271-257 thousand years ago
You’ve probably run into some version of the factoid that there is more genetic variation in Sub-Saharan Africa than in all the rest of the world. This assertion has to be handled with care. It doesn’t necessarily apply to genes that have been under strong diverging selection pressures on different continents. Consider skin pigmentation: there is not more variation inside Africa than outside it in skin color, or genes for skin color. Obviously. Likewise for hair form. But it’s true for neutral genetic variation, which is most genetic variation.
The simplest way to account for the broad Africa/non-Africa distinction would be to assume a large homogenous founder population in Africa, with a smallish number of people leaving Africa and going through a genetic bottleneck, thereby reducing their genetic variation. But recently we’ve been learning that the African situation is more complicated. Specifically, there used to be a lot of genetic differentiation between different regions within Africa. Recent population movements have smoothed out some of that variation, but recent work on ancient DNA has been bringing this more variegated past to light.
A case in point: the latest data imply that the Bushmen of Southern Africa separated from other African populations (East African, West African) around 260,000 years ago (at least), long before the major Out Of Africa venture by modern humans. What’s more, the very latest data imply that the Bushmen have received outside genetic input pretty recently, in the last 1-2 thousand years. This admixture, 9-22% of the ancestry of modern Bushmen, is absent from a 2,000 year old skeleton form Ballito Bay, South Africa.
The intruding population were probably pastoralists whose livestock, and a fraction of their genes, ultimately derived from the Near East. Another fraction of their genes originated in the Sahel or East Africa. And they probably spoke a language in the Nilo-Saharan or Afro-Asiatic family. These language families pop up as a substrate in East Africa, although largely overlain by the later expansion of Bantu speakers.
One implication: Bushman groups like the !Kung have often been presented as models for our Pleistocene hunting and gathering ancestors. Yet the most recent findings imply that there has been substantial interaction, including gene flow, between Bushmen and non-hunter-gatherers for some time.
A relevant result from twentieth century anthropology: when Nancy Howell did her classic work on the demography of the Dobe !Kung Bushmen, she found that, when you look at female genealogies, the Dobe !Kung look like a growing population, but when you look at male genealogies, they look like a shrinking population. There’s no contradiction here: the women, but not the men, in the population were sometimes having children by outsiders, neighboring pastoralists. The pastoralists in question were Bantu, having arrived in the last few centuries, but the latest genetic data imply that something similar was going long before the Bantu showed up. Since it’s not clear what effect this subaltern sexual status might have had on Bushman social organization, the social life of historic Bushmen may not be a good model for hunter-gatherer life before agriculture.
Here’s the article on Ballito Bay Boy.
And a recent review article from Nature
And assorted recent blog posts from Razib Khan, on the ancient Bushman split, and more recent African prehistory.