Can first phalanx multivariate morphometrics help document past taxonomic diversity in South American camelids?

Abstract

South American Camelids (SAC) were of great economic and cultural importance for pre-Hispanic societies in the Andes and still are important for rural Andean communities. However, understanding their specific role and function over time is hampered by the lack of reliable taxonomic identification of their archaeological remains. Although this problem has been acknowledged since the 1970s, the identification of SAC from archaeological contexts still mostly relies on the first phalanx multivariate morphometrics developed by Kent in 1982. The goal of this study is to further explore the taxonomic potential of this method using a machine learning approach including new measurements of phalanges from modern Peruvian, Bolivian and Argentine SAC with additional measurements taken from the literature. Results have shown that fore phalanges reliably distinguish the wild guanaco and vicuña, but the distinction between the domestic forms and their wild relatives still remains tentative despite the new statistics. This is largely due to limited domestic SAC comparative material, especially for the alpaca, but also underlines the presence of an intermediate size morphogroup among the domestic SAC. Morphometric variations recorded for the pre-Hispanic SAC are greater than those of current SAC, preventing reliable identification of the archaeological specimens. These results emphasize the need to improve the current SAC measurments data base and develop new multiproxy approaches in the study of pre-Hispanic camelid diversity directly from the archaeological material.