Computational protocol: Phenotypic Divergence among West European Populations of Reed Bunting Emberiza schoeniclus: The Effects of Migratory and Foraging Behaviours

Similar protocols

Protocol publication

[…] A photograph of the bill in profile was taken from 208 individuals of all populations/subspecies, and subjected to geometric morphometric analysis, a powerful method with few a priori assumptions to explicitly define shape –. This method has recently been applied to a growing number of animal groups, including in a few bird studies that compare bill shapes –. Prior to analysis, photographs were edited in Adobe Photoshop CS4 (for details see ), and then all geometric morphometric analyses were conducted in software of the tps series . A tps file was built from images using tpsUtil , and used in tpsDig , where seven landmarks and eight semi-landmarks were digitized following Foster et al. . The semi-landmarks were placed by reference to a standardized grid superimposed onto each image (cf. and ). Files containing links (between landmarks) and sliders (for each semi-landmark) were built in tpsUtil and an image list was obtained. Using the tpsSmall software , we confirmed that shape variation between the specimens was sufficiently small and therefore the distribution of points in the shape space can be represented satisfactorily by their distribution in the tangent space. We then applied a Generalized orthogonal least-squares Procrustes Analysis (GPA) , using tpsRelw , in order to standardize the size and to translate and rotate the configurations of landmark coordinates, therefore obtaining a consensus configuration. We computed partial and relative warps and extracted relative warp scores with a α = 0, using the tpsRelw software . tpsRelw output files were saved in NTS format, converted to csv using tpsUtil, and merged with the image list in Microsoft Excel. Because of logistical constraints that prevented inclusion of a standardized scale in each image, allometry was evaluated by reference to a Principal Component based on univariate measurements (see below). […]

Pipeline specifications