Ancient genomes reveal complex patterns of population movement, interaction, and replacement in sub-Saharan Africa

 Ke Wang et al. ,Ancient genomes reveal complex patterns of population movement, interaction, and replacement in sub-Saharan Africa.Sci. Adv.6,eaaz0183(2020).DOI:10.1126/sciadv.aaz0183

Tl;dr

• Eastern Africa received admixture from North African/Levantine-related groups during the Neolithic
• Individuals from the Pastoral Neolithic Tanzania and Kenya had 30-40% of that North African/Levantine component
• The authors used group of Levantine individuals " as the closest available proxy for ancient northeastern African ancestry"
• This ancestry "likely reflects a gene pool present more broadly in ancient northeastern Africa and the Levant, as identified in ancient (11, 20) and present-day northeastern African populations"
• The East and Northeast African components were diluted by later Nilotic and Bantu admixture and/or replacement
• "some regions or locations witnessing almost complete replacement from Nilotic-related migrations"
• Ancestry related to present-day Bantu speakers, which is, today, prevalent across sub-Saharan Africa, is absent from most ancient sub-Saharan African genomes analyzed to date." i.e. Eastern and Southern Africa were less "Western African" before the Bantu Expansion and spread of pastoralism
• As usual, this shatters any idea of "racial" purity or "Pan-African" population. North and Northeast Africans have always had affinities and shared ancestry with the Levant, Northeast Africans mixed with East Africans during the Neolithic, then Nilo-Saharan and Bantu speakers mixed and sometimes replaced those Eastern African pastoralists.

Fig. 2 PCA of ancient genomic data analyzed in this and previous studies, together with published modern genetic data. Modern populations shown are detailed in the legend of Fig. 1 and fig. S1. A separate PCA with only present-day populations is shown in fig. S1A. Color circles highlight the key groups discussed in this paper and summarized in Fig. 3A.

Abstract

“Africa hosts the greatest human genetic diversity globally, but legacies of ancient population interactions and dispersals across the continent remain understudied. Here, we report genome-wide data from 20 ancient sub-Saharan African individuals, including the first reported ancient DNA from the DRC, Uganda, and Botswana. These data demonstrate the contraction of diverse, once contiguous hunter-gatherer populations, and suggest the resistance to interaction with incoming pastoralists of delayed-return foragers in aquatic environments. We refine models for the spread of food producers into eastern and southern Africa, demonstrating more complex trajectories of admixture than previously suggested. In Botswana, we show that Bantu ancestry post-dates admixture between pastoralists and foragers, suggesting an earlier spread of pastoralism than farming to southern Africa. Our findings demonstrate how processes of migration and admixture have markedly reshaped the genetic map of sub-Saharan Africa in the past few millennia and highlight the utility of combined archaeological and archaeogenetic approaches.”

Excerpts from the paper

 "During the Pastoral Neolithic (PN), people related to Chalcolithic and Bronze Age Levantine groups entered eastern Africa and mixed there with individuals related to Later Stone Age foragers and with individuals related to present-day Dinka in what was proposed to have been at least a two-step process (4). Ancestry related to present-day Bantu speakers, which is, today, prevalent across sub-Saharan Africa, is absent from most ancient sub-Saharan African genomes analyzed to date."

"(ii) indications that the arrival of pastoral populations in eastern Africa resulted from the movement of several discrete groups of herders from northern to eastern Africa"

"Kenya_400BP presents extra affinity with present-day west Eurasian people (3). We further characterized genetic ancestry components of these ancient African individuals through qpAdm (19), a method to estimate ancestry proportions related to specified source populations. We found Kenya_Kakapel_3900BP has 18 ± 6% Mbuti-related ancestry, and the published Kenya_400BP has 11 ± 3% ancestry related to ancient Levantine individuals (Fig. 3 and table S3), which likely reflects a gene pool present more broadly in ancient northeastern Africa and the Levant, as identified in ancient (11, 20) and present-day northeastern African populations"

"When modeling Congo_Kindoki_150BP with qpAdm (Fig. 3B and table S5), we found a fitting model with 85 ± 7% ancestry related to Bantu speakers and 15 ± 7% ancestry related to western Eurasians. This ancestry profile would be consistent with the hypothesis that this individual has Portuguese ancestry, which would fit with the colonial history of the region (44)"

"a group of ancient Levantine individuals (24), which we take as the closest available proxy for ancient northeastern African ancestry (10, 11)"

"While the estimated proportions of Levantine-related ancestry in all samples are rather constant (around 30 to 40%)"

"Kenya_Kakapel_900BP also shares close genetic affinity with Dinka but requires an additional small ancestry component (12 ± 3%) from northeastern African/Levantine groups, similar to the ancestry component in early PN herders (Fig. 3 and table S4). We dated this admixture between Dinka- and Levantine-related ancestries in Kenya_Kakapel_900BP to around 500 ± 200 years before the death of that individual, consistent with the onset of the Iron Age in the region. This suggests that the Iron Age population represented by this single individual resulted from admixture between PN-related herders and incoming Nilotic agropastoralists, rather than resulting from a major migration of people with West African–related ancestries."

"The notable shift seen in the two Iron Age individuals from the Kakapel site to almost 90 to 100% Nilotic-related ancestry, compared to about 40% during the Pastoral Neolithic, is substantially larger than the increase in Nilotic ancestry seen in previously analyzed eastern African individuals from the Iron Age (4). In addition, the absence of ancestry related to present-day Bantu speakers in Kenya_Kakapel_900BP contrasts with the finding of this ancestry in a contemporaneous individual from the site of Deloraine farm in the Central Rift Valley of Kenya (4). This shows that patterns of dispersal and admixture in Iron Age eastern Africa resulted in a complex geography of ancestry, with some regions or locations witnessing almost complete replacement from Nilotic-related migrations (33), others seeing mixing of diverse peoples (4), and yet others demonstrating no admixture from ancestry related to Nilotic or Bantu speakers into recent centuries (as seen in Kenya_400BP)."

Fig. 3 Admixture history of ancient African populations.
(A) Overview of coexistence of distinct African ancestries through time drawing on currently available ancient genomes. (B) Ancestral components of ancient African groups/individuals according to qpAdm. We order ancient groups in the same order shown in (A) and highlight newly reported genetic groups/individuals in bold. Pvalues and estimated ancestral proportions can be found in table S3.