Patterns of range size evolution are important for developing an evolutionary biogeographical theory and supporting conservation actions. Determining those patterns is however hampered by multiple factors acting on range size and by our uncertainty with regard to species' phylogenetic relationships and geographic distribution. In addition, given the diversity of analytic approaches existing in the literature, there are concerns regarding whether different methods might lead to different patterns. Here, we addressed these uncertainties in order to test for correlations between the evolution of morphology (body size and shape) and range size. We studied lizards of the family Gymnophthalmidae, representing changes of an order of magnitude in body size and also two independent lineages with limb-reduced (snake-like) morphs. We used phylogenetic multivariate methods (pPCA) to control for allometric effects of body size over shape. Then, we performed multiple regressions under three approaches: ‘naive’ least squares, Bayesian comparative methods that accounted for uncertainties in trait optimization, tree topology and branch length, and a novel permutative procedure based on phylogenetic generalized least squares (pGLS) to assess the robustness of our relationships to uncertainties in the range sizes of the species studied. All approaches led to the same answer: only body size was related to range size. While bigger gymnophthalmids tend to have relatively large ranges, small species can have either small or large ones. Our results’ robustness to strong uncertainties in both phylogenetic relationships and range sizes shows there are opportunities for overcoming those problems and produce reliable patterns of range size evolution. However, interpretation of the processes driving patterns of range size evolution will still require advances in phylogenetic and taxonomic knowledge. We discuss the application of morphology–range size relationships to conservation planning in the light of existing uncertainties in the geographic knowledge of our studied species group and workarounds for data availability.
|Number of pages||10|
|Publication status||Published - Jun 2017|