Post Doc

Unit for Theoretical Biology, Department of Evolutionary Biology
University of Vienna

Keywords:

partial warps; geometric morphometrics; spatial scale; prWarp; phylogenetic signal; adaptive signal; shape decomposition

Abstract

Complex shape patterns can be quantified and analysed using geometric morphometrics (GMM). However, classical GMM methods have shortcomings, for example when the landmark distribution is very heterogenous, or to separate the biologically relevant signal from the noise. It is possible to bypass some of these issues by decomposing shape variation into its large-scale and small-scale components. Two different approaches are possible: (1) a biologically-informed decomposition of the overall shape variation into outline and residual shape components; and (2) a mathematical decomposition of the non-affine shape variation, i.e. the shape variation excluding linear scaling and shearing, into partial deformations for increasing spatial scales, the partial warps. Using the example of the midsagittal skull morphology in primates, I will show how these approaches can provide some insights into morphological integration and growth processes, and help to distinguish the adaptive signal from the phylogenetic signal. These methods have multiple other applications in fields as diverse as ecology, palaeontology, evolutionary biology and developmental biology. All the relevant functions and tutorials are available as part of the R package prWarp, that I developed for this purpose.