Assessing Multiple Lines of Evidence for Gene Flow in Archaeological Contexts

This multi-study dissertation assesses the ability of two skeletal analysis methods—a model-bound quantitative genetic method (Relethford-Blangero) and a model-free biological distance method (Mahalanobis’ D²)—to evaluate gene flow in the U.S. Southwest and Northwest Mexico based on archaeological models. The first study uses dental metric data from the Sonoran Desert and Mogollon Rim (c. 1600 B.C. to A.D. 1450) to pilot the Relethford-Blangero method in this context. Notably, the method shows that populations from two large sites have less than expected dental variance, failing to support a gene flow event despite material culture pointing to at least two coexisting communities. This suggests even supported conclusions from a material culture perspective will not necessarily match the biological relationships experienced by past communities. The second study compares both method groups using a sample of five sites that overlap temporally (c. A.D. 1100-1450) from the Mogollon Rim region, and also compares different groups of skeletal traits (cranial, dental, and postcranial) to each other to determine the efficacy of using each to understand gene flow. While there are some consistencies between skeletal trait groups, their congruence lacks a clear pattern. Biodistance results lack consistency regarding which groups may be most closely related, while the Relethford-Blangero method compliments archaeological data in some instances. The third study uses data from the limbs and their girdles to evaluate a larger geographic and temporal region. Sites from across the Puebloan world (c. A.D. 840-1590) are compared to address how well postcranial data mirror the archaeological models. Though not all results are consistent with archaeological models, there is a pattern showing that the ways of interacting for the Western Pueblo (Arizona) are different than those of the Eastern Pueblo (New Mexico), supporting separate cultural classifications devised by archaeologists. In addition, the lower limb appears to drive the overall postcranial signal more than the upper limb alone. Overall, while the quantitative genetic method adds additional details to the archaeological models, biodistance is sometimes difficult to interpret. The quantitative genetic method should be added to the analytical toolkit in all cases, not just those of potential ambiguity.