selection in Ranitomeya from the literature and show geographic areas with a gradient of suitability for Müllerian mimicry surrounding mimic hotspots. This approach offers an effective hypothesis generation method for studying traits that are tied to geography by explicitly connecting evolutionary patterns of traits to trends in their geographic distribution, particularly in situations where there are unknowns about drivers of trait evolution.