September 18, 2012

Out-of-Asia and Into-Africa (?)

The publication of version 2 of the Pickrell et al. paper on South Africa is as good an opportunity as any to discuss something anew something that I've been hinting at for some time now.

First things first: Pickrell et al. find West Eurasian admixture in the Hadza and Sandawe:
Both of these are consistent with west Eurasian (either European or, more likely, Arabian), gene  ow into these populations. To further examine this, we turned to ROLLOFF. We used Dinka and French as representatives of the mixing populations (since date estimates are robust to improperly speci ed reference populations). The results are shown in Supplementary Figure S22. Both populations show a detectable curve, though the signal is much stronger in the Sandawe than in the Hadza. The implied dates are 89 generations ( 2500 years) ago for the Hadza and 66 generations ( 2000 years) ago for the Sandawe. These are qualitatively similar signals to those seen by Pagani et al. [65] in Ethiopian populations.
The presence of West Eurasian ancestry in the Hadza and Sandawe was anticipated in my world9 calculator, where both these populations were shown to possess Caucasoid admixture entirely of the "Southern" component. This component peaks in Arabia, and is unaccompanied by any other type of Caucasoid element really only there. So, it is very likely that there was indeed such a migration into East Africa. What Pickrell et al. have added to our knowledge is that this migration is fairly recent.

Razib repeats one of his favorite analogies about events taking place in Africa after the pyramids were rising in Egypt. I will use a Greek epic analogy, by pointing out that at the time that Memnon the Ethiopian led his contingent to the aid of Troy, these events had not yet taken place.

Depictions of Memnon changed during classical antiquity, from a Caucasoid norm, as in the red-figure kylix on the left, to a more stereotypically African form by Roman times. This is sometimes taken as simply a consequence of the fact that the ancient Greeks were unfamiliar with African phenotypes, and changed their portraiture of Ethiopians as they became more familiar with them during Hellenistic and Roman times.

But, the very name of Aithiopes first attested in Homer (8th c. BC) attests to the fact that the Greeks were aware of what Ethiopians looked like, at least in terms of their dark pigmentation. And, there are depictions of Africans in classical art, as well as a famous quote in Herodotus which makes abundantly clear that he was aware of the physical characteristics of what we would call "Sub-Saharan Africans".

We don't only need to look at Ethiopia for evidence of the strange events that were taking place in Africa during classical antiquity. A great punch-in-the-face reminder of these events comes from the much later Greek author Pausanias who records that a statue of Athena he observed in Attica had blue eyes which he ascribed to the Libyan origin of her myth. How strange it seems to us that one would look to Africa for an explanation for the blue-eyed goddess.

Libya was of course, the ancient name for Africa, and especially Africa west of Egypt, what we might call Berber-land. Egypt was often reckoned by the ancient as part of Asia. In any case, Pausanias' strange assertion finds support in the Egyptian monuments that really do depict the ancient Libyans (=Berbers) as Caucasoid, and often lighter than Middle Eastern people. This would also accord with Coon's famous discovery of "Irish-like" Berbers among the Riffians; I often dismissed such assertions, but in a landscape of human prehistory that is getting stranger by the month, it is worth digging for gold nuggets in old texts.

A recent study claimed that there was back-to-Africa gene flow into Eurasia more than 12,000 years ago. On the other hand, both HAPMIX and StepPCO estimate the admixture in Mozabite Berbers as taking place ~120 generations ago, or, about 3.5kya assuming a generation length of 29 years as Patterson et al. (2012) do. I have observed that rolloff produces generally lower dates than these two methods, so I would not be surprised if that is the case here as well.

It seems that as recently as a few thousand years ago, West Eurasian populations were moving into Africa from both north and east. As Pickrell et al. have discovered, their eastern branch also contributed to South Africans, tagging along the dispersal of pastoralists from East-to-South Africa.

The big question is: did West and Central Africa escape this population movement?

