Arabia as Africa-across-the-sea

In antiquity ostriches and lions roamed the Syrian desert. The cheetah even still clings to a tenuous existence in the fastness of the central Iranian desert. The point being that the new finding of African modern human remains on the southeast fringe of Arabia ~85,000 years ago shouldn’t be too surprising. Old modern(ish) looking humans date to 73,000 years before the present in Southeast Asia. Modern-like ancestry can be found in eastern (Altai) Neanderthals dating to ~100,000 years ago. And the earliest humans may have arrived in Australia 65,000 years ago.

These dates are important because the genetic results indicate that much of the population divergence of modern Eurasian, Amerindian, and Oceanian peoples dates to the period between 50 to 60 thousand years ago. This was the classic epoch for the emergence of “behavioral modernity,” and the older models of “Out of Africa” which posited a rapid explosive demographic growth after a punctuated speciation even in East Africa ~60,000 years ago.

Today with remains such as Ust’-Ishim man, we can peg the admixture of Neanderthal into modern Eurasians 52,000 and 58,000 years ago. About the same period that the preponderance of the ancestry of modern Eurasians and peoples of Australia and the Americas expanded across the world, as noted above.

Most peoples in Western and Southern Eurasia also have substantial ancestry from another group which doesn’t seem to have much Neanderthal ancestry at all, the “Basal Eurasians” (BEu). This population obtained its name from the fact that it was hypothesized to have diverged from the common ancestors of northern Eurasians (the Pleistocene peoples of Europe and Siberia), eastern Eurasians, the ancestors of the Amerindians, and Oceanians, before these groups moved on and then separated (i.e., proto-Melanesians are closer to Pleistocene European hunter-gatherers than they are to BEu). These facts suggest proto-BEu was a distinct population >60,000 years ago.

The maximum range of Neanderthals

 

Because of the distribution of Neanderthal admixture across so many groups relatively evenly it probably came from a single major admixture event. Geography tells us that the most likely area of this admixture would be somewhere in the northern area of West Asia.

This implies that BEu was probably resident in the southern area of West Asia, and possibly into North Africa. We do not have any samples which are “pure BEu.” Ancient agriculturalist samples from the western Near East and the eastern Near East are high in BEu ~10,000+ years ago, but these populations are still substantially mixed with a population with affinities to Mesolithic Western European hunter-gatherers (WHG). Fu et al. 2016 use a Pleistocene transect to infer that this affinity between Near Easterners and Europeans dates to the period after ~15,000 years before the present. I presume that this late Pleistocene period was when BEu was admixed away as a pure population by an expanding hunter-gatherer culture with a nexus in Southeast Europe and into Anatolia and the trans-Caucasian region.

The recent Arabian find makes sense I think in the context of BEu and other such populations, which had diverged from the Africa metapopulation ~100,000 years ago, but had not pushed further north and east, and so mixed with Neanderthals.

But what about the older modern human remains which are showing up in eastern Eurasia? I think it is entirely likely that these populations left only a little bit of an imprint in modern groups. A paper from a few years back reported having detected such an admixture in Oceanians. The first ancient genome we have from eastern Eurasia >60,000 years ago that is from a modern human will probably yield much more satisfying results.

The big dynamic looming over the likely existence of anatomically modern human range on the edge of Africa in Arabia is that for several hundred thousand years modern humans existed within Africa as a metapopulation. The proto-Out-of-Africa population can only be understood as part of this broader metapopulation. ~100,000 years before the present humans, inclusive of Neanderthals, Denisovans, and modern humans, our species was probably defined by a set of distinct metapopulations. We know that there was gene flow between these metapopulations, but the strong evidence of purifying selection of Neanderthal and Denisovan ancestry in modern human genomes tells us that this gene flow was minimal enough that biological incompatibilities were beginning to build up and the groups were on their way to speciation as defined by the biological species concept.

There is no evidence of this between any modern populations, even the most diverged (e.g., the Khoisan, who carry Eurasian and African agriculturalist genetic material). This means that within the modern human metapopulation gene flow was sufficient to prevent incompatibilities from developing due to isolation. That being said, with the oldest (proto-)modern human skull dating to ~300,000 years, and likely discernible population structure between various African lineages going beyond 200,000 years ago, there are lots of distinct modern human groups with very long histories within Africa and on its periphery.

The earliest point that you could probably say non-African humans diverged from any African (Sub-Saharan) populations is ~100,000 years ago (and this is probably a bit too generous). A conservative estimate would suggest that modern human lineages were emerging within Africa 200,000 to 300,000 years ago. So most of modern humanity’s existence has been within Africa.

