A 3D morphometric approach to skull integration and modularity in Pan troglodytes and Homo sapiens

Cranial integration and modularity studies aim to analyse the covariation and interaction of the different bony structures that make up the cranium. The craniofacial morphology of Hominidae is largely variable and presents a series of integration patterns that may be due to both intraspecific and interspecific factors [1]. One of the keys to understanding the evolutionary mechanisms of our genus is to identify and quantify which integration patterns are shared and whether these can be applied to fossil species [2]. In the past decades, thanks to the application of new techniques and digital tools (mainly 3D), it has been possible to approach this problem from new perspectives. However, delving into studies about skull integration and modularity is crucial because of the complex interaction of individual bones in the integration of the whole system [3]. The aim of this work is to quantify the covariation patterns of the different cranial bones and modules of the two phylogenetically closest extant Hominidae species: Homo sapiens and Pan troglodytes. The main goal is to detect shared or species-specific integration patterns that can later be applied to extinct hominin species. For this purpose, we conducted a Procrustes-based 3DGMM study on the cranial surface of a sex-balanced sample of 40 adult individuals of both species. We have defined 6 regions -viscero-cranium, frontal, parietal, temporal, occipital and sphenoid - using 65 landmarks that collect homologous osteometric points and 367 curves, as well as 949 surface sliding semi-landmarks that detail the shape of the surface not measurable by the landmarks. To establish the integration patterns of both species and observe which of them are common and which are species-specific, we used the two-block partial least squares (PLS) method pooled by species [4]. This enabled us to obtain the RV coefficient between each bony structure and to create a covariation matrix between each 6 regions. To explore whether the observed patterns are shared or species-specific, we explored the slope of the PLS1 scores through a Bootstrap estimation of 95% confidence interval of the slope in each species [5]. The results obtained show that each of the 6 bone structures show significant covariation. The RVs range from 0.3-0.8 and each one shows a p-value of <0.0001, suggesting that the entire skull works like an integral functional system. The 30 modules tested also present significant covariations with RVs ranging between 0.3-0.7 and p-values of <0.0001. It should be noted that the full skull structures module configuration that provided the highest integration index, as suggested by previous studies, is composed by the (i) viscero-cranium and occipital block, and (ii) parietal, temporal and frontal block; with the sphenoid playing an important role in both of them. Even if this bone is placed in one block or the other, his location did not substantially alter the covariation index, but it does have an impact on the slopes that suggest shared or species-specific patterns. This 2-block modular configuration covariation seems shared by both species, thus making these shared morphological patterns very useful to explore covariation patterns in fossil human species. [/su_box]