A Geometric Morphometric Comparison of Pelvic and Cranial Sexual Dimorphism Kaleigh C. Best, M.S., 1 Luis L. Cabo, M.S., 2 Heather M. Garvin, Ph.D. 2 1 Southern Illinois University, 2 Mercyhurst University Materials and Methods A sample of 113 os coxae and crania from U.S. Blacks from the Hamman-Todd Osteological Collection were analyzed (Table 1). Table 1: Sample distribution. Forty-two landmarks from the cranium and 12 landmarks from the os coxae (Figure 1) were digitized using a MicroScribe ® digitizer. Landmarks were chosen in order to capture the overall shape and size of the skeletal elements. For the crania, landmarks were collected following definitions provided in the FORDISC 3.1 help file (see Jantz and Ousley 11 for more information). Landmarks for the ossa coxae were collected using guidelines from various authors (see Table 2) 14, 15 . Following a Procrustes superimposition, principal components (PCs) and cross-validated discriminant function analyses (DFA) were used to assess and compare the degree of sexual shape and form dimorphism present in both skeletal regions. Univariate analyses were performed to evaluate which specific shape changes were contributing the most to the sex differences and were visually interpreted using MorphoJ 12 and Morphologika 13 software. Centroid size was used to assess sexual size dimorphism. Table 2: Landmarks of the Ossa Coxae Sex Sample Before Outlier Detection Sample After Outlier Removal Male 66 55 Female 65 58 Total 131 113 Landmark Authors Anterior superior iliac spine Bytheway and Ross 2010 Anterior inferior iliac spine Bytheway and Ross 2010 Ischial spine Bytheway and Ross 2010 Posterior iliac spine Bytheway and Ross 2010 Apex inside the greater sciatic notch Bytheway and Ross 2010 *Acetabular point *Bytheway and Ross 2010 Most superior point on the symphyseal face Bytheway and Ross 2010 Most inferior point on the symphyseal face Bytheway and Ross 2010 Maximum os coxae height point 1 Klales et al. 2009 Maximum os coxae height point 2 Klales et al. 2009 Most posterior point of the superior border of the sciatic notch Novel landmark Midpoint of the ischiopubic ramus Novel landmark Figure 1: Landmarks of the cranium and os coxae. *Modified in this study. • If a preauricular groove was present, the post sciatic notch point was taken on the most ventrally projecting border of the groove. • If osteophytes were present, the point was taken on the adjacent non-affected surface. Introduction In humans, the os coxae and the cranium are commonly referred to as the most sexually dimorphic regions of the skeleton 1 , and thus are often used to estimate the sex of individuals in a variety of physical anthropology subfields, including paleoanthropology, bioarchaeology, and forensic anthropology. These skeletal differences between males and females have been attributed to a variety of mechanisms, including hormones 2-4 , functional constraints and biomechanical requirements for child-birth 5-9 , and sexual selection and mating preferences 9 . The direct influence each of these variables has on skeletal morphologies, however, remains unclear. If they are under the same influences, we might expect similar levels of sexual dimorphism in these two regions when evaluated in a single sample. Research Goal • To use landmark data and geometric morphometric analyses to compare sexual size and shape dimorphism in the os coxae and cranium in a single sample. These analyses will help determine: 1) which skeletal region is more sexually dimorphic (i.e., if a forensic anthropologist is faced with contradictory sex estimates from the os coxae and cranium, which estimate should they rely on); 2) whether sexual dimorphism in these skeletal regions is primarily the result of shape or size dimorphism, or a combination of both; 3) which specific regions of the os coxae and cranium contribute most to observed sex differences. Results O s Coxae Shape Of the 29 PCs, only PCs 1-6 were examined, as each explained at least 5% of the shape variation, cumulatively representing 64.2% of the variation. MANOVA results also indicate that there is a significant effect of sex on overall os coxae shape (p < 0.001). ANOVA results indicate significant sex differences in PC1, PC2, and PC3 (p <0.001). Shape differences were analyzed in each of these PCs (Figure 2). DFA results are in Table 3. Discussion and Conclusions • Sexual dimorphism was observed in both the cranium and the os coxae . • The results of the shape analyses confirm that the traditional non-metric traits are indeed the most sexually dimorphic aspects of the os coxae and cranium, and thus these traits can be used to differentiate between the sexes with relatively high accuracy. • Based on the sample, the ossa coxae appear to be