Levels of the genealogical hierarchy and the problem of hominoid phylogeny.

Molecular data are widely used to reconstruct phylogenetic relationships among species, and these phylogenies are often used as the basis for inferences about the history of evolutionary change in other nonmolecular characters. This approach is an appropriate and powerful one in many circumstances. But when several lineages diverge over a relatively short period of time, the assumption that a molecular (gene) tree will always be a valid basis for such inferences may not hold. Empirical evidence from humans, nonhuman primates, and other mammals indicates that the relationships among molecular divergence, morphological differentiation, and the origin of reproductive isolation between diverging lineages are complex. The simple dichotomously branching trees that result from molecular systematic studies of Homo, Gorilla, and Pan may be a misleading basis for reconstructions of evolutionary change in nonmolecular characters.