Canine dimensions for estimation of sex in adult and non-adult individuals with external validation by aDNA.

Abstract

Using discriminant functions obtained from canine dimensions for sex estimation in human skeletons has frequently been proposed as a promising approach within assemblages, even when used in non-adult individuals. However, applicability of this method to adult and non-adult individuals from other assemblages was rarely investigated, probably due to frequently observed inter-population differences in tooth dimensions. In the present study, discriminant functions obtained for permanent canine dimensions at the cemento-enamel junction in a previous study of the early medieval assemblage from Greding, were applied to individuals from a late medieval Jewish cemetery at Erfurt, Germany. The results were validated by aDNA analyses. Prior to the application of the functions, canine dimensions of the assemblages were compared. The comparison showed largely corresponding canine dimensions between the two assemblages. The application of the formulae obtained on the early medieval assemblage to the late medieval assemblage at Erfurt revealed a 100 % correct classification rate in the adult individuals. In non-adults, the correct classification rate was poorer, with 7 of 9 (77.8 %) individuals correctly classified. The study showed that the application of discriminant functions for sex estimation from canine measurements to assemblages other than those for which the functions were developed can lead to high correct classification rates in adults if the average canine dimensions are similar in the respective assemblages. An application to non-adult individuals should only be made with caution as canine dimensions in the "non-survivors" can lead to an over-estimation of the proportion of female non-adults.