Unless you’ve been asleep, you probably know by now that the Reich lab has come out with a paper that analyzes the remains of 4 individuals from western Cameroon, dating to 8,000 and 3,000 years ago (2 of each, with one of the older individuals yielding 18.5x coverage DNA!). The location and timing both matter.
This area of Cameroon is hypothesized to be the point of expansion for the Bantu migration. This expansion began about 3,000 years ago and swept east and south until the agricultural streams met back up in southern Africa.
Perhaps then the authors then “caught history in action” with a change between 8,000 and 3,000 years ago? No such luck actually. Here is the abstract, Ancient West African foragers in the context of African population history:
… One individual carried the deeply divergent Y chromosome haplogroup A00, which today is found almost exclusively in the same region…However, the genome-wide ancestry profiles of all four individuals are most similar to those of present-day hunter-gatherers from western Central Africa, which implies that populations in western Cameroon today—as well as speakers of Bantu languages from across the continent—are not descended substantially from the population represented by these four people. We infer an Africa-wide phylogeny that features widespread admixture and three prominent radiations, including one that gave rise to at least four major lineages deep in the history of modern humans.
Basically, just like elsewhere in Africa where the Bantu expanded, you see massive discontinuity in this region of Cameroon (the modern agriculturalists in the area are Bantu-speaking). If you have ever analyzed African genetic data, the lack of high magnitude structure of the Bantu over wide areas is pretty shocking. The reason there’s little structure seems to be two-fold
- Rapid population expansion, so not much time to accumulate distinct variants (you see this in Northern Europe too)
- Minimal admixture with local populations, at least until you get to modern-day South Africa (then there is an admixture cline with Khoisan)
Meanwhile, you have these zones of relic hunter-gatherers here and there. These samples seem to be one of those cases. I think it’s analogous to the fact that hunter-gatherers persisted in pockets for thousands of years after the initial arrival of Neolithic farmers in Europe.
There are two types of things you can take away from a paper like this. General insights. And specific details. The plot at the top of this post illustrates a model that they generated with these data. It seems quite clear that the details are not crisp, and subject to a further specific revision. But the general insights seem robust and extend what we already knew.
First, there were several human lineages that diverged 500,000 to 1 million years ago. In Eurasia, these became Neanderthals and Denisovans. In Africa, one of the branches led to what we call “modern” humans. But a variety of lines of evidence indicate that within Africa there were also highly diverged human groups, analogous to Neanderthals and Denisovans. One could call them “African Neanderthal” analogs. But within the context of this paper, they are “ghost archaics.” But those aren’t the only “ghosts.”
Extant human populations sample only a fraction of the “modern” family tree, which seems to have diversified from one of the African human groups 300,000 years ago or so.
There is now a fair amount of circumstantial evidence that Neanderthals mixed with an African lineage that is an outgroup to most other Africans and descended-from-Africans. Because of its size and warm climate, I believe that Africa was quite a good habitat for humans, and there were a variety of them across the continent. Though I don’t discount deep-time back migration of Neanderthal/Denisovan groups into Africa, I think due to the different population sizes it is probably more the case that Africans went into western Eurasia than vice versa. Additionally, Southeast Asia seems to be a good target habit for any African species due to similarities of biome (e.g., Sundaland).
Finally, there is the fact that it seems non-African ancestry is closest to the Mota sample, dated to 4,500 years ago in Ethiopia. This makes geographic sense, though I do wonder if this is an artifact of continuous gene flow back from Eurasia, as much as the likelihood that this is near the exit path of African humans.
What about the details of this paper? Look a the supplements and notice all the admixture graphs. There are lots of potential fits to the data, and more data will come in. The paper is clear to not put too much faith in one set of weights for gene flows, and different graphs might explain the patterns in the data. Additionally, a highly dense African landscape of hominins might exhibit lots of continuous gene flow and isolation by distance. There’s a lot more to learn. Nothing is being closed in this case.
1) I wonder if functional examination for phenotype related genes implicated in height in Biaka (Aka) and Mbuti will find anything unusual with these ancients, since some of the ancestry here is modeled as forming a clade with Aka. Or if these people will be perhaps more like a population that lacks the selective signals towards reduced height and size found in the Aka+Mbuti (even though much of the ancestry forms a clade with them).
2) It seems like it may be pointing to a Western Sahel-Sahara margin origin for the present day West African “Non-Basal” component. Which seems consistent with research pointing to this region for Pearl Millet and Sorghum?
