Wednesday, June 15, 2016

The African & West Eurasian elements in Aris

Aris, similar to other Horn Africans, have both African and West Eurasian ancestry. They have notably less than Somalis, Oromos, Amharas, Agaws, Wolaytas and the like but the ancestry is most certainly present:

Their West Eurasian admixture, from an autosomal perspective, also seems quite old, much like most of the West Eurasian elements in Somalis. In fact, it seems even "older" than some of what is found in Somalis in that their West Eurasian admixture, at least in the case of Ari Blacksmiths who mostly lack Somali-like/Cushitic speaker related admixture unlike Ari Cultivators, lacks more northerly West Asian affinities:

At K=5 above; you can see Somalis, Oromos and co. show "Arabian" & "European" like ancestry but Ari Blacksmiths show only Arabian-like ancestry. They seem to be bereft of the more Mediterranean and/or Caucasus-Gedrosia related elements in the Cushitic and Ethiopian Semitic speaking populations and are chiefly made up of Southwest Asian related elements. As in, very basal West Asian type ancestry much like David Wesolowski's old ENF cluster from his K=8 run or Lazaridis et al. 2013-2014's Near Eastern component.

Now, we don't have any actual time-stamps for their West Eurasian admixture (reliable ones, anyway) but the fact that it seems to be such basal looking stuff which lacks affinities that would signify later (perhaps even Post-Neolithic) West Asian-related admixture into the Horn region, such as Anatolian Neolithic-like (Mediterranean) or Caucasian Hunter-Gatherer-like (Caucasian or Caucasus-Gedrosia) ancestry, definitely implies that this is likely quite pre-historic admixture perhaps from a time when certain areas of West Asia such as the Levant lacked Caucasus Hunter-Gatherer-like and more overt Villabruna/WHG-like elements, the latter of which causes a similarity to Neolithic Western Anatolians and Early European Farmers.

It's interesting, in that case, to note that the dominant West Eurasian Y-DNA Haplogroup among Omotic speaking populations such as the Ari is actually J (J*, J1 & J2) rather than T which seems much more dominant among Cushitic speaking Horn Africans with very little to practically none of the later West Eurasian elements in their kin further inland in the Horn (I.e. Somalis). 

Is a lot of their Y-DNA J inherited from Cushitic and Ethiopian Semitic speakers? What about the interesting case of J* though? Could it be that a lot or most of their Y-DNA J lineages date back to when their ancestors first acquired the West Eurasian ancestry in the f-statistics based table from Pickrell et al. 2013 above? That most of their J lineages will turn out to be wildly divergent from some of what's found in Cushitic and Ethiopian Semitic speakers? It'll be interesting to know someday when we have more high-resolution data and can pinpoint their exact subclades.

But, there's little point in meticulously mulling over the rather low-resolution data we have for the time being.

At any rate, the non-Eurasian ("African") elements in them are much more intriguing, in my humble opinion. For one, a large segment of it seems very much like the majority of the African elements in Somalis & Habeshas or, essentially, very much like most of the ancestry in populations such as Anuaks and Dinkas:

Notice how at the K=6 results above; most of the ancestry in the two Ari groups belongs to the Nilo-Saharan cluster (essentially what I took to calling the East African cluster over a year ago) which shows among the Sudanese (Southern) and Anuak samples at a level comparable to that of the Ari Blacksmiths and Cultivators.

The rest seems to be some odd African Hunter-Gatherer like elements (as in, ancestry seemingly quite similar to but not the same as what dominates the San & African Pygmies like Mbutis). In fact, even the ancient Mota man, who shares a lot of alleles with Aris, shows an affinity for these groups who remained Hunter-Gatherers/Foragers all the way into the Modern era:

Notice how Mota breaks the line between the Somali-Beta Israel cluster & Western African cluster; he breaks off and pulls toward the "San" samples alongside Hadzas who also seem like an intriguing mixture of what looks to be very old East African-like ancestry and something similar to groups like the San and Mbutis.

The rest of the non-Eurasian ancestry in Aris (which seems to make up much more of their ancestry than this San-Mbuti-like stuff) seems to be very similar to most of the African ancestry in Somalis & populations such as Dinkas as mentioned before. In fact, despite its West Eurasian & Mbuti-San-like elements, the Ethiopic/Omotic cluster which turns up in some ADMIXTURE runs is rather close in regards to Fst divergence to the "Nilo-Saharan" / East African cluster of which it often seems to be made up of at the lower Ks:

However, I would consult an old post of mine where I basically delved into how things such as Fst can be misleading when dealing with various African populations. You see, the African elements in Somalis aren't very "divergent" from the African elements that make up about +95% of Yorubas' ancestry if we're consulting parameters such as Fst which can be good at picking up on the genetic drift between populations.

