Adaptive zones and the pinniped ankle: A three-dimensional quantitative analysis of carnivoran tarsal evolution

Abstract

Bones are functional. Stated so abruptly, this observation is a truism, but its significance depends on the context in which it is made. In an individual animal, bones support loads, resist muscular contractions, and facilitate bodily movements. Bone form both constrains, and is shaped by, force and motion. In an environmental context, a bone’s form is compatible with its owner’s size and habits and is, thus, related indirectly to habitat and environment, although any particular bone (or, more properly, musculoskeletal configuration) can cope in diverse environments, and any substrate can be traversed by animals with different skeletal forms. Form and function are inseparable at the level of joint movements (Bock and von Wahlert, 1965), but they are only loosely correlated at the level of ecology, specifically locomotor types and habitats. The coarseness of the correlation between form and ecology come from the temporal lag of phylogenetic adaptation and the many-to-many relationship between form and habitat. Even though ecophenotypic plasticity allows bones to be modified during an individual’s lifetime, bone form is largely heritable and evolutionary change requires generations of selective genetic and epigenetic reorganization (Cock, 1966; Grüneberg, 1967; Thorpe, 1981). In this paper, function and phylogeny were analyzed using a new geometric morphometric technique that quantitatively represents the entire three-dimensional surface of the bones. This method was used to associate variation in the two bones, including the size and curvature of occluding joint facets, with locomotor type, stance, number of digits, and body mass. Principal components analysis was used to describe the major axes of variation in the two bones, and multivariate analysis of variance was used to test functional categories for significance. Correlated transformations in the interlocking surfaces of the two bones were also explored using two-block partial least squares. Phylogenetic components of variation were assessed by mapping the three-dimensional shape of the bones onto a cladogram and projecting the results back into the principal component morphospace to visualize the patterns of homoplasy. Rates of morphological evolution in the several clades were calculated from the mapped shapes. Homoplasy was also quantitatively assessed by measuring the scaling coefficient between evolutionary divergence and time since common ancestry. The final aim of this paper was to develop criteria for assessing the whether functional adaptation is likely to confound phylogenetic signal in a dataset for the taxa being considered. A quantitative redefinition of Simpson’s adaptive zones was employed to assess the effect of adaptive convergence on phylogenetic divergence, and determine the circumstances in which associated homoplasy is likely to confound phylogeny reconstruction.