Testing the Giles hypothesis using geometric morphometrics.

Abstract

The Giles hypothesis posits that differences in the cranial morphology of Pan troglodytes and Gorilla gorilla are largely the result of allometric scaling. However, previous support for the Giles hypothesis was based on bivariate plots of linear measurements. This investigation uses geometric morphometric methods to retest this hypothesis and its prediction that extending the ontogenetic trajectory of a chimpanzee would produce an adult gorilla-like cranial morphology. Forty-three 3D cranial landmarks were collected from an ontogenetic series of 76 Pan troglodytes and 58 Gorilla gorilla specimens. Ontogenetic trajectories of cranial shape change were computed via multivariate regression of Procrustes aligned coordinates against LnCS (size vector) and molar eruption stage (developmental vector). These two vectors were then used in developmental simulations to extend the ontogenetic trajectories of adult chimpanzees. Allometric trajectories of chimpanzees and gorillas were also directly compared using Procrustes ANOVA. Pan and Gorilla significantly differ in their allometric trajectories, and none of the Pan developmental simulations resembled actual adult gorillas. Additionally, the more the Pan developmental vector was extended, the more morphologically distinct the simulations became from actual adult gorillas. Taken together, these results do not support the Giles hypothesis that allometric scaling is primarily responsible for observed morphological differences between chimpanzee and gorilla crania. This investigation demonstrates that neither "growing" a chimpanzee to the size of a gorilla, nor extending a chimpanzee's developmental shape trajectory will result in an adult gorilla-like cranial morphology as they differ in their patterns of allometry.