Wednesday, August 28, 2013

Homo Erectus Out Of Africa Wave Happened All At Once

Old Homo Erectus Dates In China Confirmed

Newly refined age estimates for the oldest hominin sites in China, establish that Homo Erectus spread at about the same time to Java Indonesia (1.9 million years ago), to Northern China (1.7 million years ago) and to the Southern Caucasus mountains and to a wider geographic range within Africa, all at around 1.7-1.9 million years ago from the previous core range within Africa.  The evidence for the oldest H. Erectus anywhere in Africa is a bit older.

Age estimates at the scale are about +/- 100,000 years in accuracy, and the thinness of the data in this time frame also suggests that there is a certain amount of statistical variation due to the random sampling of existing data points from all available data points for undiscovered sites of the same type.

These factors combined, informed by data points from modern human expansion and expansion of other species on how long it took those species to disperse, make the precisely differences between the ages of the various early non-core H. Erectus site dates small enough to be insignificant and suggest as single wave of H. Erectus expansion both Out of Africa and within Africa.

Implications

This new data largely refutes the alternative hypothesis that H. Erectus expanded Out of Africa in a staged migration that reached some parts of Eurasia much later than others.

Of course, expansion "all at once" is a relative thing.

It could truly mean a true single wave of expansion (and honestly, that is what I believe is the most likely scenario), but several successive waves of expansion 10,000-20,000 years apart, of the kind that may have happened in the modern human process of expansion "Out of Africa" would be indistinguishable from a single wave of expansion in the Homo Erectus case.

The new data simply shows that expansion from an original source territory to the entire ultimate Homo Erectus range probably took place over a period that was probably shorter than 200,000 years, contrary to earlier theories based upon incomplete or less accurate data from China that had suggested that there might have been a pause of 400,000 years or more before Homo Erectus spread from SE Asia and the Southern Caucasus mountains to China.

Open Questions

These days, however, the really hot issues in the prehistory of H. Erectus relate to the tail end of the story, rather than the beginning.

When did H. Erectus go extinct and why?  Was H. Erectus the source of the Denisovan genome or the H. Florensis species, and if not, as the Denisovan genome seems to suggest, what hominin species were each of these associated with, how did they end up where they did, and when?  In particular: Does the Denisovan species have a relationship to early archaic hominin evidence in China?  Did the Denisovan species replace or coexist with H. Erectus, and if so, when and where? (The distribution of lithic tools in Asia suggests that there might have been a limited replacement or co-existence of H. Erectus and Denisovans in Zomia, Malaysia and Indonesia, a path that connects most of the dots between the Denisovan cave in Southern Siberia and the island of Flores, but the more modern lithic tools are recent enough to be attributable to Homo Sapiens as well).  During which time frames, if any, did H. Erectus co-exist with modern humans?  Why isn't there a discernible trace of H. Erectus in most modern humans in Asia?  How much did H. Erectus evolve after the species left Africa?  Did H. Erectus evolve into other hominin species outside of Africa, and if so, which ones?  What impact, if any, did the Toba explosion have on H. Erectus

The picture is quite unclear for the events of the time frame from sometime after 150,000 years ago (i.e. after the period where there are no potentially more modern hominin remains) to 50,000 years ago (i.e. the period when modern humans were the undisputed dominant hominin species of Asia barring some currently unknown relict populations of archaic hominins and H. Florensis).  Any sub-periods for Asian hominin populations from ca. 1,900,000 years ago to 150,000 years ago are also quite fuzzy.  This was a period 1.5 million years plus long that was quite static relative to that of Neanderthals or modern humans over the similarly long time frames, and even relative to non-African H. Erectus populations.

Background Context

The Basic Story.

The H. Erectus who left Africa about 1.9 million years ago have not been important ancestral genetic contributors to modern humans.  It is likely, however, that Neanderthals and modern humans are among the hominin species derived from H. Erectus.

All modern humans are predominantly descended, genetically, from common modern human African ancestors.  Those modern humans evolved ca. 250,000 years ago, or so.

All non-African modern humans are predominantly descended from one or more groups of modern humans who left Africa much later. There is academic debate over when the first sustained modern human presence Out of Africa arose with the earliest estimates being around 130,000 years ago and the youngest being around 50,000-60,000 years ago with earlier Levantine modern human remains attributed to an "Out of Africa wave that failed" by supporters of that theory.  As in the case of H. Erectus there is evidence that the "Out of Africa" migration by modern humans may have coincided with a range expansion of modern humans within Africa (roughly speaking around the time the Paleo-African populations like the Khoisan and Pygmy populations broke off from other African populations).

