Using geometric morphometric visualizations of directional selection gradients to investigate morphological differentiation

Abstract

Researchers studying extant and extinct taxa are often interested in identifying the evolutionary processes that have lead to the morphological differences among the taxa. Ideally, one could distinguish the influences of neutral evolutionary processes (genetic drift, mutation) from natural selection, and in situations for which selection is implicated, identify the targets of selection. The directional selection gradient is an effective tool for investigating evolutionary process, because it can relate form (size and shape) differences between taxa to the variation and covariation found within taxa. However, although most modern morphometric analyses use the tools of geometric morphometrics (GM) to analyze landmark data, to date, selection gradients have mainly been calculated from linear measurements. To address this methodological gap, here we present a GM approach for visualizing and comparing between-taxon selection gradients with each other, associated difference vectors, and "selection" gradients from neutral simulations. To exemplify our approach, we use a dataset of 347 three-dimensional landmarks and semilandmarks recorded on the crania of 260 primate specimens (112 humans, 67 common chimpanzees, 36 bonobos, 45 gorillas). Results on this example dataset show how incorporating geometric information can provide important insights into the evolution of the human braincase, and serve to demonstrate the utility of our approach for understanding morphological evolution.

Evolvability and craniofacial diversification in genus Homo

“…Divergence vectors were calculated as the difference between an ancestor and its immediate descendant (either another internal node or a tip of the tree). These vectors capture the total shape changes accumulated along that branch over multiple generations (Weaver and Gunz ). The divergence vector on the branch leading to H. sapiens in Figure was based on four early modern human fossils rather than the more recent samples used to estimate P .…”

“…The second approach involves calculating selection gradients along branches and comparing them to neutral “selection” gradients calculated from neutral vectors simulated from a multivariate normal distribution with a mean of zero and a variance‐covariance matrix proportional to P (Weaver and Gunz ). The P matrix was subjected to eigenvalue extension based on the variance of the second derivatives as per Marroig et al.…”

Multivariate morphometrics, quantitative genetics, and neutral theory: Developing a “modern synthesis” for primate evolutionary morphology

“…35 This suggests that dissociating the human cranium into individual linear measurements may not be the most accurate means of estimating patterns of additive genetic variance relative to phenotypic variance, due to the effects of morphological integration. 36,37 As noted recently by Weaver and Gunz, 21 there is also a disconnect between the GM approach adopted by most evolutionary morphologists and the current methods for comparing directional selection gradients or detecting deviations from neutrality, which tend to be based on individual linear measurements. [12][13][14]18,19,38 Hence, new methods that combine the power of GM methods for statistical shape comparison with an analytical approach based on quantitative genetic principles such as that developed by Weaver and Gunz 21 will be important for the future development of evolutionary morphology.…”

“…[1][2][3][4] Utilization of this general framework has provided novel insights into the evolutionary history of the primate postcranium, [5][6][7][8][9][10][11] and into the evolution of cranial morphology in extinct and extant primate taxa. [12][13][14][15][16][17][18][19][20][21][22] Key to the success of this approach are three major intertwined conceptual and analytical frameworks: multivariate morphometrics, quantitative genetics, and neutral evolutionary theory (Figure 1). At a basic level, morphometrics is the methodological framework through which we quantify morphology and, thereby, delineate quantitative traits; quantitative genetics provides the framework for understanding the inheritance of complex traits and in turn, how evolutionary processes, such as drift and selection, act on morphology; and neutral theory provides a hierarchical hypothesis-testing framework for understanding the evolutionary divergence of populations.…”

Using geometric morphometrics for the genetics analysis of shape and size of lactation curves in Israeli first-parity Holstein cattle

“…Geometric morphometrics methods are powerful tools that have been developed in biology to quantify shape variation and its covariation with biotic and abiotic variables (Webster and Sheets, 2010); GMM has been successfully applied in medicine and engineering (Viscosi and Cardini, 2011; Cooke and Terhune, 2015). Individual shapes are represented by a collection of Cartesian coordinates of anatomical landmarks or semilandmarks on corresponding outlines and surfaces (Gunz and Mitteroecker, 2013;Weaver and Gunz, 2018). Geometric morphometrics methods analysis separates the morphological variation of biological forms (e.g., human skull, insect wing, or leaf) into components of size and shape, and it has been developed in the context of quantitative genetic studies of different species' shapes (Klingenberg and Leamy, 2001;Monteiro et al, 2002;Klingenberg et al, 2010).…”