Also -https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5809971/ – “The peopling of the last Green Sahara revealed by high-coverage resequencing of trans-Saharan patrilineages”. (I wonder if the ancestry which is associated with “Sub-Saharan (Western)” today is associated with “Saharan (Western)”, in the past?)
Razib,
Where do you think the people of Jebel Irhoud ( from little over 2 years ago) fit into this new phylogeny tree drawn by Reich et al.? Since their remains were dated as early as ~300,000 years ago, would they be located a little bit before Node #1 on that tree timeline wise, right before the great 4-way radiation outwards?
i wouldn’t put too much confidence in the date. i think jebel irhoud could be ‘ghost modern’
Good to see a sample of A00 in aDNA record, if you look at stuff like the ISOGG Haplogroup A tree the current modern A00 men all have very ‘long tail’ of equivalent SNP’s. At least with a new ancient sample they can probably split this block a bit and split out which of these SNP’s arose in last couple thousand years.
What this really makes like for me is that E came from Northern Africa or the Near East fairly recently down to SSA. It came in with a North African people which introduced agriculture to Subsaharan Africa and intermixed heavily with local foragers especially in West Africa, from where they spread. Similar to Neolithics in Europe, they largely replaced after a successful package and fast expanding population sufficiently adapted to the new habitat was established.
If they dig deeper they will see that Mota is to a large degree “a new arriver” in East Afria as well, there were older layers in the region with the first archaic Homo sapiens being most likely related to Jebel Irhoud and Florisbad-like, but more waves coming in.
So I think the map in Fig. 4 showing South Sudan/Uganda/Kenia as the epicentre for all human dispersions is plain wrong and cannot be based on anything we know for certain by now.
I wonder how the chances are to find remains older than 100.000 years in Egypt – obviously Egypt should have been inhabited by sapiens throughout most of its existence.
“But a variety of lines of evidence indicate that within Africa there were also highly diverged human groups, analogous to Neanderthals and Denisovans. One could call them “African Neanderthal” analogs. But within the context of this paper, they are “ghost archaics.” But those aren’t the only “ghosts.””
I think its pretty likely that the “archaic ghost” admxiture being represented by Iwo Eleru skulls:
https://www.bbc.com/news/science-environment-14947363
They just mention it offhand in the paper without naming it in the text. But what else should it have been, if this human variety was actually there?
The significant difference between the Central African foragers (“Pygmies”) and West Africans (“Negroid”) was always noted (there is no “one Subsaharan” population, adding Khoisan even less so) and one model to solve that involved admixture of a population related to the CA hunters with an early, more (Basal) Eurasian-like, shifted population, resulting in modern West African forms in West Africa and which spread with the Bantu expansion East and South. The local West African part of the fusion might still be divergent in a surprising way, which complicates current models without a fitting sample for comparison. Actually it could be said there is no proper fit for the Northern and the Western African component involved by now. Shum Laka is not right, but was just taken by the wave and surrounded.
I doubt they find haplogroup E before the transition with the Northern component in West Africa in higher frequency, in any West African population. East Africa, I’m not as sure, but most likely not as well, because like I said before, there is no definite proof for an East Afrian origin not just of E, but of fully modern human forms at all.
The authors know that of course and write correctly about “might” and “could” etc. Its no closed case yet.
Question: The Sahara has only served as a major demographic boundary for the past 5,000 years. How is it then that northern Africans are caucasoid? Was there a major displacement of the original N. A. population?
@Jamie: My understanding is that two things are to consider concerning North Africans:
1st I doubt there was ever a truly “Subsaharan”/Negroid majority in North Africa, because like I proposed, this human variety is the result of admixture with older human layers in West Africa.
2nd There were whole series of migrations into Africa, especially North Africa, pulling the regional population even more towards West Eurasians.
But the question is also interesting for what I speculated above, because the earlist sample we have is from Taforalt afaik, and Taforalt is already clearly more West Eurasian/Caucasoid in most respects, Wikipedia says:
“The Taforalt individuals show closest genetic affinity for ancient Epipaleolithic Natufian individuals, with slightly better affinity for the Natufians than later Neolithic Levantines. A two-way admixture scenario using Natufian and modern West African samples as reference populations inferred that the Taforalt individuals bore 63.5% Natufian-related and 36.5% West African-related ancestries, with no evidence for additional gene flow from the Epigravettian culture of Upper Paleolithic Europe. The Taforalt individuals also show evidence of limited Neanderthal ancestry.”
