PHYLOGENETIC HYPOTHESES UNDER THE ASSUMPTION OF NEUTRAL QUANTITATIVE-GENETIC VARIATION

There are many situations in which the only available characters for reconstructing phylogenies are morphological. Those traits that are subject only to the forces of mutation and random genetic drift can be used to obtain unbiased estimates of phylogenetic relationships. However, the accurate recovery of a phylogeny from information on neutral characters requires the procurement of data for a large number of independent traits, individuals, and populations. Phylogenetic trees fit to data from more than five species will almost always contain topological errors, even with very large data sets. The population-genetic consequences of the neutral model are reviewed, and some statistical methods for testing whether the diversification of a phylogeny is compatible with such a model are outlined. The theory is then applied to a very large data set on cranial morphology in modern man. The results are consistent with the hypothesis that interracial differences in human skull dimensions are a simple consequence of random drift and mutation.