Ancient genomes from S Africa pushes modern human divergence beyond 260,000 years ago https://t.co/HaO7GO2KAs WANTED TO GO TO SLEEP!!!
— Razib Khan (@razibkhan) June 6, 2017
In response to a little bit of fatigue at the constant stream of ancient DNA, John Hawks digs the knife in a bit deeper. There’s more. Since Hawks is co-author with Lee Berger of Almost Human: The Astonishing Tale of Homo Naledi and the Discovery That Changed Our Human Story, I’m hoping for something related to naledi. But barring that, there is still lots to come down the pipeline (Alexander Kim has a Twitter account worth following to keep up on this line of research).
But I thought I’d enter something else into the record about what we’re starting to understand about the origins of modern human genetic variation.
- The “Out of Africa” movement is a real thing. That is, ~50,000 years ago a population seems to have diverged from a broader pan-African group, and replaced archaic hominins across Eurasia (with some admixture at low levels), and then pushed the boundaries of our genus to Oceania and the New World. This does not mean that there were not earlier “Out of Africas.” Just that the dominant signal of variation is due to the pulse that swept out ~50,000 years ago.
- Within Africa the story is different and more complicated. The expansion of anatomically modern humans out of Africa occurred relatively rapidly from a small founder population (within an order of magnitude of ~1,000 individuals seems correct). The archaeology and genetics are in pretty good alignment. But within Africa it looks like the lineages which led to modern humans are much deeper, and preserve structure that may be hundreds of thousands of years old. Just as we see admixture events giving rise to new lineages outside of Africa over the last 50,000 years, the same dynamic probably applied to within Africa far earlier.
One way to think about it is that the old “Africa Eve” model is very useful for the 10,000 year period around 50,000 years ago. And, it applies to non-Africans.
The story within Africa though may be more like the old multi-regionalist model, though with stronger biases of gene flow between populations, so that at one time one lineage may be preponderant. Over the last 10,000 years the expansion of certain populations within Africa though (in particular Bantus) has collapsed a lot of the deep structure, and now Africa resembles Eurasia much more (with some Eurasian back-migration too).
Note: I am talking about “modern human genetic variation” because I am starting to think talking about “modern humans” obscures far more than it illuminates.
Cushitic nomads of the southern Cushitic branch wer present in northern Kenya around 4580 ybp and present in the Cape region of South Africa some 2000 ybp with cattle and the Somali donkey and mat tents. They carried lineages E-123 EV22 and minor T1a and mtdna N and lots of I and some roa and L3 they were 60% AEA and 40% Basal and UHG Eurasian
Exciting stuff.
Razib, I’m so glad you’re around to help me make sense of all these new discoveries.
The new discovery of archaic homo sapiens fossils in Morocco dating to 300,000 years ago makes an interesting addition to this story.
https://www.theguardian.com/science/2017/jun/07/oldest-homo-sapiens-bones-ever-found-shake-foundations-of-the-human-story
With respect to both this post and your previous one, that new skull from Morocco seems to add weight to the evidence of modern human admixture in Siberian Neanderthals from a group that diverged from us ~300kya. So I guess we’re just the most recent and possibly most successful OoA event.
If modern humans emerged and started splitting up and even leaving SSA around 300kya, and Neanderthals split from us ~500kya, it seems like splits between different groups of modern humans are deeper than the split between Neanderthals and modern human was when modern humans emerged. That’s not phrased well but I hope it’s clear?
I think Dienekes’ idea that there may have been a back migration into Africa around 300kya bears greater consideration too.
“Australian Aborigines seem to have been resident in their current locations for ~50,000 years, but this seems the exception, not the rule. Do we really think that the ancestors of the Bushmen were living in southern Africa for five times as long as Australian Aborigines?”
Pama–Nyungan languages seem to be as young as 5,000 years old – coincident with dingoes arriving in Australia maybe, so I wouldn’t take for granted that there weren’t these big sweeps in Australia either. We just have a hard time detecting them now is all.
“I wouldn’t take for granted that there weren’t these big sweeps in Australia either.”
Assuming that were true, and it is contradicted by Eske Willerslev’s findings, it would still mean that Aboriginal people were genetically isolated, without population turnover, for a very long time.
Re: Aboriginal Australians and Malaspinas 2016, there are lots of parameters and modelling assumptions in Malaspinas 2016’s tree structure (Figures 3 and 4), including how they model population history between Europeans and East Asians and Africans which sets the clock on divergence time for Australians.
I don’t really see anything there that would give strong assumptions about time depth and a lack of Holocene or very late Upper Paleolithic admixture from Sahul (and then for Sahul from East Asia) without adna. Some degree of broad continuity in region over 40 kya seems at least to be likely though (also true for Europe, despite glaciation being almost certain to have disrupted continuity to a greater degree).
