The ancient origins of African pygmies

Inferring the Demographic History of African Farmers and Pygmy Hunter-Gatherers Using a Multilocus Resequencing Data Set:

The transition from hunting and gathering to farming involved a major cultural innovation that has spread rapidly over most of the globe in the last ten millennia. In sub-Saharan Africa, hunter-gatherers have begun to shift toward an agriculture-based lifestyle over the last 5,000 years. Only a few populations still base their mode of subsistence on hunting and gathering. The Pygmies are considered to be the largest group of mobile hunter-gatherers of Africa. They dwell in equatorial rainforests and are characterized by their short mean stature. However, little is known about the chronology of the demographic events--size changes, population splits, and gene flow--ultimately giving rise to contemporary Pygmy (Western and Eastern) groups and neighboring agricultural populations. We studied the branching history of Pygmy hunter-gatherers and agricultural populations from Africa and estimated separation times and gene flow between these populations. We resequenced 24 independent noncoding regions across the genome, corresponding to a total of ~33 kb per individual, in 236 samples from seven Pygmy and five agricultural populations dispersed over the African continent. We used simulation-based inference to identify the historical model best fitting our data. The model identified included the early divergence of the ancestors of Pygmy hunter-gatherers and farming populations ~60,000 years ago, followed by a split of the Pygmies' ancestors into the Western and Eastern Pygmy groups ~20,000 years ago. Our findings increase knowledge of the history of the peopling of the African continent in a region lacking archaeological data. An appreciation of the demographic and adaptive history of African populations with different modes of subsistence should improve our
understanding of the influence of human lifestyles on genome diversity.

Most people know the standard Out of Africa model. ~50-10,000 years before the present modern humans left the African continent, therefore extant human populations today are descendants of this migration event. The main argument on the margins is only about the possibility of introgression of genetic variants from other non-African lineages into the human gene pool as a supplementary assimilation to the dominant dynamic of replacement. But the story in Africa did not end with that. It is famously well known that Africans have the most genetic diversity of human populations, arguably more than all other populations combined (from mtDNA, Y lineages and more recent autosomal studies). There is population structure, Africa did not remain in stasis after the ancestors of non-Africans left ~75,000 years ago. This paper addresses some of the deeper questions about African structure.

The hypotheses being tested by the researchers in this paper are outlined in Figure 5. To the left you can see the 4 models. "AGR" = Agricultural, "WPYG" = Western Pygmies and "EPYG" = Eastern Pygmies. The usage of a cultural term like "Agricultural" is due to the fact that the Pygmies are distinctive in their language, as they speak the dominant dialects of their region, but their lifestyle, which is dissimilar from that of their neighbors. The Pygmies still engage hunter-gathering, while other populations do not. Additionally they are characterized by the physical peculiarity of relatively short stature, ergo the name "Pygmy."

Pygmies have been of intense scholarly interest, most famously the work of Colin Turnbull, but also by the geneticist L. L. Cavalli-Sforza in his book African Pygmies. While the cultural details of the Pygmies were of interest to anthropologists, the short stature naturally would pique the curiosity of geneticists. Additionally the Pygmy groups could not be understood except in light of a relatively recent event, the Bantu Expansion. As I noted above Africa has not been in stasis. Genetically adaptations to malaria are relatively recent, and almost certainly due to ecological changes wrought by the rise of agriculture as the dominant mode of life across much of the continent. The spread of the Bantu seems to fall into L. L. Cavalli-Sforza's classic demic diffusion, whereby a population which adopts farming begins to rapidly increase in population up to the Malthusian limit, and then advance as a demographic wave which carries their cultural toolkit forward at a rapid clip. From Cameroun these peoples swept east across the savanna, then south along the littoral of the Indian ocean until they reached the margins of the Cape of Good Hope, where a Mediterranean climatic regime renders the Bantu agricultural techniques relatively inefficient. Another group pushed south through central Africa and the Congo basin down into what is today Angola. All the while these peoples admixed with the local populations, as evidenced by the obvious Khoisan phenotypes among the Xhosa people (e.g., Nelson Mandela) who were along the leading edge of the demographic advance. But they retained their language, whose relatively recent diversification is indicated by the similarities of the Bantu languages from the margins of West Africa all the way into modern South Africa (I have read that most Bantu languages are intelligible, though obviously can not attest to this personally).

