Disentangling long bones of foxes (Vulpes vulpes and Alopex lagopus) from arctic archaeological sites

To discern the presence of two anatomically close species within an archaeological assemblage is always a problem for the analyst. A particularly interesting case is the distinction between the red fox (Vulpes vulpes) and the arctic fox (Alopex lagopus), which is often determined using non-rigorous criteria and producing results that must be regarded with caution. Classical methods for separating these two species and determining sexual dimorphism, and eventually the sex ratio, such as analyses using uni- and bivariate plots, are based on arbitrarily fixed limits between sexes. In this text, a more robust statistical method termed mixture analysis (MA) is used to determine the species of limb bones from foxes. First, the MA is applied to a sample of each species using metric data taken from a collection of present-day fox skeletons. Afterwards, the MA is applied to archaeological samples dated to the Dorset period and retrieved from the Tayara site (near Salluit, South Hudson Strait, Nunavik). The results clearly demonstrate that the greatest length (GL) of long bones, especially the humerus and tibia, is the best measurement for distinguishing the species, followed by the distal breadth and the proximal breadth; and that GL is better suited for sexing foxes. The results of MA applied to the Tayara site collection lead to the conclusion that the red fox is present along with the arctic fox among the assemblage, and that there is no fox body size change during the last two millennia in the Eastern Arctic.