Department of Anthropology, Temple University
March 26, 2015 , Archview Ballroom
Morphological characters such as sexual size dimorphism (SSD), residual testes weight (RTW) and sperm midpiece volume (MPV) differ among species and between groups of species with similar mating systems. However, because species-specific intensities of pre- and post-copulatory, male-male competition are difficult to quantify and mating system categories are broad, hypotheses of sexual selection are difficult to test. We used principal components analysis (PCA) to reduce the dimensions of a morphological dataset and create a 2-dimensional morphospace. We analyzed a correlation matrix of data on RTW, MPV and three measures of SSD for 27 species. Species’ PC scores were plotted on a bivariate graph. The first two PCs accounted for 80.13% of variance in the dataset (48.48% and 31.65%, respectively). PC1 correlated with SSD traits, and PC2 with MPV and RTW. We calculated the area of the convex hull of species’ positions in morphospace and asked whether species clustered more tightly by mating system or phylogeny. We then compared mean PC1 and mean PC2 values between mating systems using ANOVA. Species clustered more closely when grouped by mating system than when grouped by phylogeny and mean PC values differed between mating systems (p < 0.005). In post-hoc analysis, we separated polygyandrous species above the fourth quintile of PC1 from those below, which tightened clustering by both measures. These results suggest reproductive skew may be highest in strongly dimorphic, polygynandrous primates. Further, variation in the intensity of precopulatory competition is an important driver of phenotypic diversity in polygynandrous primates.
Andrew Van Horn is supported by a Temple University Fellowship