Deshpande et al. (2008) on Out of Africa

This is an important new paper which adds some complexity to the Out of Africa theory. Much existing work has focused on a "tree-like" story of the emergence of modern humans, with an African source population at the root, and other populations being less diverse the further they are (geographically) from the source.

This new model is not limited on colonization, i.e., the movement of a subset of a territory's population into a new uninhabited territory, but also on "lateral" gene exchange between pre-established populations.

From the paper:

Unlike previous models, ours separated colonization events from the continued exchange of people between occupied territories. Our estimates of the exchange rate between neighbouring populations were very low (below 0.01), with carrying capacities ranging from approximately 600 to 1200. Assuming that the census size is three times this effective population size, we derive a census size of approximately 1800–3600 people in each deme. Since each deme has dimensions of 125x125 km, this corresponds to a population density of approximately 0.11–0.23 persons m^-2, well within the range for hunter–gatherers referred to by Liu et al. (2006).

Related: Geographic and genetic distance in human populations, A Geographically Explicit Genetic Model of Worldwide Human-Settlement History

Proceedings of the Royal Society B doi: 10.1098/rspb.2008.0750

A serial founder effect model for human settlement out of Africa

Omkar Deshpande, Serafim Batzoglou, Marcus W. Feldman, L. Luca Cavalli-Sforza

Abstract

The increasing abundance of human genetic data has shown that the geographical patterns of worldwide genetic diversity are best explained by human expansion out of Africa. This expansion is modelled well by prolonged migration from a single origin in Africa with multiple subsequent serial founding events. We discuss a new simulation model for the serial founder effect out of Africa and compare it with results from previous studies. Unlike previous models, we distinguish colonization events from the continued exchange of people between occupied territories as a result of mating. We conduct a search through parameter space to estimate the range of parameter values that best explain key statistics from published data on worldwide variation in microsatellites. The range of parameters we use is chosen to be compatible with an out-of-Africa migration at 50–60Kyr ago and archaeo–ethno–demographic information. In addition to a colonization rate of 0.09–0.18, for an acceptable fit to the published microsatellite data, incorporation into existing models of exchange between neighbouring populations is essential, but at a very low rate. A linear decay of genetic diversity with geographical distance from the origin of expansion could apply to any species, especially if it moved recently into new geographical niches.

Link