Abstract
Previous osteological studies of East Polynesian populations have primarily focused on the identification of ancestral homelands and have ignored issues relating to regional and local populations. This thesis examines East Polynesian population history and addresses questions concerning East Polynesian prehistory from a bioanthropological perspective. The results are considered in conjunction with previous studies drawing from archaeology, linguistics and biological anthropology. Of specific interest is bioanthropological support for models of prehistoric East Polynesian colonisation and interaction networks, as well as regional and local population variation.
Multivariate analysis of cranial nonmetric data from 984 individuals demonstrates a high degree of gene flow between central East Polynesian populations, including the Southern Cook, Society and Tuamotu islands. The results provide support for the existence of a central East Polynesian (CEP) interaction sphere in prehistory. The Marquesas Islands were peripheral to the major areas of gene flow and were not part of the main CEP interaction network. Although geographically marginal, Hawaii likely received gene flow from both the Marquesas Islands and the CEP interaction sphere. I suggest that Mangareva was influential in southeast Polynesia, possibly serving as a crossroads between the Tuamotus and the smaller islands to the east. The ancestral homeland of the Rapa Nui people remains uncertain, although either Mangareva or a location in the Tuamotus remains the most likely option. New Zealand Maori origins are probably from the CEP interaction sphere, while the Chatham Islands were likely settled from New Zealand.
Analysis of local populations suggests patterned variation within the Marquesas Islands and the North Island of New Zealand but not for Rapa Nui. In addition, I find no evidence for prehistoric gene flow between South America and Rapa Nui. The results highlight the potential value of bioanthropological studies of local East Polynesian populations.
I also raise a number of relevant methodological issues. In particular, I demonstrate that the inclusion of samples of varying size within the Mean Measure of Divergence analysis significantly alters the results of the statistic. In addition, I argue that group homogeneity should be established a priori to analysis and that constructing samples based on other criteria should be avoided.
Finally, this thesis demonstrates the important role biological anthropology can play in the study of East Polynesian prehistory. The lack of relevant bioanthropological studies of local East Polynesian populations and an almost singular bioanthropological focus on ancestral homelands have helped restrict a holistic approach to East Polynesian prehistory. It is my intention that this study will help fill the previous bioanthropological void in East Polynesian studies.