Now what I really wonder, and this is a question to all more knowledgeable people on genetic models, what about Taforalt being just closer to the source of the Northern component in Negroid, rather than gene flow from West Africa producing them?
Like in the paper they wrote:
“We can also obtain a good fit for the Shum Laka individuals in a less-parsimonious alternative model using three components, replacing the basal West African source with a combination of ancestry from inside the clade defined by the other West African populations and from a source entirely outside the West African clade (near one lineage that contributes to the Taforalt individuals) (Extended Data Fig. 5, Supplementary Information section 3). However, two-component models for the Shum Laka individuals that have the majority source splitting closer to other West or East Africans are rejected (Z = 7.1 and Z = 3.7, respectively).”
Why do they think that the model in Fig. 5 is less parsimonious? Its actually very close to the wave-model for fully modern Homo sapiens expansions into West Africa from North Afria/West Eurasia I would propose.
It makes much more sense everything considered, whereas the model they suggest as more likely is just a mess and doesn’t fit as good into the bigger picture.
This would just mean that Shum Laka already was “part modernised” by gene flow from Northern Africa, which they should have been, considering their age (!). The modern Negroid population would be the result of yet another, more closely Northern African related major incursion which finally spread E and brought it to the modern frequencies.
Its clear that we need this “Ghost North African” and “Basal West African” samples to come to better models, but still I think the second, alternative model they show makes sense, the first much less so.
@Obs
This debate’s been going on for a long time, and the Shum Laka data doesn’t do a lot to settle it. Regarding E, why have we found E-M35 alone in Morocco and the Levant from the final Pleistocene on – where was the other E? Why is E-M35 today frequent and diverse north of the Sahara, and commonly accompanied by North African/West Eurasian south of the Sahara, while other E branches are not? A recent North African origin is possible, but not without difficulties.
@Metalophias: To elaborate on my current understanding of the developments in Africa: First I assume there was a developmental centre in North Eastern Africa and/or the Southern Near East, from which pre-Basal Eurasians first and Basal Eurasians proper second expanded towards Africa. First North and East Africa, second Subsaharan Africa.
West Africa, the tropical regions in general, were never population centres and hotspots for human evolution, for developing fully modern Homo sapiens traits. That’s why human varieties like Iwo Eleru persisted there for until recently. The tropics became just recently more human-friendly by introducing adapted agriculture, plants and techniques. For foragers, that was a second choice habitat, always.
What I propose for North-West and West Africa is therefore that multiple waves came in, and Taforalt being caught during such a transition, with Basal Eurasian proper and E1b1b1 entering the scene and replacing the earlier wave (“Ghost North African”). This earlier population however did not disappear, but existed just South of it in the green Sahara.
When the conditions in the Sahara deteriorating while at the same time Neolithic technologies became fully established, this population which I would assume carried E1b1a. Under the pressure from the North and the climate, with new technologies at their hand, they pushed on South, meeting there with populations more similar to Shum Laka, with the mixture creating modern West Africans.
Now this doesn’t mean E1b1a wasn’t in Africa, even more South, long ago, but:
– They came from North Africa, probably starting moving South under pressure of the movement represented by Taforalt the first time.
– The major push took place when the Sahara became a worse place to live and through cultural and genetic adaptation (Malaria resistence!), including mixture with locals, adaptation of domesticated plants, they starting pushing into the tropical forests and beyond (especially Bantu expansion). Probably Shum Laka is like Taforalt, at a turning point, but while in the case of Taforalt we see the decisive result (Basal Eurasian taking over), in Shum Laka we see the start of the transition, first small scale admixture being probably present, some generations later the Negroid/Niger-Kordofan/Bantu expansion would have been in full effect.
So I’d assume E1b1a spread like R-V88, it came from a completely different region, was once more widespread in very different places, but after its initial expansion, the carriers were pushed on, until many landed in West Africa too. My assumption is, the history of E1b1a will prove to be very, very similar, just even much more successful in Subsaharan Africa.
But that the Bantu expansion started in situ, from within a tropical context like in Cameroon was always extremely unlikely for me. A more recent push after the adaptation in a borderzone is much more likely.
For getting the ancestors of E1b1a in North Africa, quite obviously, one would need to search before the more recent Basal Eurasian wave entered the scene, before Taforalt.
How do the Dinka population fit in the papers graph?
Special request for Matt: can you run these Shum Laka samples through your G25 PCA ringer and see if you can tease out anything interesting? I think David already added them to the spreadsheets.
Into Africa.