Culturally the Pygmies were swept up in the Bantu Expansion, as they speak Bantu languages. This is not exceptional for indigenous peoples, the Vedda of Sri Lanka speak Sinhalese. The similarites and interdependencies between the agriculturalists and Pygmies in a given region have of course given rise to models which suggest that the Pygmies are not autochthonous at all, but rather instances of convergent evolution across the forest zone of Central Africa which were shaped by local adaptation to the hunter-gatherer lifestyle. The recurrent physical motif of Pygmy and Pygmoid peoples across Island Southeast Asia without any suggestion of a phylogenetic closeness between these groups opens up this as a plausible option. The geographic separation between the West Africa and East African Pygmies is not trivial, on the order of 1,000 miles or 1/3 the distance across of the United States. And the terrain across the interval is fragmented by forests and river systems. In fact the Congo river serves as the north-south division line between Bonobos and the Common Chimpanzee, suggesting the power of geography in driving speciation.

The data reported here do not support the position that the Pygmies evolved recently from agricultural populations. Here is a reedit of Figure 1 from the paper (so it can fit on the screen easily). You see the agricultural populations the researchers used, as well as the two Pygmy groups, and where they are located. The second set of figures shows how the results are displayed using Structure. Assuming a number of populations defined by K you see where individuals within these populations sort out. Admixed individuals exhibit the admixture in their genotypes, while those who are not do not. At K = 2 you see immediately that there is a difference between Pygmies and non-Pygmies. Additionally the researchers observe that the two Pygmy populations which are sometimes termed "Pygmoid" because of their comparative acculturation to Bantu populations among whom they live show the greatest evidence of admixture and gene flow. Then at K = 3 the two Pygmy groups separate out. In other words, the agriculturalists are the outgroup to the Pygmies, and Pygmies are a phylogenetically supported group, though one with two branches. This does not prove that Pygmy stature is necessarily a common derived feature inherited from the ancestral population, for that you would need to look closely at the genetic architecture of the trait in question and the phylogeny of the haplotypes on the implicated loci. In any case the second set of Structure results are filtered for individuals with less than 20% admixture, and so the between group differences are even starker. It seems likely then that whatever the origin of the Pygmy phenotype, they are the descendants of the pre-Bantu substrate which was extant across Central Africa before the great expansion, and these pre-Bantu populations had a common ancestry.

Now the relationship the populations in question is established, the temporal dynamics are next up. Using a standard molecular evolutionary method, Tajimas D, they infer that the agricultural populations have been through a demographic expansion recently. Going back to the Bantu Expansion this gels well with what we know from non-genetic disciplines. Farming populations expand fast up to the Malthusian limit. But what about the Pygmies? If the model of Pygmies as non-farmers is true, combined with their long residence in their localities, one would assume less evidence of a recent expansion. The carrying capacity for a hunter-gatherer lifestyle is lower, the rapid expansion of farmers is due to their switch to that lifestyle, or their emigration to "virgin land." Neither of these would hold for hunter-gatherers who are indigenous of long standing. Though classical tests implied constant population size, putting more assumptions into the model (ones which are likely or supported, such as gene flow between populations and so forth) gives a more nuanced result:

A bottleneck beginning 2,500-25,000 years ago with an 80% decrease in population size, followed by a recovery starting 125 years later with a size increase of between 100% and 400% (Figure 4), fitted the WPYG data significantly better than the constant-sized population model (P = 0.04, see Materials and Methods). For the EPYG group, a bottleneck starting 250-2,500 years ago with a 90 to 95% decrease in population size (Figure 4) fitted the observed genetic diversity significantly better than the constant-sized population model