I seriously suspect that it did not. I base that assertion on several arguments, of varying strength:

  1. Why would they? If they inundated East and North Africa, why would they not venture further?
  2. Living Sub-Saharan African farmers are not symmetrically related to West and East Eurasians: they are closer to the former. West Eurasian back-migration would explain this phenomenon.
  3. The Great Event in Sub-Saharan Africa was doubtlessly the Bantu explosion, and it is a curious coincidence that this took place precisely close to the time of these events
  4. The Iwo Eleru crania from Nigeria are of late Pleistocene age, archaic in character, and unlike modern West Africans. Something did happen in West Africa over the course of, say, the last 10,000 years
And, I always try to remind myself of the Kiffians and Tenerians. I have not seen any follow-up work on them, but if anyone has an ancient DNA lab, I'd think they would be prime candidates for a study.

Speaking of ancient DNA, this unexpected archaeogenetic study from the University of Khartoum, hints at important changes in Africa:

The area known today as Sudan may have been the scene of pivotal human evolutionary events, both as a corridor for ancient and modern migrations, as well as the venue of crucial past cultural evolution. Several questions pertaining to the pattern of succession of the different groups in early Sudan have been raised. To shed light on these aspects, ancient DNA (aDNA) and present DNA collection were made and studied using Y-chromosome markers for aDNA, and Y-chromosome and mtDNA markers for present DNA. Bone samples from different skeletal elements of burial sites from Neolithic, Meroitic, Post-Meroitic and Christian periods in Sudan were collected from Sudan National Museum. aDNA extraction was successful in 35 out of 76 samples, PCR was performed for sex determination using Amelogenin marker. Fourteen samples were females and 19 were males. To generate Y-chromosome specific haplogroups A-M13, B-M60, F-M89 and Y Alu Polymorphism (YAP) markers, which define the deep ancestral haplotypes in the phylogenetic tree of Y-chromosome were used. Haplogroups A-M13 was found at high frequencies among Neolithic samples. Haplogroup F-M89 and YAP appeared to be more frequent among Meroitic, Post-Meroitic and Christian periods. Haplogroup B-M60 was not observed in the sample analyzed.
I was reminded of it recently when this curious abstract came up, which I still believe is missing a zero somewhere, but these days you never know.


Evidence that Sub-Saharan Africans too have experienced gene flow from West Eurasians occasionally comes up, but formal tests of admixture, e.g., f3(Yoruba; San, French) usually do not achieve significance. But, we must be cautious: South Africans do appear admixed between San and East Africans, but this is a consequence of the fact that admixture is recent, leaving a trail of populations of varying East African ancestry, and the San still exist and can serve as one pole in a comparison of admixture.

David Reich has hinted at dual origins for West Africans. I am looking forward to learning what he means by it, but I would not be surprised if it involves admixture between a Eurasian-like population with a Palaeoafrican population of indigenous West African hunter-gatherers.

In any case, ex Africa semper aliquid novi even today. But, interdum, aliquid novum in Africam.

UPDATE: Pickrell and co-authors discuss their paper here.

13 comments:

pconroy said...

Interesting.

I see no reason why R1 or R1b carriers originating in the Mehrgarh area, and bringing Pastoralism to Northern Iran, and across into the Northern Fertile Crescent - leaving descendants among today's Armenians, Assyrians and Alawites - and then by boat to Cyprus to Western Europe, and/or via North Africa. Certainly Libyans were named "The Red People" by the Egyptians - owing to their skin pigmentation - and some had red or blonde hair and some had blue eyes.

There is very little trace of R1 or R1b (except for the V88 variety) in North Africa today, so I expect later male mediated migrations from the Horn of Africa (bringing E) and later from the Near East (bringing J), wiped out much of it.

Supposedly Pharoah Seti I had red hair, as did Ramesses the great, and his grandson King Tut may have been R1b.

rouge77 said...

It's not certain that the Aethiopia of Homer was intended to be situated in the same location as Roman Aethiopia - vaguely south, sometimes south-east of Egypt - nor that the vase makers would have thought so.

Ponto said...

Considering the likelihood of West Eurasian admixture from Arabia and the Levant into Africa, does that not make Admixture estimations of African admixture whether sub Saharan, East African and Pygmy/Khoisanid rather shaky for Southern and Western Europeans? The assumption is always that Europeans are the admixed and receiving population compared with sub Saharan Africans, and that the Near East and North Africa have contributed more to Europeans than the converse. Maybe all those assumptions are wrong.

