Modeling of Rate Heterogeneity in Datasets Compiled for Use With Parsimony

Abstract

In recent years, there has been an increased interest in modeling morphological traits using Bayesian methods. Much of the work associated with modeling these characters has focused on the substitution or evolutionary model employed in the analysis. However, there are many other assumptions that researchers make in the modeling process that are consequential to estimated phylogenetic trees. One of these is how among-character rate variation (ACRV) is parameterized. In molecular data, a discretized gamma distribution is often used to allow different characters to have different rates of evolution. Morphological data are collected in ways that fundamentally differ from molecular data. In this paper, we appraise the use of standard parameters for ACRV and provide recommendations to researchers who work with morphological data in a Bayesian framework.