Research Training Program

Commingled assemblages are a common occurrence in many faunal investigations. In the past methods of sorting mixed assemblages have focused on taphonomic characteristics, provenience, simple metrics, or best conjecture. This project focuses on sorting complete commingled human femur and tibia remains using biometric and discriminate functions, and fragmented remains using a z-score statistical analysis. Two population groups were analyzed, 201 black and white individuals from the anatomical Robert J. Terry Collection, and 28 individuals from the South Dakota Mobridge Site (39WW1), representing individuals from the Arikara people dated between 1600 and 1830 AD. Forty-four variables were obtained for each skeleton on the femur and tibia. A discriminate function analysis comparing sides of elements from the same individual with the sides of elements of different individuals yielded sectioning points to sort complete elements with greater then 85 percent accuracy. Averages and standard deviations of variables, variable differences, and variable differences represented as a percentage, have been calculated for the Terry Collection sample. These calculations help define the nature of lower limb asymmetry, variable asymmetry variation, and can be used in sorting fragmented femur and tibia remains with a z-score statistical analysis. The methods used in this project will be valuable in sorting commingled remains in archaeological, and paleontological, and forensic assemblages. Defining elemental relationships in commingled faunal assemblages will improve understanding of their biologic and taphonomic characteristics, giving a more refined understanding of the nature of paleoecological systems.