There’s an interesting mention in their supplement of “Note that we estimate a very recent East Asian contribution of 18% into north-eastern Australia (95%CI: 9-29%), consistent with other analyses (S05), but we have limited power to estimate the timing of such recent events” . About 10 kya, probably something to do with Pama–Nyungan languages, but seems to have not percolated much autosomal turnover by the time it gets to SW Australia. NE Australia is also represented as receiving a 41% New Guinean pulse at around 15-20 kya which SW Australia is excluded from. Both of these will be easier to check properly with adna; like in the case of Southern Africa, ancient samples may constitute a new floor.
SW Australia became geographically pretty isolated relatively early due to climate change and being surrounded by ocean and country with a harsh climate – it’s an isolated pocket of, in the deep SW where I was born, a lot of dense temperate forest; very tall hardwood eucalypts (not Redwood tall, but tall, like 200m or so going straight up – you have to travel to the eastern side of Australia to find trees that tall anywhere else). Also, the SW has some unique extant (and some recently extinct, unfortunately) fauna that are not found anywhere else in Australia.
Plus the people there seem to have avoided exogamy due to cultural reasons. Aboriginal men generally practised some pretty spectacular forms of circumcision throughout Australia (I won’t go into details, but it qualified as pretty full-on genital mutilation in very many cases); the Noongar people of the SW are the only Aboriginal people I know of who did not practise any form of circumcision, and they resisted pressure from contacts with other Aboriginal people to do it. Hence Eske Willerslev’s finding that they became genetically isolated pretty early on.
So, knowing the country and the people first hand, it’s no surprise to me that the SW would be excluded by those events, although they spoke Pama-Nyungan languages (Nyunga = Noongar), which I presume they got largely by cultural diffusion through trade contacts with other groups. The seemingly rapid expansion of Pama-Nyungan languages appears to have been accompanied by new stone tool technology, which also seems to have facilitated survival in some of the more harsh desert environments, so it seems to have been associated with a general population increase, as people could expand and inhabit a lot of areas that were previously very difficult to survive in.
I might have exaggerated tree height through mixing up my units – more like 80 metres. Still notably very tall for Australian trees. Eucalyptus diversicolor, if you want to look them up.
“NE Australia is also represented as receiving a 41% New Guinean pulse at around 15-20 kya which SW Australia is excluded from.”
Well, the mixture should be expected, as 18,000 years ago, the Sahul Shelf was exposed as dry land. You could just walk from New Guinea to Australia. But the lack of spread of that ancestry shows that any more recent language sweep did not occur by mass migration of people.
“But the lack of spread of that ancestry shows that any more recent language sweep did not occur by mass migration of people.”
No, it just says that any mass migration could not have originated in a population like present day NE Australia.
Rick: “But the lack of spread of that ancestry shows that any more recent language sweep did not occur by mass migration of people.”
Well, the thing is, as I understand it, that the model they have can work out that NE Australia has these extra pulses of ancestry because they can check NE Australia’s position relative to SW Australia, and see where they are more related to East Asia / New Guinea compared to if there was a simple bifurcation between NE and SW only (on their choice of statistic, in this case Site Frequency Spectrum).
But for SW Australia, you can’t really calculate the floor, because they are the floor. It’s sort of an assumption.
In every other case we’ve ever had, ancient dna has shown that modern populations we might assume to be an unadmixed floor actually show admixture relative to ancients and that admixture is more profound than expected and divergence older/deeper (Sardinia, Khoi San, etc.).
This being the case, I don’t have a lot of confidence that language shift among hunter gatherer populations without genetic change at the end of the cline of expansion would, in this instance, stand to ancient dna (if it was ever obtained). Not impossible that an autosomal shift could be so watered down by the end of the cline that it is nothing, but I would not bet on it.
I do not have the time at this moment to properly reply, and (as we all know) the genetic history of Australia is far from complete, but it still looks like there are very major genetic distances between groups that share quite close linguistic ties.
Perhaps with more sampling, the nature of this will be more clear. I hope so.
It is fascinating.
One thing I just stumbled across by the way – Periplus of the Erythraean Sea describes the East African coast as populated by broadly Afro-Asiatic looking folks. So in terms of East African and Eurasian admixture in the San, I think that’s a good candidate.
One thing I just stumbled across by the way – Periplus of the Erythraean Sea describes the East African coast as populated by broadly Afro-Asiatic looking folks. So in terms of East African and Eurasian admixture in the San, I think that’s a good candidate.
gold!
i pointed out a while back that ethiopians in the iliad are not described as african looking. may be that dark age greeks didn’t know about ethiopians. or, it may be that admixture was only recent and uneven then….