Glossy said...

"...the Greeks were aware of what Ethiopians looked like..."

I'd say so.

Grey said...

"The Great Event in Sub-Saharan Africa was doubtlessly the Bantu explosion, and it is a curious coincidence that this took place precisely close to the time of these events"

My (very rough and possibly wrong) understanding is that there were breeds of large but unresistant to tropical diseases cattle in Africa that came in from the Levant and there were small but resistant native breeds in West Africa. If so then it seems to me the successful cross breeding of these breeds might create a large and resistant breed and the neccessary foundation for the Bantu expansion.

The Zebu breeds could have spread naturally from tribe to tribe from the north or north-east towards West Africa somewhere adjacent to where the disease resistance was needed and the cattle bred with the resistant breeds.

"David Reich has hinted at dual origins for West Africans."

Alternatively outsider traders of some description could have brought Zebu cattle down the Atlantic coast or over ground from the north following the trade routes, particularly the gold route.

http://en.wikipedia.org/wiki/File:Niger_saharan_medieval_trade_routes.PNG

http://www.metmuseum.org/toah/hd/gold/hd_gold.htm

"but formal tests of admixture, e.g., f3(Yoruba; San, French) usually do not achieve significance"

If any of this is related to the gold trade (and maybe it isn't) the admixture itself may not have been as significant as the import of external "pots" simply due to weight of numbers (e.g. British in India)

or

maybe it's concentrated among populations along those old trade routes so that would be the place to look?

eurologist said...

Considering the likelihood of West Eurasian admixture from Arabia and the Levant into Africa, does that not make Admixture estimations of African admixture whether sub Saharan, East African and Pygmy/Khoisanid rather shaky for Southern and Western Europeans?

Probably. I wonder if Dienekes could do a qp3Pop study using SW Asian (either real or zombie) as "C," and South Asian as "B."

On the one hand, I think it's going to be really hard to detect truly ancient SW Asian signatures - on the other, I think it's too early to give up.

dalouh said...

"There is very little trace of R1 or R1b (except for the V88 variety) in North Africa today, so I expect later male mediated migrations from the Horn of Africa (bringing E) and later from the Near East (bringing J), wiped out much of it."

huh ? R1b is averaging 10 % in coastal North Africa...the Mesolithic females of S.W European origin found at Taforalt,did not cross over to Africa without male companions..

and forget about the horn of Africa origin of E..

modern day Berbers of Tripolitania

http://s1.zetaboards.com/anthroscape/topic/4873913/1/#new

Roy said...

@pconroy

This also suggests a sizable Nordic migration as far as the Mahreb in modern times, as the blue-eyed mutation has been dated to about 8000 ybp. See Peter Frost's "European hair and eye color
A case of frequency-dependent sexual selection?"

rouge77 said...

When it comes to red hair of the 19th dynasty Pharaohs, it could have easily come from the Berbers for example.

The problem with the family characteristics of the 19th dynasty pharaohs is that we can't know how well they represented the overall population of the time in the Nile delta where the family seems to have come from.

terryt said...

"Alternatively outsider traders of some description could have brought Zebu cattle down the Atlantic coast or over ground from the north following the trade routes, particularly the gold route".

The arrival of zebu in Africa is usually assumed to indicate a connection with India, where humped cattle appear to have originated.

"does that not make Admixture estimations of African admixture whether sub Saharan, East African and Pygmy/Khoisanid rather shaky for Southern and Western Europeans?"

Good point.

Grey said...

rouge77
"The problem with the family characteristics of the 19th dynasty pharaohs is that we can't know how well they represented the overall population of the time in the Nile delta where the family seems to have come from."

I think this is possibly a key point. A lot of known Egyptian history revolves around mercenaries rebelling and setting up a dynasty. So King Tut may have been descended from Libyan or Sherden mercenaries (or something) rather then being representative of the majority Egyptian population.