New archaeological data and increasingly refined understandings of population genetic data increasingly favors Out of Africa dates that are closer to the older date in this range, although the appearance of a younger age in some respects requires a fairly complex narrative of human expansion beyond Africa to fit the data precisely.  For example, a population bottleneck for the Out of Africa population, or a second wave of expansion after a first less numerous one, could make non-Africans look genetically younger, on average, than the time when their earliest ancestors actually left Africa.

The Tweaks To The Story Associated With Archaic Admixture

* Neanderthal Admixture.

All modern humans who are descended from "Out of Africa" modern humans have significant traces of genetic admixture from Neanderthals (estimates have ranged from 2% to 10% with some individual and group variation).  Apart from this Neanderthal admixture sometime after leaving Africa and before reaching Southeast Asia, modern humans are not directly descended from Neanderthals who were the dominant hominin species in Europe before modern humans arrived there over a period from ca. 50,000 years ago to ca. 42,000 years ago.  Neanderthals were replaced in Europe over thousands of years of co-existence with modern humans in Europe (less in any one place) and were extinct or moribund as a species by 29,000 years ago.  The details of the timing and scale and structure of this admixture process are the subject of ongoing research (e.g. described here).

* Denisovan Admixture.

There are also genetic signs of "Denisovan" admixture in aboriginal Australians, and in indigenous Papuans, Melanesians and Polynesians, in addition to their Neanderthal admixture which they share with other non-African modern humans.  These populations have the highest known percentage of archaic admixture in their genomes of any modern human populations, but are still overwhelmingly genetic descendants of early modern human "Out of Africa" migrants.

No one archaeologically classified species of archaic hominin has been definitively identified with the non-Neanderthal archaic hominin admixture seen in the small number of modern humans that have genetic traces of Denisovan admixture that have been identified.  The Denisovan genome is based on bones found in Siberia that are too fragmentary to make an archaic hominin species identification (although the Denisovans were not modern human and not pure-blooded Neanderthals) at a place far removed from where existing modern humans showing signs of this archaic admixture are found.  It is also likely that there are traces within the Denisovan genome of archaic admixture between them and a previous archaic hominin species.

* Archaic admixture in Africa.

There are genetic signs of other kinds of archaic hominin admixture with modern humans are present in a couple of groups of Africans (one a pygmy group, and another found more widely adjacent to pygmy or former pygmy territories in tropical West Africa).  These populations still have less archaic admixture than almost all non-Africans.  No particular archaelogically classified species of archaic hominin has been definitively identified with the non-Neanderthal archaic hominin admixture seen in the small number of modern humans that have genetic traces of African non-Denisovan, non-Neanderthal admixture that have been identified.  The genetic evidence points to relatively recent admixture with relict populations of archaic hominin species who had previously not been known to exist that late in time in Africa.

The traces of admixture with archaic hominins in Africa are at much lower levels than for Neanderthal and Denisovan admixture in the relevant populations, although the African case presents the strongest evidence yet for a single Y-DNA haplogroup that introgressed from an archaic hominin into modern humans.

* Mostly, archaic admixture DNA is selectively neutral.

Immune system related HLA complex genes appear to have been the main part of the archaic admixture package outside Africa that conferred selective benefit and has left a non-trace level mark in the region's genomes.  The vast majority of archaic admixture in modern human genomes shows statistical frequencies and patterns consistent with the selective neutrality of those genes.  Any archaic admixture sourced genes that were selective fitness reducing would have vanished from the modern human genome long before the time from which our oldest available ancient DNA samples were left behind.

Regional Evolution Compared

Notably, contrary to the original "regional evolution" hypothesis, the main phenotype distinctions (i.e. visible differences) between regional populations of modern humans described as "race" do not have archaic admixture with different archaic hominins as an important source.  Serial founder effects and selective adaptation effects not related to archaic admixture are the source for most of these differences.

In short, while some of the processes associated with a regional evolution hypothesis did take place, they did not have the impact, or involve the kind of narrative, that the original proponents of the hypothesis suggested.

3 comments:

terryt said...

