Anthropology and Illinois Informatics Institute, University of Illinois
Friday 3:30-3:45, Parlors
The Cheverud Conjecture (CC) states that genetic and phenotypic covariance matrices (G and P, respectively) are very nearly proportional for morphological characteristics. If the CC holds, estimates of selection gradients, estimates of the degree of divergence among groups, and tests of the neutral model of phenotypic evolution derived using P will be proportional to those calculated using G. This would be convenient for those who wish to use evolutionary quantitative genetic theory and method to test hypotheses about evolutionary process in humans and other primates as it is extraordinarily difficult to estimate G in most species of primates. Using published estimates of G and P, I test a series of hypotheses meant to judge the soundness of the CC and the sensitivity of evolutionary quantitative genetic tests to violations of the proportionality assumption. Almost without exception, the CC holds, and G and P appear to be proportional when the vagaries of sampling are taken into account. Uncertainty in estimates of P due to sampling and the undue effects of the badly estimated dimensions of P and G are much more pressing issues for the study of the effects of selection and drift on the evolution of morphology. Estimates of P are adequate to answer some, but not all, questions about the evolution of morphology in humans and other primates. Quantitative genetic investigations into how P and G evolve are critical for understanding the causes of the differences among primates in their patterns of modularity and integration.
I thank the National Science Foundation (BCS 0962903), the Center for Advanced Study, the University of Illinois Research Board, Illinois ATLAS, and the Illinois Informatics Institute for their support.