I wouldn't get too caught up in the specific numbers here, but the evidence of a population bottleneck is suggestive in light of what we know from history and anthropology when farmers and hunter-gatherers come into contact: the latter are often decimated by disease and warfare due to the appearance on the scene of the former. The genetic data opens up the a more detailed historical narrative which would be impossible otherwise because there are no extant records from this period in this region, and archaeological evidence is difficult to extract due to the unfavorable ecology for preservation. The bounce-back of the Pygmy population could be because after the initial shock of contact with the relatively rapacious agricultural societies the hunter-gatherers retreated to the regions where their lifestyle was still at a comparative advantage and eventually established a symbiotic mondus vivendi whereby both groups might benefit. But that's just a story, more numbers. Using a combination of the molecular genetic data, demographic assumptions and simulations they come up with some time frames for separation of the populations:

...finding that the ancestors of AGR and PYG populations diverged ca. 60 Kya is consistent with our recent single-locus estimation based on the mtDNA diversity of African farmers and Western Pygmies...Most of the large waves of population expansion and migration, both within and out of Africa, have been dated at ca. 50-80 Kya, based on several genetic markers...It has been suggested that these early population movements within and out of Africa may have been triggered by rapid environmental changes. During this period, sub-Saharan Africa witnessed a major episode of climatic change: a sharp oscillation towards a drier climate, with annual rainfall decreasing by up to 50%...In this context, our estimated date of the initial divergence between the ancestors of present-day farmers and Pygmies implies that this period was characterized not only by major human movements, but also by early episodes of population differentiation within the African continent.

Our evidence for a separation of the ancestors of Western and Eastern Pygmy groups ca. 20 Kya is also consistent with a previous mtDNA study, dating the time of separation of these two Pygmy groups to at least 18 Kya...These estimates coincide with another period of major climatic change, the Last Glacial Maximum, which led to a massive retreat of tropical forests in Central Africa...Our genetic results thus support the anthropological hypothesis that the ancestors of present-day forest specialists -- Western and Eastern Pygmies -- began to diverge at the same time as the rainforest retreated into refugia, ~20 Kya...Finally, our estimates of gene flow between each group of Pygmies and agricultural populations yielded contrasting values, with levels of gene flow between WPYG and AGR populations three to seven times higher than those between EPYG and AGR populations...

These data suggest the intriguing possibility that concurrently with the Out of Africa event the seeds for the structure in modern African populations was already being laid. No doubt there were many populations between the ancestors of the Bantus and the Pygmies whom they enveloped, but as I suggest above it may be that the Pygmy groups we see are trivial remnants of the extant populations which were pushed aside by the agriculturalists as they expanded. The Western and Eastern Pygmies may have had the good luck to be in close proximity to ecological domains where their lifestyle was at a comparative advantage and so were able to maintain some sort of demographic and population genetic integrity in the face of the agricultural onslaught. Ergo, the genetic discontinuity might be an outcome of these details of history, as the groups which span the gap between the proto-Bantu and the Pygmies left very few descendants.

I wonder therefore the need to appeal to refugia ~20,000 years ago, as opposed to the Western and Eastern Pygmy simply being the western and eastern edges of a continuous population of which they are the last fragments. I do know that there are data which suggest that the speciation of some African monkeys is a function of the fragmentation of the rainforest during the Last Glacial Maximum, but most monkeys are by nature forest animals, while humans are not. Therefore I am skeptical of the thesis that ecological fragmentation is the reason behind the genetic differences between the two Pygmy groups dating to ~20,000 before the present. Much more likely that the Pygmy populations retreated to regions where hunter-gathering was still viable, and their genetic distance is simply an echo of population structure since erased by the Bantu demographic expansion. On the other hand, if subsequent data showed that the Pygmy phenotype is of very ancient pedigree, then perhaps the concept of perpetually forest resident populations might seem more plausible.

Hopefully there will be some examination of the other pre-Bantu populations of Southern Africa in the near future, the Hadza and Khoisan.

Citation: Patin E, Laval G, Barreiro LB, Salas A, Semino O, et al. (2009) Inferring the Demographic History of African Farmers and Pygmy Hunter-Gatherers Using a
Multilocus Resequencing Data Set. PLoS Genet 5(4): e1000448. doi:10.1371/journal.pgen.1000448