a more reliable indicator for sex estimation than the crania. • None of the first four cranial PCs displayed significant sex differences, indicating that most cranial variation, as captured by the landmarks in this study, is not related to sexual differences. • Including size in the analysis (i.e., form) did not significantly increase classification accuracy in either the cranium or the os coxae, which may be explained by size obscuring shape differences. • Possible explanations for higher rates of sexual dimorphism in the ossa coxae include: the functional constraints of childbirth on the pelvis and lack of functional constraints on the crania, the influence of sexual selection of traits on the cranium, the role of environment on morphology, differential effects of sexual steroids in both elements, or landmark selection in the study. Figure 2: Shape changes in PC1, 2, (left) and 3 (right) in the os coxae . Figure 3: Shape changes in PC5 in the cranium. Cranium Shape Out of 64 PCs extracted, only the first four account for circa 5% or more of the shape variation, cumulatively adding to 54.1% of the variance. MANOVA revealed sexual differences on overall cranial shape (p <0.001). ANOVA narrowed these differences to components PC5 (p<0.001), PC17 (p = 0.045), PC18 (p = 0.035), PC26 (p <0.001), and PC27 (p = 0.046). Figure 3 illustrates sexual shape differences in PC5, which accounts for the larger share (4.8%) of overall shape variation among the above. Table 3 displays DFA results. Element Wilks ƛ P value (p <) Predicted Group Total Classification Accuracy (%) Cross -validated Accuracy (%) M F Cranium 0.495 0.001 46 9 55 83.6 84.1 9 49 58 84.5 Os coxae 0.126 0.001 55 0 55 100 99.1 1 57 58 98.3 Centroid Size and Form (Shape + Size) Centroid size alone allowed assignment of sex with 70.8% (os coxae ) and 74.3% (cranium) accuracies (cross-validated). As expected, form (size + shape) rendered higher accuracies: 98.2% for the os coxae and 77.9% for the cranium) . Table 3: Shape Discriminant Function Analysis by Element Acknowledgements The authors are immensely grateful to the faculty, staff and graduate students in the Department of Applied Forensic Sciences at Mercyhurst University. Thanks also go to Lyman Jellema at the Cleveland Museum of Natural History for access to the Hamann-Todd Osteological Collection. References 1 Spradley MK, and Jantz RL. 2011. Sex Estimation in Forensic Anthropology: Skull versus Postcranial Elements J Forensic Sci. 56 (2): 289-296. 2 Compston JE. 2001. Sex Steroids and Bone. Physiological Reviews. 81 (1). 3 Frank GR. 2003. Role of Estrogen and Androgen in Pubertal Skeletal Physiology. Med Pediatric Oncolo. 41: 217-221. 4 Bilfeld MF, Dedouit F, Sans N, Rousseau H, Rouge D, and Telmon N. 2015. Ontogeny of Size and Shae Sexual Dimorphism in the Pubis: A Multislice Computed Tomography Study by Geometric Morphometry. J of Forensic Sci. doi: 10.1111/156-4029.12761. 5 Tague, RG. 1992. Sexual dimorphism in the human bony pelvis, with a consideration of the Neadnertal pelvis from Kebara cave, Israel. Am. J Phys Anth. 88(1): 1-21. 6 Bigoni L, Veleminska J, and Bruzek J. 2010. Three Dimensional geometric morphometric analysis of cranio-facial sexual dimorphism in a Central European sample of known sex. J of Comparative Human Biology. 61: 16-32. 7 Kolesova O and Vetra J. 2011. Sexual Dimorphism of Pelvic Morphology Variation in Live Humans. Papers on Anthropology XX. 209 -217. 8 Dunsworth HM, Warrener AG, Deacon T, Ellison PT, Pontzer H. 2012. Metabolic hypothesis for human altriciality. PNAS. 109 (38): 15212-15216. 9 Frayer DW and Wolpoff MH. 1985. Sexual Dimorphism. Annual Review of Anthropology 14:429-473. 10 Stull KE, Kenyhercz MW, and E L’Abbe. 2013. Non-metric Cranial and Pelvic Traits as a Measure of Sexual Dimorphism in a Modern South African Population. Proceedings of the American Association of Physical Anthropologists 82 nd Annual Meeting. 11 Jantz RL, and Ousley SD. 2010. FORDISC 3.1: computerized forensic discriminant functions. Version 3.1 University of Tennessee, Knoxville, TN. 12 Klingenberg CP. 2011. MorphoJ: an integrated software package for geometric morphometrics. Molecular Ecology Resources. 13: 353-357. 12 O’Higgins P, and Jones N. 1998. Facial growth in Cercocebus torquatus: An application of three dimensional geometric morphometric techniques to the study of morphological variation. J of Anatomy. 193: 251-272. 14 Bytheway JA, and Ross AH. 2010. A Geometric Morphometric Approach to Sex Determination of the Human Adult os coxae . J Forensic Sci. 55(4). 15 Klales AR, Vollner JM, and Ousley SD. 2009. A New Metric Procedure for the Estimation of Sex and Ancestry from the Human Innominate. Proceedings of the American of Academy of Forensic Sciences. Vol. 15.