.
terryt
"The arrival of zebu in Africa is usually assumed to indicate a connection with India, where humped cattle appear to have originated."

Yes the origin of zebu into Africa itself. I'm wondering about possible cross-breeding with the tsetse resistant native west african cattle at some much later point.

If that happened it could have happened indigenously. However if it was externally triggered then i'd suggest the old gold routes as a possible vector and so if you were looking for evidence of eurasian admixture i'd look along those old gold routes also or cattle cultures near the gold routes.

terryt said...

"I'm wondering about possible cross-breeding with the tsetse resistant native west african cattle at some much later point".

Almost immediately. Daniel Bradley (et al.) has found the Y-chromosome of African humpbacked cattle is zebu while the mt-DNA is local.

http://www.google.co.nz/url?sa=t&rct=j&q=&esrc=s&source=web&cd=2&ved=0CDAQFjAB&url=http%3A%2F%2Fwww.sist.sn%2Fgsdl%2Fcollect%2Fpubli%2Findex%2Fassoc%2FHASHb1dd%2F7c484906.dir%2Fdoc.pdf&ei=BhhcUJ3UOMfdigeT04CoCg&usg=AFQjCNGc6GgBnwn-caDNZ2b6Vt41Y1ikDg

And:

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1208036/

"Substantial differences in the distribution of alleles at 10 of these loci were observed between zebu and taurine cattle. These markers subsequently proved very useful for investigations of gene flow and admixture in African populations. When these data were considered in conjunction with previous mitochondrial and Y chromosomal studies, a distinctive male-mediated pattern of zebu genetic introgression was revealed. The introgression of zebu-specific alleles in African cattle afforded a high resolution perspective on the hybrid nature of African cattle populations and also suggested that certain West African populations of valuable disease-tolerant taurine cattle are under threat of genetic absorption by migrating zebu herds".

"If that happened it could have happened indigenously".

Seems it did so.

"However if it was externally triggered then i'd suggest the old gold routes as a possible vector and so if you were looking for evidence of eurasian admixture i'd look along those old gold routes also or cattle cultures near the gold routes".

No need to look for anything too complicated. There was no need to bring in huge numbers of cattle. Just a few bulls would do.

terryt said...

"huh ? R1b is averaging 10 % in coastal North Africa.."

But is that R1b actually all R1b-V88? Most of the R1b through Africa is, as pconroy said, R1b1c-V88. It is one of three basal branches within R1b, the others being R1b1b-M335 (found in Turkey) and R1b1a (found as far east as the Hazaris of Afghanistan, who were possibly originally from the Iranian Plateau). That spread of R1b into Africa from the Levant/Anatolia/Iran could be quite ancient, certainly long before any R1b had entered Europe.

"and forget about the horn of Africa origin of E.."

I'm beginning to agree with that. It's from further west. As suggested in the extract:

"The area known today as Sudan may have been the scene of pivotal human evolutionary events, both as a corridor for ancient and modern migrations, as well as the venue of crucial past cultural evolution".

The 'Sudan' would include the Bahr el Ghazal which forms a single eco-region along the margin between forest and savanna all the way west to Cameroon. Someone commented recently that the modern Y-DNA looks to have first appeared in Cameroon. And from looking at E's modern distribution we could claim that it too could easily have first appeared there. E2 is widespread but it owes much of its spread to the Bantu expansion. It is not common in East Africa. There its place is taken by E1.

E1a, though, is basically a West Sahelian haplogroup, however it has made it across the Mediterranean to Italy at some time. I have no idea where E1b2 has been found but E1b1a looks also to be basically a West Sahelian haplogroup. So that leaves just E1b1b as an East African haplogroup. The Y-DNA involved in the OoA would then be CF and D. but can we really believe that haplogroups CF and D did not leave Africa until the E1b1a/E1b1b split? Surely that is likely to be merely the last in a process that has been continuous since A first appeared, not a single eruption. To me the most obvious interpretation of the modern haplogroup distribution is that C, D, E and F simply represent the geographic points of CT's original expansion. In other words they were spread through Dienekes' postulated North Africa/Arabia/Levant region. The lines managed to escape beyond that region when the climate permitted.