Detecting biological uniqueness using geometric morphometrics: an example case from the Vancouver Island marmot

The Vancouver Island marmot (M. vancouverensis) consists of a small insular population with an uncommon dark-brown pelage colour. This population provides an excellent example of how morphology and molecules can tell very different stories about the evolution of a group. Mitochondrial DNA indicates that the degree of divergence of M. vancouverensis is minimal and actually comparable or even smaller than among populations of continental species of marmots. Accurate geometric morphometric analyses on crania and mandibles from modern and subfossil samples, in contrast, convincingly demonstrate that the Vancouver Island marmot is one of the most distinctive populations among all living marmots. Thus, M. vancouverensis is not simply a dark variant of its sister species M. caligata but represents a highly morphologically derived population whose evolutionary significance is overlooked by molecular analyses. In this study, we show how geometric morphometrics can complement studies of genetic divergence by measuring the phenotypic dis- tinctiveness of a population. This was done by performing cross-validated discriminant analyses on mandibular shape in the Vancouver Island marmot and its closest Nearctic relatives. Thus, we found that using mandibles we can achieve in M. vancouverensis a classification accuracy as high as for ‘good species’ (i.e., species which are also significantly divergent in molecular analyses) and much higher than in populations within those same species. The approach used in this study is simple, effective and, at least in this specific case, fairly robust to sampling error in small samples. Thus, such morphological analysis seems promising as a tool complementary to molecular studies to aid detection of biological distinctiveness in natural populations. We suggest that its potential should be explored and its usefulness examined also in Palearctic marmots and other mammals. If successful, the ‘geometric morphometric approach to the study of biological diversity‘ would add a phenotypic dimension to the characterization of the evolutionary significance of populations. In turn, this would allow a more inclusive understanding of the complexity of adaptive variation with potential implications for conservation biology.