Three-dimensional geometric morphometric analysis of the skull of Protoceratops andrewsi supports a socio-sexual signalling role for the ceratopsian frill

Abstract

Socio-sexual selection is predicted to be an important driver of evolution, influencing speciation, extinction and adaptation. The fossil record provides a means of testing these predictions, but detecting its signature from morphological data alone is difficult. There are, nonetheless, some specific patterns of growth and variation which are expected of traits under socio-sexual selection. The distinctive parietal-squamosal frill of ceratopsian dinosaurs has previously been suggested as a socio-sexual display trait, but evidence for this has been limited. Here, we perform a whole-skull shape analysis of an unprecedentedly large sample of specimens of Protoceratops andrewsi using a high-density landmark-based geometric morphometric approach to test four predictions regarding a potential socio-sexual signalling role for the frill. Three predictions—low integration with the rest of the skull, significantly higher rate of change in size and shape during ontogeny, and higher morphological variance than other skull regions—are supported. One prediction, sexual dimorphism in shape, is not supported, suggesting that sexual differences in P. andrewsi are likely to be small. Together, these findings are consistent with mutual mate choice or selection for signalling quality in more general social interactions, and support the hypothesis that the frill functioned as a socio-sexual signal in ceratopsian dinosaurs.

Sexual selection leads to positive allometry but not sexual dimorphism in the expression of horn shape in the blue wildebeest, Connochaetes taurinus

“…Both sexes follow identical patterns of growth and variation across the skull, and are separated only by size. These results reflect attempts to recover dimorphism in extinct taxa where sex is not known a priori (Mallon, 2017;Knapp et al, 2021). Reproductive biology, life history and intensity of sexual selection may all affect the diversity, magnitude of sexual dimorphism and relative growth of secondary sexual traits, but the basic effects of sexual selection on morphology are likely to remain detectable to some extent.…”

“…2). ML approaches are known to preferentially support highly-paramaterised datasets, so we followed previous studies (Bardua et al 2019a, Knapp et al, 2021) and examined the between-element correlation values to determine if any elements could reasonably be merged into a single module. Using a correlation value threshold of within 0.1 of either connected element we merged three pairs of elements (frontal and parietal, occipital and sphenoid, and premaxilla and maxilla), resulting in a 9-module configuration.…”

“…All additional analyses performed on subsets of the dataset are presented in the supplementary data. Landmarks were reflected across the sagittal plane to reduce lateral inaccuracies in aligning the bilaterally symmetrical skull (Knapp et al, 2021). Specimens were then aligned using a generalised Procrustes alignment (GPA) in the R package geomorph (Adams and Otárola-Castillo, 2013), and the reflected landmarks were removed, leaving the original, Procrustes-aligned right-side landmarks for analysis.…”

Complex macroevolutionary dynamics underly the evolution of the crocodyliform skull

“…This simplistic portrayal of crocodyliform evolution is partly driven by the data used to quantify their morphology. Whereas three-dimensional morphometrics has long been the standard for comparative analyses of vertebrate cranial evolution [ 35 – 39 ], studies of crocodyliforms still rely overwhelmingly on two-dimensional geometric morphometric data. Even studies that celebrate their extinct diversity generally measure it using dorsal view images that exclude much of the complexity of the cranium.…”

“…R. Soc. B 288: 20210919 comparative analyses of vertebrate cranial evolution [35][36][37][38][39], studies of crocodyliforms still rely overwhelmingly on twodimensional geometric morphometric data. Even studies that celebrate their extinct diversity generally measure it using dorsal view images that exclude much of the complexity of the cranium.…”