Variation in anthropoid vertebral formulae: implications for homology and homoplasy in hominoid evolution

Variation in vertebral formulae within and among hominoid species has complicated our understanding of hominoid vertebral evolution. Here, variation is quantified using diversity and similarity indices derived from population genetics. These indices allow for testing models of hominoid vertebral evolution that call for disparate amounts of homoplasy, and by inference, different patterns of evolution. Results are interpreted in light of "short-backed" (J Exp Zool (Mol Dev Evol) 302B:241-267) and "long-backed" (J Exp Zool (Mol Dev Evol) 314B:123-134) ancestries proposed in different models of hominin vertebral evolution. Under the long-back model, we should expect reduced variation in vertebral formulae associated with adaptively driven homoplasy (independently and repeatedly reduced lumbar regions) and the relatively strong directional selection presumably associated with it, especially in closely related taxa that diverged relatively recently (e.g., Pan troglodytes and Pan paniscus). Instead, high amounts of intraspecific variation are observed among all hominoids except humans and eastern gorillas, taxa that have likely experienced strong stabilizing selection on vertebral formulae associated with locomotor and habitat specializations. Furthermore, analyses of interspecific similarity support an evolutionary scenario in which the vertebral formulae observed in western gorillas and chimpanzees represent a reasonable approximation of the ancestral condition for great apes and humans, from which eastern gorillas, humans, and bonobos derived their unique vertebral profiles. Therefore, these results support the short-back model and are compatible with a scenario of homology of reduced lumbar regions in hominoid primates. Fossil hominin vertebral columns are discussed and shown to support, rather than contradict, the short-back model.