A demographic simulation model to assess prehistoric migrations.

Abstract

This study revisits a traditional anthropological theme of prehistoric migrations. Anthropologists typically assume that changes in culture, as reflected in the archaeological record, reflect changes in populations due to prehistoric migrations. This work addresses this presumed correlation between migrations and culture change and focuses on the question, did people migrate? The study develops a demographic simulation model that draws from ethnographic, paleodemographic and population genetic theory. The model uses mitchondrial DNA frequency data applied to computer simulations to test the following null hypothesis: <italic>the genetic change between two populations separated by time can be explained solely by the processes of random genetic drift, gene flow and/or population growth within a single temporally continuous breeding population</italic>. In other words, this is a null hypothesis of population continuity. If a continuity hypothesis can be refuted, then other processes such as prehistoric migration could account for the temporal changes in genetic frequency data. To demonstrate how the model works, the study evaluates a migration hypothesis from the prehistoric Great Basin of North America. In this case, archaeologists (Bettinger and Baumhoff 1982) and linguists (Lamb 1958) propose that a population replacement via an expansion occurred about 1,000 years ago. Kaestle and Smith (2001) tested this Numic Expansion hypothesis by comparing the mitochondrial frequencies of ancient (pre-1,000 years) and modern population samples. Based on large differences in the DNA frequencies of these samples, they concluded that a prehistoric population replacement had likely occurred in the Great Basin. This study reevaluates the significance of their results under the assumptions of the demographic simulation model. The Great Basin example demonstrates that for this as well as any other data set, a hypothesis of population continuity through time can be rejected (and a migration scenario possibly supported) only under a specific range of conditions of prehistoric population sizes, gene flow and population growth rates, and number of generations separating the ancient and modern population samples.