This is partly because pre-historic African populations seemingly had much larger population sizes than Eurasian ones [5] and thus didn't seemingly diverge from one another as much as Eurasians did via genetic drift with population dips such as the "Eurasian bottle-neck" (the shared ancestral population between all Out-of-Africa populations from Native Americans, Eastern Non-Africans and West Eurasians) can be quite conducive to. And Humans mostly diverge via these two mechanisms:

Genetic Drift

For genetic drift what happens is; Population A & Population B are more or less "identical" and each have 1,000 people in them but they are geographically separated for 10,000 years... Population A doesn't experience any odd population losses and just continues to prosper from the original 1,000 roster but Population B loses 800 people due to an Earthquake very early on and the descendants of Population B find themselves being descended, 10,000 years into the future, from only those 200 folks who survived the Earthquake; losing a lot of their population's prior genetic diversity and diverging from Population A as a result of this alongside later mutations and selection and so on.

"Admixture" on the other hand would mean Population B's 1,000 remained intact in terms of numbers but they discovered a far away Population C with ancestry distinct from Population A & B and then intermixed with them to form a new mixed population that is now divergent from Population A due to being a mixture between Population B & C.

Between various African populations; genetic drift is not something that was as relevant during their pre-histories given their generally larger population sizes so they largely maintained and built upon their prior genetic diversity rather than experiencing the constant founder effects and bottlenecks, on an autosomal level, found outside of Africa. Though some notable drift definitely occurred.

So, when consulting something like Fst which focuses on differences in allele frequencies; things may not prove as useful when comparing various African populations as they might when comparing Africans to Out-of-Africa populations or OoA populations to each other. What might in fact prove more useful would be Haplotypic divergence datings or the general differences implied by Haplogroups (Y-DNA and mtDNA). 

For example, observe the "Khoesan" cluster (makes up pretty much all of the ancestry in the San samples) & the "Niger-Congo" cluster (makes up pretty much all of the ancestry in Yoruba samples) and notice that the Niger-Congo cluster is closer to the Khoesan cluster than it is to the East Asian one. However, when observing divergence dating between various populations such as Yorubas and the San:

The Yoruba are actually closer to all Out-of-Africa populations than they are to the San. As in, they seemingly share an ancestral population with Out-of-Africa populations that they do not share with the San which is why you very often see diagrams like this one being thrown around by various studies when trying to simplify the Homo Sapien Sapien family-tree:

The Yoruba and the San are not as drifted from one another as East Asians (i.e. The Han) and the Yoruba are but it seems obvious that when grouping these populations together; the Yoruba look to belong under the same branch as West Eurasians and Eastern Non-Africans which then shares an ancestor with the San. [important note]

What was my point in explaining this? Well, the point is simply that given people like the Aris' distinct Y-DNA profile; despite the majority of their ancestry not seeming particularly "divergent" from most of the similar African ancestry in Somalis and Nilo-Saharan speakers such as the Dinka; it may very well have diverged from the predominant African elements in these populations a very long time ago. Just as Haplogroups imply to us that the African element in Yorubas, as I mentioned here, whilst seemingly very close to the African element in Somalis, likely diverged from it well-over 20,000 years ago. 

So when observing various African populations (the ones that mostly lack Eurasian admixture); I would not allow myself to be fooled by the sometimes low, or simply not as high as you would expect, divergence between them when utilizing analyses such as IBS, Fst and so on which, in a way, have a bit of a Eurasian bias about them in that they're used so much because they make sense for Eurasian/Out-of-Africa populations but not as much for numerous African populations.

In fact, Wolaytas, despite being substantially Ari Blacksmith-related in ancestry, will have Somalis seem closer to them in terms of IBS (Identity-by-State) than Tigrinyas who have more Ari Blacksmith-related admixture than Somalis:

That's because Wolaytas are only a little bit less West Eurasian, on average, than Somalis whilst most of the African elements in them, however long-ago some of it cut-off from the African elements in Somalis, is still not very drifted/diverged from what's in Somalis so the two populations ultimately seem quite genetically close even though, in terms of actual shared ancestry within the last 3,000-10,000 years; Somalis are much closer to Tigrinyas.

I for one will be interested in seeing what we learn via ancient DNA across Africa period by period when the time comes. We do have some samples on the way such as 10,000 year old samples from Kenya from what I recall... For all I know; a lot of what I've touched upon here will be very redefined by ancient DNA data.

Reference List:

1. Early Back-to-Africa Migration into the Horn of Africa, Hodgson et al.

2. Variation in Y chromosome, mitochondrial DNA and labels of identity on Ethiopia, Plaster et al.

3. mtDNA variation in East Africa unravels the history of afro-asiatic groups, Boattini et al.

4. Ancient west Eurasian ancestry in southern and eastern Africa, Pickrell et al.

5. Khoisan hunter-gatherers have been the largest population throughout most of modern-human demographic history, Kim et al.

6. Admixture into and within sub-Saharan Africa, Busby et al.

7. Genetic history of an archaic hominin group from Denisova Cave in Siberia, Reich et al.

8. Ancient gene flow from early modern humans into Eastern Neanderthals, Kuhlwilm et al.

No comments:

Post a Comment