Postcranial indicators of primate sexual dimorphism: implications for reconstructing fossil hominin sexual dimorphism and hominin palaeoecology

Abstract

Sexual dimorphism can be used to reconstruct various aspects of hominin palaeoecology; however, previous studies have highlighted problems with the current methodologies for estimating both sex and body mass. This includes the influence of body mass estimation techniques on the prediction of dimorphism and choosing the correct comparative sample. Increased understanding of sexual dimorphism within the primate order may improve the accuracy of methods used for estimation. Here the structure of sexual dimorphism for nine primate species was investigated through twelve postcranial indicators of skeletal dimorphism. Discriminant function analysis was used to assess the best metric discriminators of sex and was evaluated as a method for classifying sex in fossil hominin specimens. Differences in skeletal metric pair correlation coefficient values between the sexes were also used to investigate variation in the structure of sexual dimorphism. Skeletal dimorphism within the primate order was found to be non-isometric, with upper limb metrics being generally better discriminators of sex for dimorphic primates. This includes Homo sapiens upper limb metrics, although femoral head diameter is a higher ranked discriminator of sex for Homo sapiens than it is for other dimorphic primates. Discriminant function analysis achieves greater accuracy in estimating sexual dimorphism than previous methods and accuracy is sustained when using a smaller number of the best skeletal metric discriminators. The pattern of skeletal metric correlation coefficient difference between males and females varies across the primate order and similarities between species do not consistently reflect phylogenetic relationships. Separating the estimation of body mass and the determination of sex within fossil hominin species is important because it reduces the risk of error being introduced through the prediction of sex from body mass. The sustained accuracy of sex estimation through the best skeletal metric discriminators makes discriminant function analysis a practicable method of classifying sex for fossil hominin specimens. Patterns of shape dimorphism, analysed through skeletal metric correlation coefficient values, supplies another method for analysing the complexities of scaling relationships in males and females. Increased accuracy in estimation will lead to greater confidence when inferring various aspects of hominin palaeoecology.