The non-African populations descend from a group which underwent a period of reduced population size vis-a-vis all the African groups. But one thing I think is important to remember is that this was probably not exceptional. We know now that over the past 5,000 years African population genetic structure has been reshaped by events such as the Bantu expansion. But there were surely small and marginal groups with low effective population sizes within Africa that either went extinct or were absorbed by other populations.

The difference in the non-African population is that it was on the edge of the modern human range, and likely occupied territory that was relatively isolated from other modern humans due to the dry nature of the Sahara during most of the Pleistocene. This prevented its absorption into more numerous groups of modern humans further south and to the west. And the strong cultural and genetic barriers with the Neanderthals probably limited gene flow as well.

But even in the inclement conditions of North Africa and West Asia for most of the past 100,000 years, modern humans may have had a larger effective population size than archaic Eurasian hominins. And with this larger effective population size, one can imagine that greater cultural creativity and genetic robustness to dynamics such as population declines gave the modern humans a long-term advantage. In this context, the existence of modern human remains in a diverse array of places across warmer areas of Eurasia before 60,000 isn’t that surprising. And, the demographic wave that swallowed Neanderthals and Denisovans probably swallowed the earlier modern humans who ventured into eastern Eurasia before 60,000 years ago!

0

8 thoughts on “Arabia as Africa-across-the-sea

  1. It seems to be most likely that the most significant modern human expansion started from a place which is more likely in Eurasia rather than SSA. The modell of a massive back migration gains more and more credibility.
    Also, hows your take on the origin of DE?
    And the possibility of much higher Neandertal admixture masked by back migration to Africa? Possible or not?
    If a possibility, how about modern behaviour postdating the hybridisation?

    0
  2. There seems almost a fractal pattern here of metapopulations overrunning each other in turn.

    1000s of years: examples. Bantu’s expand in Africa, pastoralists overrun farmers who overrun hunter gatherers
    100k years: modern humans overrun old modern humans, also over run Denisovans and Neanderthals

    Makes one wonder if the pattern goes farther in both directions. homo erectus overruns other hominans at 1M year level. And on shorter time period various groups overrun each other at 500 year level.

    And I wonder if this is common across all mammals. Which should be testable, especially with genomics and aDNA.

    0
  3. Fascinating.

    And with this larger effective population size, one can imagine that greater cultural creativity and genetic robustness to dynamics such as population declines gave the modern humans a long-term advantage… And, the demographic wave that swallowed Neanderthals and Denisovans probably swallowed the earlier modern humans who ventured into eastern Eurasia before 60,000 years ago!

    1) Were these earlier modern humans behaviorally modern or only anatomically so?
    2) Presumably they had not generated the greater cultural creativity and genetic robustness because of small effective population size?

    0
  4. re: back-migration. there is some. but i don’t think massive. SS-african distance from eurasians is proportional to recent admixture. unless the back-migration is TOTALLY even or TOTAL that doesn’t make sense.

    Also, hows your take on the origin of DE?

    uniparental is weird.

    If a possibility, how about modern behaviour postdating the hybridisation?

    well ‘modern behavior’ clearly postdates modern humans. but what you’re talking about seems coincidence with the absorption of neanderthals denisovans and probably other humans.

    Makes one wonder if the pattern goes farther in both directions.

    i think you are on to something.

    remember, neanderthals descend from post-erectus africans who presumably replaced relic eurasian post-erectus hominins.

    1) Were these earlier modern humans behaviorally modern or only anatomically so?
    2) Presumably they had not generated the greater cultural creativity and genetic robustness because of small effective population size?

    anatomically.

    as for cultural creativity etc. who knows. but i think sufficient census pop size is clearly necessary, but not sufficient.

    didn’t want to get into this in the piece but historically and from cultural ev. theory frontiers are innovative. perhaps that’s why africans on the margin of african range expanded?

    0
  5. What I meant is that it probably was no coincidence, but causal. The admixture event with Neandertals led to a behavioural change and this change spread back into Africa.
    On the way, by mixing with different Basal Eurasian and African populations, the new Neandertal-admixed population was diluted.

    I think most of us agree by now that Mota isn’t the unadmixed SSA?
    What if the replacement was huge and just relic groups were absorbed, of which some were local African archaic hominids, resulting in the genetic variation pattern we see today? As far as I know, which is not that much on that matter, nobody can falsify that hypothesis by now? So it should be considered as an option and tested with more material available.

