Sex-biased migration in humans: what should we expect from genetic data?

Different patterns of mitochondrial and Y-chromosome diversity have been cited as evidence of long-term patrilocality in human populations. However, what patterns are expected depends on the nature of the sampling scheme. Samples from a local region reveal only the recent demographic history of that region, whereas sampling over larger geographic scales accesses older demographic processes. A historical change in migration becomes evident first at local geographic scales, and alters global patterns of genetic diversity only after sufficient time has passed. Analysis of forager populations in the ethnographic record suggests that patrilocality may not have predominated among pre-agricultural humans. The higher female migration rate inferred by some genetic studies may reflect a shift to patrilocality in association with the emergence of agriculture. A recent global survey does not show the expected effects of higher female migration, possibly because the sampling scheme used for this study is accessing pre-agricultural human migration patterns. In this paper, we show how the demographic shift associated with agriculture might affect genetic diversity over different spatial scales. We also consider the prospects for studying sex-biased migration using the X-linked and autosomal markers. These multi-locus comparisons have the potential of providing more robust estimates of sex differences than Y-linked and mitochondrial data, but only if a very large number of loci are included in the analysis.