1Anthropology, University of Tennessee, Knoxville, 2Anthropology, University of New Mexico
March 26, 2015 , Gateway Ballroom 2
While recent studies have demonstrated that global patterns of human autosomal variation have been shaped by serial founder effects, the extent to which this and other processes have shaped mitochondrial diversity is as yet unresolved. Here we focus on mitochondrial genomic variation to illuminate the role of maternally mediated gene flow and founder effects on global patterns of variation. We examined the correlation between gene identities in 248 globally distributed populations versus geographic distance from a potential East African founder location, and used a generalized hierarchical modeling framework to identify the root location of population trees and to test the fit of the serial founder process to the genetic data. We used partial Mantel tests of gene identity, genetic structure, and geographic distance to identify the independent effects of gene flow and founder effects among populations. Results are then compared to those obtained from autosomal STR data.
The best fitting mitochondrial tree was rooted in East Africa. Gene identities increased with increasing geographic distance from the root location, though the correlation was weak and variation within global regions deviated significantly from the pattern predicted by a strict serial founder process. Additionally, gene flow played a stronger role relative to founder effects in shaping patterns of diversity. We show that incongruences between mitochondrial and autosomal data sets reflect differences in information content, effective sizes, and sex-biased gene flow. We explore the implications of these results for reconstructions of past demography.