    Can anybody be sure what’s the bottom line of sapiens vs neandertal? Isn’t it just that most researchers assume that SSA have no significant Neandertal admixture because of the dominant recent OOA model?

    Where do you think is the limit for retrieving ancient DNA? Will it be possible to analyse 200.000 year old remains? Or older? Is it impossible or just unlikely?

    0
  6. The admixture event with Neandertals led to a behavioural change and this change spread back into Africa.

    this is an old hypothesis. cochran et al. proposed it.

    the issue i have is that small admixture events here and then seem common. so hard to establish this was obv. causal cuz it’s so exceptional. the siepel data indicates that the reciprocal occurred to altai neanderthals. and there was probably an earlier one into neanderthals too.

    I think most of us agree by now that Mota isn’t the unadmixed SSA?

    there is no the SSA. there are lots of types.

    i think we can agree that a small % of back-migration across africa seems likely in the post-100K interval. i said that above. but a lot? skeptical.

    What if the replacement was huge and just relic groups were absorbed, of which some were local African archaic hominids, resulting in the genetic variation pattern we see today? As far as I know, which is not that much on that matter, nobody can falsify that hypothesis by now? So it should be considered as an option and tested with more material available.

    yes. but the problem is if it was back-migration from the same OoA pop then it should show the same bottleneck in PSMC. it doesn’t. it doesn’t seem like africans share that demographic history. that could be because of admixture…but the the smaller the % of substrate admixture the bigger the signal of the OoA bottleneck should be.

    Isn’t it just that most researchers assume that SSA have no significant Neandertal admixture because of the dominant recent OOA model?

    to some extent, yes. but i outlined the issues above. the admixture has to be even. there’s no variance of allele sharing with neanderthals in africans that isn’t accounted for by post-10K BP backflow.

    Where do you think is the limit for retrieving ancient DNA? Will it be possible to analyse 200.000 year old remains? Or older? Is it impossible or just unlikely?

    i believe it is possible…seeing as 430K BP proto-neanderthals were analyzed

    https://www.scientificamerican.com/article/oldest-ancient-human-dna-details-dawn-of-neandertals/

    and here’s the 700K BP horse genome from 5 years ago
    https://www.wired.com/2013/06/ancient-horse-genome/

    0
  7. From Prufer (2017): “in contrast to earlier analyses of chromosome 21 data for the European Neandertals (13), analyses of the full genomes suggest that the putative early modern human gene flow into Neandertals occurred prior to the divergence of the populations ancestral to the Vindija and Altai Neandertals ~130-145 thousand years ago”.

    One has to wonder if this flow was the same one responsible for Neandertals and modern humans having relatively similar mitochondrial DNA, whatever the direction of flow.

    0
  8. Razib: remember, neanderthals descend from post-erectus africans who presumably replaced relic eurasian post-erectus hominins.

    Though on that point, Reich’s proposed a model in his opus in which AMH ancestors evolved in Eurasia roughly between 1.8mya prior to return to Africa 300kya and then an out of Africa pulse (or rather a successful and replacing ooA pulse) at just before 50kya.

    This is order to more parsimoniously explain the adna structure (the “Africans are a subset of modern+adna Eurasian variation” theme of Dienekes) albeit he acknowledges lack of adna from non-AMH African fossils.

    Sort of verging into off topic here, I was thinking about how what that would then mean, morphologically and in terms of environmental selection. If the AMH lineage were evolving out of Africa, and largely this is going to mean China, SE Asia, India as well as to a degree in cooler climates, then when proto-AMH return to Africa 300kya, they might be more likely to harbour and maintain at low levels “Eurasian” variants that would allow rapid adaptation to the Eurasian climate later in history when the 50kya pulse ooA happens. Eurasia has a somewhat more variable climate than Africa(?), perhaps useful for our evolution into a fairly cosmopolitan species with a good amount of variation to evolve and re-evolve into forms adapted for climatic areas?

    In that light, a relatively low degree of introgression of variants from Neanderthal that we find might be more explicable – if proto-AMH and then AMH already held variants adapting them to Eurasian climate, then Neanderthal didn’t really have a morphological “card” to play that would prevent replacement by this Homo species if they had even a little mental/behavioural/resource consumption edge.

    (Though in other places, key variants may still have been lacking, because they were very specific to regions, like the high altitude adaptations, needing introgression from what in Reich’s model would be a specific population within the overarching set that gives rise to the back to Africa proto-AMH).

    Plus it might also help re: time some of variants might have needed to be probable to evolve in low population size scenarios.

    0

Comments are closed.