"The Denisovan genome is based on bones found in Siberia that are too fragmentary to make an archaic hominin species identification (although the Denisovans were not modern human and not pure-blooded Neanderthals) at a place far removed from where existing modern humans showing signs of this archaic admixture are found".

But that doesn't mean the admixture necessarily occurred close to the region where the remaining admixture is found today. Earkier in your article you wrote:

"there might have been a limited replacement or co-existence of H. Erectus and Denisovans in Zomia, Malaysia and Indonesia, a path that connects most of the dots between the Denisovan cave in Southern Siberia and the island of Flores, but the more modern lithic tools are recent enough to be attributable to Homo Sapiens as well"

And I think the connection is through Homo sapiens, not any archaic human.

"the African case presents the strongest evidence yet for a single Y-DNA haplogroup that introgressed from an archaic hominin into modern humans".

And yet modern human Y-DNA is descended from a close relative of that West African Y-DNA. To me the evidence as it stands suggests the modern Y-DNA originated in West Africa while the modern mt-DNA originated in East Africa. The big modern human expansion seems to coincide with the meeting of the two lines somewhere around Lake Chad or in Cameroon. Hybrid vigour?

andrew said...

Re Y-DNA

I think the most plausible inference is that relict archaic hominin populations endured longest in the depths of the Congo jungle, that admixture between modern human pygmies and archaic individuals happened there, that Bantu admixed with pygmies, and that this brought the archaic Y-DNA to nearby West Africa.

The archaic species that admixed with pygmies in the Congo may have had a much wider rider prior to the advent of modern humans and may have been from a species ancestral to modern humans, explaining the close relationship.

terryt said...

"I think the most plausible inference is that relict archaic hominin populations endured longest in the depths of the Congo jungle, that admixture between modern human pygmies and archaic individuals happened there, that Bantu admixed with pygmies, and that this brought the archaic Y-DNA to nearby West Africa".

I'm very much inclined to disagree with much of that. To me A00 is very much a Central/West African haplogroup in origin. Probably Gabon/Cameroon, but not 'Pygmy' as such. Although A00 has been found in Pygmy populations, and members of the A0 branch in the phylogeny, are present in the Bakola Pygmies, it is probable the haplogroup is a comparatrively recent arrival in Pygmy populations. A0 for example, is also present as far north as Algerian Berbers and both haplogroups are strongly represented in non-Pygmy populations in the Cameroons. Both A00 and A0 appear to have originated in much the same region as each other, and so are unlikely to represent separate species or even subspecies.

Even when we consider A0's brother haplogroup,A1, we find A1a is very much a West African haplogroup, although reaching further west than the previously mentioned haplogroups, to Guinea-Bissau. With A1b1 we are on firmer ground for postulating an origin further east, perhaps as far as Kenya and even reaching southern Africa. But A1b1 is a brother haplogroup to BT, members of which we can be sure were involved in the Ooa whenever it occurred. So A's emergence from West/Central Africa may not be much more ancient than the OoA.

And B itself looks to have a West African origin. B1 is concentrated in West Africa from Cameroon west to Birkina Faso and Mali. Furthermore B2(xB2a,B2b) tends to be concentrated mainly in Cameroon Pygmies. B2a and B2b are the big haplogroup B expanders, possibly along with CT and A1b1.

Even E's origin can be more convincingly placed somewhere in the western Sahel than in anywhere in East Africa. E2(xE2a,E2b) is actually spread through much of Africa but especially in the west. And Xhosa E2b is most easily explained as being associated with the Bantu expansion. Turning to E1: E1a is definitely West African while E1b is the big expander amoung the E haplogroups. But within that haplogroup E1b1a tends to be basically West/Central African. E1b1b is the only branch within E at all convincingly suggesting an easterly origin.

As a result of the basal Y-DNA distribution it seems most likely to me that the modern human Y-DNA did not leave West Africa until some time after its diversification, possibly not much before the OoA, possibly soon after it had met and mixed with a westward expanding population that contained mt-DNA haplogroup L.

"The archaic species that admixed with pygmies in the Congo may have had a much wider rider prior to the advent of modern humans and may have been from a species ancestral to modern humans, explaining the close relationship".

The archaic species seems to have admixed long after the OoA and the mixing appears to have occurred in Nigeria, somewhat west of where A00 looks most likely to have originated.