A framework for studying behavioral evolution by reconstructing ancestral repertoires

Abstract

Although extensive behavioral changes often exist between closely related animal species, our understanding of the genetic basis underlying the evolution of behavior has remained limited. Here, we propose a new framework to study behavioral evolution by computational estimation of ancestral behavioral repertoires. We measured the behaviors of individuals from six species of fruit flies using unsupervised techniques and identified suites of stereotyped movements exhibited by each species. We then fit a Generalized Linear Mixed Model to estimate the suites of behaviors exhibited by ancestral species, as well as the intra- and inter-species behavioral covariances. We found that much of intraspecific behavioral variation is explained by differences between individuals in the status of their behavioral hidden states, what might be called their "mood." Lastly, we propose a method to identify groups of behaviors that appear to have evolved together, illustrating how sets of behaviors, rather than individual behaviors, likely evolved. Our approach provides a new framework for identifying co-evolving behaviors and may provide new opportunities to study the genetic basis of behavioral evolution.