Quantifying Sexual Dimorphism in Scapular Morphology

Abstract

This project explores sexual differences in scapular morphology through linear measurements (LMs) and geometric morphometric (GM) methods. Traditional LMs assess size while GM methods primarily assess shape. Our two hypotheses are as follows: 
the human scapula expresses sexual dimorphism in both size and shape, and given that shape differences exist, this bone will produce
higher correct classification rates when assessed for sex using GM rather than using LMs. Three-dimensional data were obtained from the UTK Donated Skeletal Collection (n = 106); linear data were obtained from the Forensic Anthropology Data Bank (n = 1,252).

We modeled these data separately to quantify levels of sexual dimorphism in size (LM) and shape (GM) variability. Linear measurements correctly identified 93% of the sample; the GM method could only correctly identify ~70% of the sample when following the GM protocol for the scapular landmark collection outlined in Uhl et al. (2007). However, GM data produced correct classification rates of over 93% when checked for correlation of centroid size. These results indicate that both size and shape drive differences between female and male shoulder girdle morphology, though the primary contributor is sexual size dimorphism. Size-free and/or allometric differences were also noted within this sample but with a much smaller impact on morphology. The human scapula produces high accuracy rates for sex estimation, regardless of whether LMs or GMs are utilized.